Metro Goldwyn Mayer Wiki
Metro Goldwyn Mayer Wiki
Metro-Goldwyn-Mayer Wiki
984
pages
in: Pages with template loops, All articles with dead external links, Films,
and 8 more

Raging Bull

Sign in to edit

Raging Bull is a 1980 American biographical sports drama film directed by Martin Scorsese, produced by Robert Chartoff and Irwin Winkler and adapted by Paul Schrader and Mardik Martin from Jake LaMotta's memoir Raging Bull: My Story. The film, distributed by United Artists, stars Robert De Niro as Jake LaMotta, an Italian-American middleweight boxer whose self-destructive and obsessive rage, sexual jealousy, and animalistic appetite destroyed his relationship with his wife and family. Also featured in the film are Joe Pesci as Joey, LaMotta's well-intentioned brother and manager who tries to help Jake battle his inner demons, and Cathy Moriarty as his wife. Nicholas Colasanto, Theresa Saldana, and Frank Vincent all play supporting roles in the film.

Scorsese was initially reluctant to develop the project, though he eventually came to relate to LaMotta's story. Schrader re-wrote Martin's first screenplay, and Scorsese and De Niro together made uncredited contributions thereafter. Pesci was an unknown actor prior to the film, as was Moriarty, who was suggested for her role by Pesci. During principal photography, each of the boxing scenes was choreographed for a specific visual style and De Niro gained approximately 60 pounds (27 kg) 60 pounds (27 kg) to portray LaMotta in his later post-boxing years. Scorsese was exacting in the process of editing and mixing the film, expecting it to be his last major feature.

Raging Bull premiered in New York on November 14, 1980 and was released in theaters on December 19, 1980. The film had a lukewarm box office of $23.4 million against its $18 million budget. The film received mixed reviews upon its release; while De Niro's performance and the editing were widely acclaimed, it garnered criticism due to its violent content. Despite the mixed reviews, the film was nominated for eight Academy Awards at the 53rd Academy Awards (tying with The Elephant Man as the most nominated film of the ceremony), including Best Picture and Best Director, and won two: Best Actor for De Niro (his second Oscar) and Best Editing.

After its release, Raging Bull went on to garner a high critical reputation, and is now often considered Scorsese's magnum opus and one of the greatest films ever made. In 1990, it became the first film to be selected for preservation in the National Film Registry in its first year of eligibility.

Plot[]

In 1964, an aging, overweight Italian American, Jake LaMotta, practices a comedy routine.

In 1941, LaMotta is in a major boxing match against Jimmy Reeves, where he suffered his first loss. Jake's brother, Joey LaMotta, discusses a potential shot for the middleweight title with one of his Mafia connections, Salvy Batts. Some time thereafter, Jake spots a fifteen-year-old girl named Vickie at an open-air swimming pool in his Bronx neighborhood. He eventually pursues a relationship with her, even though he is already married. In 1943, Jake defeats Sugar Ray Robinson, and has a rematch three weeks later. Despite the fact that Jake dominates Robinson during the bout, the judges surprisingly rule in favor of Robinson and Joey feels Robinson won only because he was enlisting into the Army the following week. By 1945, Jake marries Vickie.

Jake constantly worries about Vickie having feelings for other men, particularly when she makes an off-hand comment about Tony Janiro, Jake's opponent in his next fight. His jealousy is evident when he brutally defeats Janiro in front of the local Mob boss, Tommy Como, and Vickie. As Joey discusses the victory with journalists at the Copacabana, he is distracted by seeing Vickie approach a table with Salvy and his crew. Joey speaks with Vickie, who says she is giving up on his brother. Blaming Salvy, Joey viciously attacks him in a fight that spills outside of the club. Como later orders them to apologize, and has Joey tell Jake that if he wants a chance at the championship title, which Como controls, he will have to take a dive first. In a match against Billy Fox, after briefly pummeling his opponent, Jake does not even bother to put up a fight. He is suspended shortly thereafter from the board on suspicion of throwing the fight, though he realizes the error of his judgment when it is too late. He is eventually reinstated, and in 1949, wins the middleweight championship title against Marcel Cerdan.

A year later, Jake asks Joey if he fought with Salvy at the Copacabana because of Vickie. Jake then asks if Joey had an affair with her; Joey refuses to answer, insults Jake, and leaves. Jake directly asks Vickie about the affair, and when she hides from him in the bathroom, he breaks down the door, prompting her to sarcastically state that she had sex with the entire neighborhood (including his brother, Salvy, and Tommy Como). Jake angrily walks to Joey's house, with Vickie following him, and assaults Joey in front of Joey's wife Lenora and their children before knocking Vickie unconscious. After defending his championship belt in a grueling fifteen-round bout against Laurent Dauthuille in 1950, he makes a call to his brother after the fight, but when Joey assumes Salvy is on the other end and starts insulting and cursing at him, Jake says nothing and hangs up. Estranged from Joey, Jake's career begins to decline slowly and he eventually loses his title to Sugar Ray Robinson in their final encounter in 1951.

By 1956, Jake and his family have moved to Miami. After he stays out all night at his new nightclub there, Vickie tells him she wants a divorce (which she has been planning since his retirement) as well as full custody of their kids. She also threatens to call the cops if he comes anywhere near them. He is later arrested for introducing under-age girls to men in his club. He tries and fails to bribe his way out of his criminal case using the jewels from his championship belt instead of selling the belt itself. In 1957, he goes to jail, sorrowfully questioning his misfortune and crying in despair. Upon returning to New York City in 1958, he happens upon Joey, who forgives him but is elusive.

Again in 1964, Jake now recites the "I coulda been a contender" scene from the 1954 film On the Waterfront, where Terry Malloy complains that his brother should have been there for him but is also keen enough to give himself some slack. After a stagehand informs him that the auditorium where he is about to perform is crowded, Jake starts to chant "I'm the boss" while shadowboxing.

Cast[]

  • Robert De Niro as Jake LaMotta[3]
  • Joe Pesci as Joey LaMotta[3]
  • Cathy Moriarty as Vickie LaMotta[3]
  • Nicholas Colasanto as Tommy Como (based upon Frankie Carbo)[3]Template:Verify source
  • Theresa Saldana as Lenora LaMotta (Joey's wife)[3]
  • Frank Vincent as Salvy Batts[3]
  • Johnny Barnes as Sugar Ray Robinson
  • Coley Wallace as Joe Louis[4]
  • Lori Anne Flax as Irma LaMotta (Jake's previous wife)
  • Martin Scorsese as Barbizon stagehand (voice only)

Production[]

Development[]

Raging Bull came about when De Niro read the autobiography upon which the film is based on the set of The Godfather Part II. Although disappointed by the book's writing style, he became fascinated by the character of Jake LaMotta. He showed the book to Martin Scorsese on the set of Alice Doesn't Live Here Anymore in the hope that he would consider the project.[5] Scorsese repeatedly turned down the opportunity to direct the film, claiming he had no idea what Raging Bull was about, even though he had read some of the text. Never a sports fan, when he found out what LaMotta used to do for a living, he said, "A boxer? I don't like boxing...Even as a kid, I always thought that boxing was boring... It was something I couldn't, wouldn't grasp." His overall opinion of sport in general is, "Anything with a ball, no good."[6] The book was then passed onto Mardik Martin, the film's eventual co-screenwriter, who said "the trouble is the damn thing has been done a hundred times before—a fighter who has trouble with his brother and his wife and the mob is after him". The book was even shown to producers Robert Chartoff and Irwin Winkler by De Niro, who were willing to assist only if Scorsese agreed.[7] After nearly dying from a drug overdose, Scorsese agreed to make the film, not only to save his own life but also to save his career. Scorsese began to relate very personally to the story of Jake LaMotta, and in it he saw how the boxing ring can be "an allegory for whatever you do in life”, which for him paralleled moviemaking, “you make movies, you’re in the ring each time.”[8][9][10][11]

File:JakeLaMotta with DeNiro.jpg

Robert De Niro in training with the real Jake LaMotta

Preparation for the film began with Scorsese shooting some 8 mm color footage featuring De Niro boxing in a ring. One night when the footage was being shown to De Niro, Michael Chapman, and his friend and mentor, the British director Michael Powell, Powell pointed out that the color of the gloves at the time would have only been maroon, oxblood, or even black. Scorsese decided to use this as one of the reasons to film Raging Bull in black and white. Other reasons would be to distinguish the film from other color films around the time and to acknowledge the problem of fading color film stock—an issue Scorsese recognized.[12][13][14] Scorsese went to two matches at Madison Square Garden to aid his research, picking up on minor but essential details such as the blood sponge and later, the blood on the ropes (which would later be used in the film).[14] According to the brief comments on the inlay card of the Raging Bull DVD, Scorsese was not—and still is not—a fan of sports or boxing, which he describes as boring. When he saw the blood-soaked sponges being dipped in a bucket, he recalls thinking to himself, 'And they call this sport.'

Screenplay[]

Under the guidance of Chartoff and Winkler, Mardik Martin was asked to start writing the screenplay.[15] According to De Niro, under no circumstances would United Artists accept Martin's script.[16] The story was based on the vision of journalist Pete Hamill of a 1930s and 1940s style, when boxing was known as "the great dark prince of sports". De Niro was unimpressed when he finished reading the first draft, however.[17] Taxi Driver screenwriter Paul Schrader was swiftly brought in to rewrite the script around August 1978.[17] Some of the changes that Schrader made to the script included a rewrite of the scene with the uncooked steak and inclusion of LaMotta seen masturbating in a Florida cell. The character of LaMotta's brother, Joey, was finally added, previously absent from Martin's script.[16][17] United Artists saw a massive improvement on the quality of the script. However, its chief executives, Steven Bach and David Field, met with Scorsese, De Niro, and producer Irwin Winkler in November 1978 to say they were worried that the content would be X-rated material and have no chance of finding an audience.[12]

According to Scorsese, the script was left to him and De Niro, and they spent two-and-a-half weeks on the island of Saint Martin extensively re-building the content of the film.[8] The most significant change would be the entire scene when LaMotta fixes his television and then accuses his wife of having an affair. Other changes included the removal of Jake and Joey's father; the reduction of organized crime's role in the story and a major re-write of LaMotta's fight with Tony Janiro.[18][19] They were also responsible for the end sequence where LaMotta is all alone in his dressing room quoting the "I could have been a contender" scene from On the Waterfront.[19] An extract of Richard III had been considered but Michael Powell thought it would be a bad decision within the context of an American film.[8] According to Steven Bach, the first two screenwriters (Martin and Schrader) would receive credit, but since there was no payment to the writer's guild on the script, De Niro and Scorsese's work would remain uncredited.[19]

Casting[]

One of Scorsese's trademarks was casting many actors and actresses new to the profession.[20] De Niro, who was already committed to play Jake LaMotta, began to help Scorsese track down unfamiliar names to play his on-screen brother, Joey, and wife, Vikki.[21][22] The role of Joey LaMotta was the first to be cast. De Niro was watching a low budget television film called The Death Collector when he saw the part of a young career criminal played by Joe Pesci (then an unknown and struggling actor) as an ideal candidate. Prior to receiving a call from De Niro and Scorsese for the proposal to star in the film, Pesci had not worked in film for four years and was running an Italian restaurant in New Jersey.

The role of Vikki, Jake's second wife, had interest across the board, but it was Pesci who suggested actress Cathy Moriarty from a picture he once saw at a New Jersey disco.[22] Both De Niro and Scorsese believed that Moriarty could portray the role after meeting with her on several occasions and noticing her husky voice and physical maturity. The duo had to prove to the Screen Actors Guild that she was right for the role when Cis Corman showed 10 comparing pictures of both Moriarty and the real Vikki LaMotta for proof she had a resemblance.[22] Moriarty was then asked to take a screen test which she managed—partly aided with some improvised lines from De Niro—after some confusion wondering why the crew were filming her take. Joe Pesci also persuaded his former show-biz pal and co-star in The Death Collector, Frank Vincent, to try for the role of Salvy Batts. Following a successful audition and screen test, Vincent received the call to say he had received the part.[23] Charles Scorsese, the director's father, made his film debut as Tommy Como's cousin, Charlie.[23]

While in the midst of practicing a Bronx accent and preparing for his role, De Niro met both LaMotta and his ex-wife, Vikki, on separate occasions. Vikki, who lived in Florida, told stories about her life with her former husband and also showed old home movies (that later inspired a similar sequence to be done for the film).[13][24] Jake LaMotta, on the other hand, served as his trainer, accompanied by Al Silvani as coach at the Gramercy club in New York, getting him into shape. The actor found that boxing came naturally to him; he entered as a middleweight boxer, winning two of his three fights in a Brooklyn ring dubbed "young LaMotta" by the commentator. According to Jake LaMotta, De Niro was one of the top 20 best middleweight boxers of all time.[13][22]

Principal photography[]

File:Filming the boxing scenes.jpg

The filming of the boxing scenes with director, Scorsese (center left, with beard) and the director of photography, Michael Chapman (center right, with white shirt).

According to the production mixer, Michael Evje, the film began shooting at the Los Angeles Olympic Auditorium on April 16, 1979. Grips hung huge curtains of black duvetyne on all four sides of the ring area to contain the artificial smoke used extensively for visual effect. On May 7, the production moved to the Culver City Studio, Stage 3, and filmed there until the middle of June. Scorsese made it clear during filming that he did not appreciate the traditional way in films of showing fights from the spectators' view.[14] He insisted that one camera operated by the Director of Photography, Michael Chapman, would be placed inside the ring as he would play the role of an opponent keeping out of the way of other fighters so that viewers could see the emotions of the fighters, including those of Jake.[22] The precise moves of the boxers were to be done as dance routines from the information of a book about dance instructors in the mode of Arthur Murray. A punching bag in the middle of the ring was used by De Niro between takes before he aggressively came straight on to do the next scene.[22][25] The initial five-week schedule for the shooting of the boxing scenes took longer than expected, putting Scorsese under pressure.[22]

According to Scorsese, production of the film was then closed down for around four months with the entire crew being paid, so De Niro could go on a binge eating trip around northern Italy and France.[13][25] When he did come back to the United States, his weight increased from 145 to 215 pounds (66 to 97 kg).[22] The scenes with the heftier Jake LaMotta—which include announcing his retirement from boxing and LaMotta ending up in a Florida cell—were completed while approaching Christmas 1979 within seven to eight weeks so as not to aggravate the health issues which were already affecting De Niro's posture, breathing, and talking.[22][25][26]

According to Evje, Jake's nightclub sequence was filmed in a closed-down San Pedro club on December 3. The jail cell head-banging scene was shot on a constructed set with De Niro asking for minimal crew to be present—there was not even a boom operator.

The final sequence where Jake LaMotta is in front of his mirror was filmed on the last day of shooting, requiring 19 takes, with only the 13th being used for the film. Scorsese wanted to have an atmosphere that would be so cold that the words would have an impact as he tries to come to terms with his relationship with his brother.[8]

Post-production[]

The editing of Raging Bull began when production was temporarily put on hold and was completed in 1980.[25][27] Scorsese worked with the editor, Thelma Schoonmaker, to achieve a final cut of the film. Their main decision was to abandon Schrader's idea of LaMotta's nightclub act interweaving with the flashback of his youth and instead they just followed along the lines of a single flashback where only scenes of LaMotta practicing his stand-up would be left bookending the film.[28] A sound mix arranged by Frank Warner was a delicate process taking six months.[27] According to Scorsese, the sound on Raging Bull was difficult because each punch, camera shot, and flash bulb would be different. Also, there was the issue of trying to balance the quality between scenes featuring dialogue and those involving boxing (which were done in Dolby Stereo).[25] Raging Bull went through a test screening in front of a small audience including the chief executives of United Artists, Steven Bach and Andy Albeck. The screening was shown at the MGM screening room in New York around July 1980. Later, Albeck praised Scorsese by calling him a "true artist".[27]

According to the producers Robert Chartoff and Irwin Winkler, matters were made worse when United Artists decided not to distribute the film but no other studios were interested when they attempted to sell the rights.[27] Scorsese made no secret that Raging Bull would be his "Hollywood swan song" and he took unusual care of its rights during post-production.[9] Scorsese threatened to remove his credit from the film if he was not allowed to sort a reel which obscured the name of a whisky brand known as "Cutty Sark" which was heard in a scene. The work was completed only four days shy of the premiere.[29]

In 2012, Raging Bull was voted by the Motion Picture Editors Guild as the best-edited film in history.[30]

Copyright litigation[]

Paula Petrella, heir to Frank Petrello whose works were allegedly sources for the film, filed for copyright infringement in 2009 based on MGM's 1991 copyright renewal of the film. In 2014, the Supreme Court held, in Petrella v. Metro-Goldwyn-Mayer, Inc., that Petrella's suit survived MGM's defense of "laches", the legal doctrine that protects defendants from unreasonable delays by potential plaintiffs. The case was remanded to lower courts, meaning that Petrella could receive a decision on the merits of her claim.[31] MGM settled with Petrella in 2015.[32]

Reception[]

Box office[]

The brew of violence and anger, combined with the lack of a proper advertising campaign, led to the film's lukewarm box office intake of only $23 million, when compared to its $18 million budget. It only earned $10.1 million in theatrical rentals (about $27 million in 2013 dollars).[33] Scorsese became concerned for his future career and worried that producers and studios might refuse to finance his films.[34] According to Box Office Mojo, the film grossed $23,383,987 in domestic theaters.[35]

Critical reception[]

When it first premiered in New York on November 14, 1980, the initial release of Raging Bull was met with polarized reviews, but the film would later be met with much more positive reception over time.[27][28] On review aggregator Rotten Tomatoes, the film currently holds a 96% approval rating based on 68 reviews. Jack Kroll of Newsweek called Raging Bull the "best movie of the year".[27] Vincent Canby of The New York Times said that Scorsese "has made his most ambitious film as well as his finest" and went on to praise Moriarty's debut performance saying: "either she is one of the film finds of the decade or Mr. Scorsese is Svengali. Perhaps both."[34] Time praised De Niro's performance since "much of Raging Bull exists because of the possibilities it offers De Niro to display his own explosive art".[34] Steven Jenkins from the British Film Institute's (BFI) magazine, Monthly Film Journal, said "Raging Bull may prove to be Scorsese's finest achievement to date".[34]

Accolades[]

Award Category Recipient Result
Academy Awards[34][36][37] Best Picture Irwin Winkler and Robert Chartoff Template:Nom
Best Director Martin Scorsese Template:Nom
Best Actor Robert De Niro Template:Won
Best Supporting Actor Joe Pesci Template:Nom
Best Supporting Actress Cathy Moriarty Template:Nom
Best Cinematography Michael Chapman Template:Nom
Best Film Editing Thelma Schoonmaker Template:Won
Best Sound Mixing Donald O. Mitchell, Bill Nicholson,
David J. Kimball, and Les Lazarowitz 		Template:Nom
BAFTA Awards Best Actor Robert De Niro Template:Nom
Best Editing Thelma Schoonmaker Template:Won
Most Promising Newcomer Joe Pesci Template:Won
Cathy Moriarty Template:Nom
Golden Globe Awards Best Motion Picture – Drama Template:Nom
Best Actor – Motion Picture Drama Robert De Niro Template:Won
Los Angeles Film Critics Association Best Film Template:Won
Best Actor Robert De Niro Template:Won
National Board of Review Awards Top 10 Films Template:Draw
Best Actor Robert De Niro Template:Won
Best Supporting Actor Joe Pesci Template:Won
National Society of Film Critics Best Film Template:Nom
Best Director Martin Scorsese Template:Won
Best Actor Robert De Niro Template:Nom
Best Supporting Actor Joe Pesci Template:Won
Best Supporting Actress Cathy Moriarty Template:Nom
Best Cinematography Michael Chapman Template:Won

The Oscars were held the day after President Ronald Reagan was shot by John Hinckley Jr., who did it as an attempt to impress Jodie Foster, who played a child prostitute in another of Scorsese's famous films, Taxi Driver.[38] Out of fear of being attacked, Scorsese went to the ceremony with FBI bodyguards disguised as guests who escorted him out before the announcement of the Academy Award for Best Picture was made (the winner being Robert Redford's Ordinary People).

The Los Angeles Film Critics Association voted Raging Bull the best film of 1980 and De Niro best actor. The National Board of Review also voted De Niro best actor and Pesci best supporting actor. The Berlin International Film Festival chose Raging Bull to open the festival in 1981.[34]

The 2012 Parajanov-Vartanov Institute Award honored screenwriter Mardik Martin "for the mastery of his pen on iconic American films" Mean Streets and Raging Bull.[39]

Legacy[]

By the end of the 1980s, Raging Bull had cemented its reputation as a modern classic. It was voted the best film of the 1980s in numerous critics' polls and is regularly pointed to as both Scorsese's best film and one of the finest American films ever made.[40] Several prominent critics, among them Roger Ebert, declared the film to be an instant classic and the consummation of Scorsese's earlier promise. Ebert proclaimed it the best film of the 1980s,[41] and one of the ten greatest films of all time.[42] The film has been deemed "culturally, historically, and aesthetically significant" by the United States Library of Congress and was selected for preservation in the National Film Registry in 1990.[43] The film currently holds a 95% "Certified Fresh" rating on the review aggregate website Rotten Tomatoes, with an average rating of 9/10.[44] The similarly themed Metacritic rates the movie 92/100 which represents "universal acclaim".[45]

Raging Bull was listed by Time magazine as one of the All-TIME 100 Movies.[46] Variety magazine ranked the film number 39 on their list of the 50 greatest movies.[47] Raging Bull was fifth on Entertainment Weekly's list of the 100 Greatest Movies of All Time.[48] The film tied with The Bicycle Thieves and Vertigo at number 6 on Sight & SoundTemplate:'s 2002 poll of the greatest movies ever.[49] When Sight & SoundTemplate:'s directors' and critics' lists from that year are combined, Raging Bull gets the most votes of any movie that has been produced since 1975.[50] In 2002, Film4 held a poll of the 100 Greatest Movies, on which Raging Bull was voted in at number 20.[51] Halliwell's Film Guide, a British film guide, placed Raging Bull seventh in a poll naming their selection for the "Top 1,000 Movies".[52] TV Guide also included the film on their list of the 50 best movies.[53] Movieline magazine included the film on its list of the 100 best movies.[54] Leonard Maltin included Raging Bull on his 100 Must-See Films of the 20th Century list.[55] Video Detective also included the film on its list of the top 100 movies of all time.[56] Roger Ebert named "Robert De Niro's transformation from sleek boxer to paunchy nightclub owner in Raging Bull" as one of the 100 Greatest Movie Moments.[57] The National Society of Film Critics ranked it #75 on their 100 Essential Films list.[58] Rolling Stone magazine ranked it #6 on their list of the 100 Maverick Movies in the Last 100 Years.[59]

A 1997 readers poll conducted by the L.A. Daily News ranked the film #64 on a list of the greatest American movies.[60] The Writers Guild of America named the film as the 76th best screenplay of all time.[61] Raging Bull is #7 on Time Out Film Guide's "Centenary Top 100" list,[62] and it also tied at #16 (with Lawrence of Arabia) on their 1998 readers poll.[63] In 2008, Empire magazine held a poll of The 500 Greatest Movies of All Time, taking votes from 10,000 readers, 150 film makers, and 50 film critics in which Raging Bull was placed at number 11.[64] It was also placed on a similar list of 1000 movies by The New York Times.[65] In 2010, Total Film selected the film as one of The 100 Greatest Movies of All Time.[66] FilmSite.org, a subsidiary of American Movie Classics, placed Raging Bull on their list of the 100 greatest movies.[67] Additionally, Films101.com ranked the film as the 17th best movie of all time in a list of the 10,790 most notable.[68]

In 2012, the Motion Picture Editors Guild listed the film as the best-edited film of all time based on a survey of its membership.[69][70] Contemporaries of Scorsese, like Francis Ford Coppola, have included it routinely in their lists for favorite films of all time.

American Film Institute recognition[]

  • AFI's 100 Years... 100 Movies: #24[71]
  • AFI's 100 Years... 100 Thrills: #51[72]
  • AFI's 100 Years... 100 Movies (10th Anniversary Edition): #4[73]
  • AFI's 10 Top 10: #1 Sports[74]

Soundtrack[]

Martin Scorsese decided to assemble a soundtrack made of music that was popular at the time using his personal collection of 78s. With the help of Robbie Robertson, the songs were carefully chosen so they would be the ones that a person would hear on the radio, at the pool or in bars and clubs which reflected the mood of that particular era.[75][76] Some lyrics from songs would also be slipped into some dialogue. The Intermezzo from Cavalleria rusticana by Italian composer Pietro Mascagni would serve as the main theme to Raging Bull after a successful try-out by Scorsese and the editor, Thelma Schoonmaker, over the film's opening titles.[76] Two other Mascagni pieces were used in the film: the Barcarolle from Silvano, and the Intermezzo from Guglielmo Ratcliff.[77] A two-CD soundtrack was released in 2005, long after the film was released, because of earlier difficulties obtaining rights for many of the songs, which Scorsese selected from his childhood memories growing up in New York.

Proposed sequel[]

In 2006, Variety reported that Sunset Pictures was developing a sequel entitled Raging Bull II: Continuing the Story of Jake LaMotta, chronicling LaMotta's early life, as told in the sequel novel of the same name.[78] Filming began on June 15, 2012 with William Forsythe as an older LaMotta and Morjean Aria as the younger version (before the events of the first film).[79] The film, directed by Martin Guigui also stars Joe Mantegna, Tom Sizemore, Penelope Ann Miller, Natasha Henstridge, Alicia Witt, Ray Wise, Harry Hamlin, and James Russo as Rocky Graziano.[80][81] In July 2012, MGM, owners of United Artists, filed a lawsuit against LaMotta and the producers of Raging Bull II to keep the new film from being released. The former party argued that they have rights to make any authorized sequel to the original book, which goes back to an agreement LaMotta and co-author Peter Savage made with Chartoff-Winkler, producers of the original film. In addition, MGM argues that the defendants are publicly claiming the film to be a sequel to the original film, which could most likely "tarnish" its predecessor's reputation.[82] In August 2012, the producers retitled the film The Bronx Bull, disassociating itself as a sequel to Raging Bull, and the lawsuit was subsequently dropped.[83]

Notes[]

  1. ↑ Raging Bull. British Board of Film Classification. Retrieved on July 29, 2013.
  2. ↑ 2.0 2.1 Raging Bull (1980) – Financial Information. Retrieved on December 21, 2014.
  3. ↑ 3.0 3.1 3.2 3.3 3.4 3.5 Evans, Mike The Making of Raging Bull 2006 p.177.
  4. ↑ Archived copy.
  5. ↑ Biskind, Peter, Easy Riders, Raging Bulls, 1999, p. 254.
  6. ↑ Biskind, Peter Easy Riders, Raging Bulls 1998, p. 378.
  7. ↑ Biskind, Peter, Easy Riders, Raging Bulls, 1998, p. 315.
  8. ↑ 8.0 8.1 8.2 8.3 Thompson, David and Christie, Ian, Scorsese on Scorsese, pp. 76/77.
  9. ↑ 9.0 9.1 Friedman Lawrence S. The Cinema of Martin Scorsese, 1997, p. 115.
  10. ↑ Phil Villarreal. "Scorsese's 'Raging Bull' is still a knockout", The Arizona Daily Star (Tucson, AZ), February 11, 2005, p. E1.
  11. ↑ Kelly Jane Torrance. "Martin Scorsese: Telling stories through film", The Washington Times (Washington, D.C.), November 30, 2007, p. E1.
  12. ↑ 12.0 12.1 Biskind, Peter, Easy Riders, Raging Bulls, 1998, p. 389.
  13. ↑ 13.0 13.1 13.2 13.3 Total Film, The 100 greatest films of all time, pp. 180–181.
  14. ↑ 14.0 14.1 14.2 Thompson, David and Christie, Ian, Scorsese on Scorsese, p. 80.
  15. ↑ Biskind, Peter Easy Riders, Raging Bulls 1998, p. 379.
  16. ↑ 16.0 16.1 Biskind, Peter Easy Riders, Raging Bulls pp. 384–385
  17. ↑ 17.0 17.1 17.2 Baxter John De Niro A Biography, pp. 186–189.
  18. ↑ Biskind, Peter Easy Riders, Raging Bulls, p. 390.
  19. ↑ 19.0 19.1 19.2 Baxter, John De Niro A Biography, p. 193.
  20. ↑ Evans, Mike The Making of Raging Bull, p. 65.
  21. ↑ Evans, Mike The Making of Raging Bull, p. 61.
  22. ↑ 22.0 22.1 22.2 22.3 22.4 22.5 22.6 22.7 22.8 Baxter, John De Niro A Biography pp. 196–201
  23. ↑ 23.0 23.1 Evans, Mike, The Making of Raging Bull, pp. 65/66.
  24. ↑ Baxter, John, De Niro: A Biography, p. 192.
  25. ↑ 25.0 25.1 25.2 25.3 25.4 Thompson and Christie, Scorsese on Scorsese, pp. 83–84.
  26. ↑ Baxter, John, The Making of Raging Bull, p. 83.
  27. ↑ 27.0 27.1 27.2 27.3 27.4 27.5 Biskind, Peter Easy Riders, Raging Bulls 1999, p.399.
  28. ↑ 28.0 28.1 Evans, Mike The Making of Raging Bull, p. 90.
  29. ↑ Baxter, John De Niro A biography, p. 204.
  30. ↑ The 75 Best Edited Films Of All Time. IndieWire. Motion Picture Editors Guild (February 2015).
  31. ↑ Eriq Gardner, "Supreme Court Allows 'Raging Bull' Heiress to Sue MGM for Copyright Damages" --[[ ---------------------------------- Lua module implementing the Template loop detected: Template:Webarchive template. A merger of the functionality of three templates: Template:Wayback, Template:Webcite and Template:Cite archives ]] --[[--------------------------< D E P E N D E N C I E S >------------------------------------------------------ ]] require('Module:No globals'); local getArgs = require ('Module:Arguments').getArgs; --[[--------------------------< F O R W A R D D E C L A R A T I O N S >-------------------------------------- ]] local categories = {}; -- category names local config = {}; -- global configuration settings local digits = {}; -- for i18n; table that translates local-wiki digits to western digits local err_warn_msgs = {}; -- error and warning messages local excepted_pages = {}; local month_num = {}; -- for i18n; table that translates local-wiki month names to western digits local prefixes = {}; -- service provider tail string prefixes local services = {}; -- archive service provider data from local s_text = {}; -- table of static text strings used to build final rendering local uncategorized_namespaces = {}; -- list of namespaces that we should not categorize local uncategorized_subpages = {}; -- list of subpages that should not be categorized --[[--------------------------< P A G E S C O P E I D E N T I F I E R S >---------------------------------- ]] local non_western_digits; -- boolean flag set true when data.digits.enable is true local this_page = mw.title.getCurrentTitle(); local track = {}; -- Associative array to hold tracking categories local ulx = {}; -- Associative array to hold template data --[[--------------------------< S U B S T I T U T E >---------------------------------------------------------- Populates numbered arguments in a message string using an argument table. ]] local function substitute (msg, args) return args and mw.message.newRawMessage (msg, args):plain() or msg; end --[[--------------------------< tableLength >----------------------- Given a 1-D table, return number of elements ]] local function tableLength(T) local count = 0 for _ in pairs(T) do count = count + 1 end return count end --[=[-------------------------< M A K E _ W I K I L I N K >---------------------------------------------------- Makes a wikilink; when both link and display text is provided, returns a wikilink in the form D; if only link is provided, returns a wikilink in the form L; if neither are provided or link is omitted, returns an empty string. ]=] local function make_wikilink (link, display, no_link) if nil == no_link then if link and ( ~= link) then if display and ( ~= display) then return table.concat ({'', display, ''}); else return table.concat ({'', link, ''}); end end return display or ; -- link not set so return the display text else -- no_link if display and ( ~= display) then -- if there is display text return display; -- return that else return link or ; -- return the target article name or empty string end end end --[[--------------------------< createTracking >----------------------- Return data in track[] ie. tracking categories ]] local function createTracking() if not excepted_pages[this_page.fullText] then -- namespace:title/fragment is allowed to be categorized (typically this module's / template's testcases page(s)) if uncategorized_namespaces[this_page.nsText] then return ; -- this page not to be categorized so return empty string end for _,v in ipairs (uncategorized_subpages) do -- cycle through page name patterns if this_page.text:match (v) then -- test page name against each pattern return ; -- this subpage type not to be categorized so return empty string end end end local out = {}; if tableLength(track) > 0 then for key, _ in pairs(track) do -- loop through table table.insert (out, make_wikilink (key)); -- and convert category names to links end end return table.concat (out); -- concat into one big string; empty string if table is empty end --[[--------------------------< inlineError >----------------------- Critical error. Render output completely in red. Add to tracking category. This function called as the last thing before abandoning this module ]] local function inlineError (msg, args) track[categories.error] = 1 return table.concat ({ 'Error in ', -- open the error message span config.tname, -- insert the local language template name ' template: ', substitute (msg, args), -- insert the formatted error message '.', -- close the span createTracking() -- add the category }) end --[[--------------------------< inlineRed >----------------------- Render a text fragment in red, such as a warning as part of the final output. Add tracking category. ]] local function inlineRed(msg, trackmsg) if trackmsg == "warning" then track[categories.warning] = 1; elseif trackmsg == "error" then track[categories.error] = 1; end return '' .. msg .. '' end --[[--------------------------< base62 >----------------------- Convert base-62 to base-10 Credit: https://de.wikipedia.org/wiki/Modul:Expr ]] local function base62( value ) local r = 1 -- default return value is input value is malformed if value:match ('%W') then -- value must only be in the set [0-9a-zA-Z] return; -- nil return when value contains extraneous characters end local n = #value -- number of characters in value local k = 1 local c r = 0 for i = n, 1, -1 do -- loop through all characters in value from ls digit to ms digit c = value:byte( i, i ) if c >= 48 and c <= 57 then -- character is digit 0-9 c = c - 48 elseif c >= 65 and c <= 90 then -- character is ascii a-z c = c - 55 else -- must be ascii A-Z c = c - 61 end r = r + c * k -- accumulate this base62 character's value k = k * 62 -- bump for next end -- for i return r end --[[--------------------------< D E C O D E _ D A T E >-------------------------------------------------------- Given a date string, return it in iso format along with an indicator of the date's format. Except that month names must be recognizable as legitimate month names with proper capitalization, and that the date string must match one of the recognized date formats, no error checking is done here; return nil else ]] local function decode_date (date_str) local patterns = { ['dmy'] = {'^(%d%d?) +([^%s%d]+) +(%d%d%d%d)$', 'd', 'm', 'y'}, -- %a does not recognize unicode combining characters used by some languages ['mdy'] = {'^([^%s%d]+) (%d%d?), +(%d%d%d%d)$', 'm', 'd', 'y'}, ['ymd'] = {'^(%d%d%d%d) +([^%s%d]+) (%d%d?)$', 'y', 'm', 'd'}, -- not mos compliant at en.wiki but may be acceptible at other wikis }; local t = {}; if non_western_digits then -- this wiki uses non-western digits? date_str = mw.ustring.gsub (date_str, '%d', digits); -- convert this wiki's non-western digits to western digits end if date_str:match ('^%d%d%d%d%-%d%d%-%d%d$') then -- already an iso format date, return western digits form return date_str, 'iso'; end for k, v in pairs (patterns) do local c1, c2, c3 = mw.ustring.match (date_str, patterns[k][1]); -- c1 .. c3 are captured but we don't know what they hold if c1 then -- set on match t = { -- translate unspecified captures to y, m, and d [patterns[k][2]] = c1, -- fill the table of captures with the captures [patterns[k][3]] = c2, -- take index names from src_pattern table and assign sequential captures [patterns[k][4]] = c3, }; if month_num[t.m] then -- when month not already a number t.m = month_num[t.m]; -- replace valid month name with a number else return nil, 'iso'; -- not a valid date form because month not valid end return mw.ustring.format ('%.4d-%.2d-%.2d', t.y, t.m, t.d), k; -- return date in iso format end end return nil, 'iso'; -- date could not be decoded; return nil and default iso date end --[[--------------------------< makeDate >----------------------- Given year, month, day numbers, (zero-padded or not) return a full date in df format where df may be one of: mdy, dmy, iso, ymd on entry, year, month, day are presumed to be correct for the date that they represent; all are required in this module, makeDate() is sometimes given an iso-format date in year: makeDate (2018-09-20, nil, nil, df) this works because table.concat() sees only one table member ]] local function makeDate (year, month, day, df) local format = { ['dmy'] = 'j F Y', ['mdy'] = 'F j, Y', ['ymd'] = 'Y F j', ['iso'] = 'Y-m-d', }; local date = table.concat ({year, month, day}, '-'); -- assemble year-initial numeric-format date (zero padding not required here) if non_western_digits then -- this wiki uses non-western digits? date = mw.ustring.gsub (date, '%d', digits); -- convert this wiki's non-western digits to western digits end return mw.getContentLanguage():formatDate (format[df], date); end --[[--------------------------< I S _ V A L I D _ D A T E >---------------------------------------------------- Returns true if date is after 31 December 1899 (why is 1900 the min year? shouldn't the internet's date-of-birth be min year?), not after today's date, and represents a valid date (29 February 2017 is not a valid date). Applies Gregorian leapyear rules. all arguments are required ]] local function is_valid_date (year, month, day) local days_in_month = {31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31}; local month_length; local y, m, d; local today = os.date ('*t'); -- fetch a table of current date parts if not year or == year or not month or == month or not day or == day then return false; -- something missing end y = tonumber (year); m = tonumber (month); d = tonumber (day); if 1900 > y or today.year < y or 1 > m or 12 < m then -- year and month are within bounds TODO: 1900? return false; end if (2==m) then -- if February month_length = 28; -- then 28 days unless if (0==(y%4) and (0~=(y%100) or 0==(y%400))) then -- is a leap year? month_length = 29; -- if leap year then 29 days in February end else month_length=days_in_month[m]; end if 1 > d or month_length < d then -- day is within bounds return false; end -- here when date parts represent a valid date return os.time({['year']=y, ['month']=m, ['day']=d, ['hour']=0}) <= os.time(); -- date at midnight must be less than or equal to current date/time end --[[--------------------------< decodeWebciteDate >----------------------- Given a URI-path to Webcite (eg. /67xHmVFWP) return the encoded date in df format returns date string in df format - webcite date is a unix timestamp encoded as bae62 or the string 'query' ]] local function decodeWebciteDate(path, df) local dt = {}; local decode; dt = mw.text.split(path, "/") -- valid URL formats that are not base62 -- http://www.webcitation.org/query?id=1138911916587475 -- http://www.webcitation.org/query?url=http..&date=2012-06-01+21:40:03 -- http://www.webcitation.org/1138911916587475 -- http://www.webcitation.org/cache/73e53dd1f16cf8c5da298418d2a6e452870cf50e -- http://www.webcitation.org/getfile.php?fileid=1c46e791d68e89e12d0c2532cc3cf629b8bc8c8e if dt[2]:find ('query', 1, true) or dt[2]:find ('cache', 1, true) or dt[2]:find ('getfile', 1, true) or tonumber(dt[2]) then return 'query'; end decode = base62(dt[2]); -- base62 string -> exponential number if not decode then return nil; -- nil return when dt[2] contains characters not in %w end dt = os.date('*t', string.format("%d", decode):sub(1,10)) -- exponential number -> text -> first 10 characters (a unix timestamp) -> a table of date parts decode = makeDate (dt.year, dt.month, dt.day, 'iso'); -- date comparisons are all done in iso format with western digits if non_western_digits then -- this wiki uses non-western digits? decode = mw.ustring.gsub (decode, '%d', digits); -- convert this wiki's non-western digits to western digits end return decode; end --[[--------------------------< decodeWaybackDate >----------------------- Given a URI-path to Wayback (eg. /web/20160901010101/http://example.com ) or Library of Congress Web Archives (eg. /all/20160901010101/http://example.com) or UK Government Web Archive (eg. /ukgwa/20160901010101/http://example.com or /tna/20160901010101/http://example.com) return the formatted date eg. "September 1, 2016" in df format Handle non-digits in snapshot ID such as "re_" and "-" and "*" returns two values: first value is one of these: valid date string in df format - wayback date is valid (including the text string 'index' when date is '/*/') empty string - wayback date is malformed (less than 8 digits, not a valid date) nil - wayback date is '/save/' or otherwise not a number second return value is an appropriate 'message' may or may not be formatted ]] local function decodeWaybackDate(path, df) local msg, snapdate; snapdate = path:gsub ('^/web/', ):gsub ('^/all/', ):gsub ('^/ukgwa/', ):gsub ('^/tna/', ):gsub ('^/', ); -- remove leading /web/, /all/, /ukgwa/, /tna/, or / snapdate = snapdate:match ('^[^/]+'); -- get timestamp if snapdate == "*" then -- eg. /web/*/http.., etc. return 'index'; -- return indicator that this url has an index date end snapdate = snapdate:gsub ('%a%a_%d?$', ):gsub ('%-', ); -- from date, remove any trailing "re_", dashes msg = ; if snapdate:match ('%*$') then -- a trailing '*' causes calendar display at archive .org snapdate = snapdate:gsub ('%*$', ); -- remove so not part of length calc later msg = inlineRed (err_warn_msgs.ts_cal, 'warning'); -- make a message end if not tonumber(snapdate) then return nil, 'ts_nan'; -- return nil (fatal error flag) and message selector end local dlen = snapdate:len(); if dlen < 8 then -- we need 8 digits TODO: but shouldn't this be testing for 14 digits? return , inlineRed (err_warn_msgs.ts_short, 'error'); -- return empty string and error message end local year, month, day = snapdate:match ('(%d%d%d%d)(%d%d)(%d%d)'); -- no need for snapdatelong here if not is_valid_date (year, month, day) then return , inlineRed (err_warn_msgs.ts_date, 'error'); -- return empty string and error message end snapdate = table.concat ({year, month, day}, '-'); -- date comparisons are all done in iso format if 14 == dlen then return snapdate, msg; -- return date with message if any else return snapdate, msg .. inlineRed (err_warn_msgs.ts_len, 'warning'); -- return date with warning message(s) end end --[[--------------------------< decodeArchiveisDate >----------------------- Given an Archive.is "long link" URI-path (e.g. /2016.08.28-144552/http://example.com) return the date in df format (e.g. if df = dmy, return 28 August 2016) Handles "." and "-" in snapshot date, so 2016.08.28-144552 is same as 20160828144552 returns two values: first value is one of these: valid date string in df format - archive.is date is valid (including the text string 'short link' when url is the short form) empty string - wayback date is malformed (not a number, less than 8 digits, not a valid date) nil - wayback date is '/save/' second return value is an appropriate 'message' may or may not be formatted ]] local function decodeArchiveisDate(path, df) local snapdate if path:match ('^/%w+$') then -- short form url path is '/' followed by some number of base 62 digits and nothing else return "short link" -- e.g. http://archive.is/hD1qz end snapdate = mw.text.split (path, '/')[2]:gsub('[%.%-]', ); -- get snapshot date, e.g. 2016.08.28-144552; remove periods and hyphens local dlen = string.len(snapdate) if dlen < 8 then -- we need 8 digits TODO: but shouldn't this be testing for 14 digits? return , inlineRed (err_warn_msgs.ts_short, 'error'); -- return empty string and error message end local year, month, day = snapdate:match ('(%d%d%d%d)(%d%d)(%d%d)'); -- no need for snapdatelong here if not is_valid_date (year, month, day) then return , inlineRed (err_warn_msgs.ts_date, 'error'); -- return empty string and error message end snapdate = table.concat ({year, month, day}, '-'); -- date comparisons are all done in iso format if 14 == dlen then return snapdate; -- return date else return snapdate, inlineRed (err_warn_msgs.ts_len, 'warning'); -- return date with warning message end end --[[--------------------------< serviceName >----------------------- Given a domain extracted by mw.uri.new() (eg. web.archive.org) set tail string and service ID ]] local function serviceName(host, no_link) local tracking; local index; host = host:lower():gsub ('^web%.(.+)', '%1'):gsub ('^www%.(.+)', '%1'); -- lowercase, remove web. and www. subdomains if services[host] then index = host; else for k, _ in pairs (services) do if host:find ('%f[%a]'..k:gsub ('([%.%-])', '%%%1')) then index = k; break; end end end if index then local out = {}; -- empty string in [1] so that concatenated result has leading single space ulx.url1.service = services[index][4] or 'other'; tracking = services[index][5] or categories.other; -- build tail string if false == services[index][1] then -- select prefix table.insert (out, prefixes.at); elseif true == services[index][1] then table.insert (out, prefixes.atthe); else table.insert (out, services[index][1]); end table.insert (out, make_wikilink (services[index][2], services[index][3], no_link)); -- add article wikilink if services[index][6] then -- add tail postfix if it exists table.insert (out, services[index][6]); end ulx.url1.tail = table.concat (out, ' '); -- put it all together; result has leading space character else -- here when unknown archive ulx.url1.service = 'other'; tracking = categories.unknown; ulx.url1.tail = table.concat ({, prefixes.at, host, inlineRed (err_warn_msgs.unknown_url, error)}, ' '); end track[tracking] = 1 end --[[--------------------------< parseExtraArgs >----------------------- Parse numbered arguments starting at 2, such as url2..url10, date2..date10, title2..title10 For example: Template loop detected: Template:Webarchive Three url arguments not in numeric sequence (1..4..7). Function only processes arguments numbered 2 or greater (in this case 4 and 7) It creates numeric sequenced table entries like: urlx.url2.url = <argument value for url4> urlx.url3.url = <argument value for url7> Returns the number of URL arguments found numbered 2 or greater (in this case returns "2") ]] local function parseExtraArgs(args) local i, j, argurl, argurl2, argdate, argtitle j = 2 for i = 2, config.maxurls do argurl = "url" .. i if args[argurl] then argurl2 = "url" .. j ulx[argurl2] = {} ulx[argurl2]["url"] = args[argurl] argdate = "date" .. i if args[argdate] then ulx[argurl2]["date"] = args[argdate] else ulx[argurl2]["date"] = inlineRed (err_warn_msgs.date_miss, 'warning'); end argtitle = "title" .. i if args[argtitle] then ulx[argurl2]["title"] = args[argtitle] else ulx[argurl2]["title"] = nil end j = j + 1 end end if j == 2 then return 0 else return j - 2 end end --[[--------------------------< comma >----------------------- Given a date string, return "," if it's MDY ]] local function comma(date) return (date and date:match ('%a+ +%d%d?(,) +%d%d%d%d')) or ; end --[[--------------------------< createRendering >----------------------- Return a rendering of the data in ulx[][] ]] local function createRendering() local displayfield local out = {}; local index_date, msg = ulx.url1.date:match ('(index)(.*)'); -- when ulx.url1.date extract 'index' text and message text (if there is a message) ulx.url1.date = ulx.url1.date:gsub ('index.*', 'index'); -- remove message if 'none' == ulx.url1.format then -- For Template:Wayback, Template:Webcite table.insert (out, '['); -- open extlink markup table.insert (out, ulx.url1.url); -- add url if ulx.url1.title then table.insert (out, ' ') -- the required space table.insert (out, ulx.url1.title) -- the title table.insert (out, ']'); -- close extlink markup table.insert (out, ulx.url1.tail); -- tail text if ulx.url1.date then table.insert (out, ' ('); -- open date text; TODO: why the html entity? replace with regular space? table.insert (out, 'index' == ulx.url1.date and s_text.archive or s_text.archived); -- add text table.insert (out, ' '); -- insert a space table.insert (out, ulx.url1.date); -- add date table.insert (out, ')'); -- close date text end else -- no title if index_date then -- when url date is 'index' table.insert (out, table.concat ({' ', s_text.Archive_index, ']'})); -- add the index link label table.insert (out, msg or ); -- add date mismatch message when url date is /*/ and |date= has valid date else table.insert (out, table.concat ({' ', s_text.Archived, '] '})); -- add link label for url has timestamp date (will include mismatch message if there is one) end if ulx.url1.date then if 'index' ~= ulx.url1.date then table.insert (out, ulx.url1.date); -- add date when data is not 'index' end table.insert (out, comma(ulx.url1.date)); -- add ',' if date format is mdy table.insert (out, ulx.url1.tail); -- add tail text else -- no date table.insert (out, ulx.url1.tail); -- add tail text end end if 0 < ulx.url1.extraurls then -- For multiple archive URLs local tot = ulx.url1.extraurls + 1 table.insert (out, '.') -- terminate first url table.insert (out, table.concat ({' ', s_text.addlarchives, ': '})); -- add header text for i=2, tot do -- loop through the additionals local index = table.concat ({'url', i}); -- make an index displayfield = ulx[index]['title'] and 'title' or 'date'; -- choose display text table.insert (out, '['); -- open extlink markup table.insert (out, ulx[index]['url']); -- add the url table.insert (out, ' '); -- the required space table.insert (out, ulx[index][displayfield]); -- add the label table.insert (out, ']'); -- close extlink markup table.insert (out, i==tot and '.' or ', '); -- add terminator end end return table.concat (out); -- make a big string and done else -- For Template:Cite archives if 'addlarchives' == ulx.url1.format then -- Multiple archive services table.insert (out, table.concat ({s_text.addlarchives, ': '})); -- add header text else -- Multiple pages from the same archive table.insert (out, table.concat ({s_text.addlpages, ' '})); -- add header text table.insert (out, ulx.url1.date); -- add date to header text table.insert (out, ': '); -- close header text end local tot = ulx.url1.extraurls + 1; for i=1, tot do -- loop through the additionals local index = table.concat ({'url', i}); -- make an index table.insert (out, '['); -- open extlink markup table.insert (out, ulx[index]['url']); -- add url table.insert (out, ' '); -- add required space displayfield = ulx[index]['title']; if 'addlarchives' == ulx.url1.format then if not displayfield then displayfield = ulx[index]['date'] end else -- must be addlpages if not displayfield then displayfield = table.concat ({s_text.Page, ' ', i}); end end table.insert (out, displayfield); -- add title, date, page label text table.insert (out, ']'); -- close extlink markup table.insert (out, (i==tot and '.' or ', ')); -- add terminator end return table.concat (out); -- make a big string and done end end --[[--------------------------< P A R A M E T E R _ N A M E _ X L A T E >-------------------------------------- for internaltionalization, translate local-language parameter names to their English equivalents TODO: return error message if multiple aliases of the same canonical parameter name are found? returns two tables: new_args - holds canonical form parameters and their values either from translation or because the parameter was already in canonical form origin - maps canonical-form parameter names to their untranslated (local language) form for error messaging in the local language unrecognized parameters are ignored ]] local function parameter_name_xlate (args, params, enum_params) local name; -- holds modifiable name of the parameter name during evaluation local enum; -- for enumerated parameters, holds the enumerator during evaluation local found = false; -- flag used to break out of nested for loops local new_args = {}; -- a table that holds canonical and translated parameter k/v pairs local origin = {}; -- a table that maps original (local language) parameter names to their canonical name for local language error messaging local unnamed_params; -- set true when unsupported positional parameters are detected for k, v in pairs (args) do -- loop through all of the arguments in the args table name = k; -- copy of original parameter name if 'string' == type (k) then if non_western_digits then -- true when non-western digits supported at this wiki name = mw.ustring.gsub (name, '%d', digits); -- convert this wiki's non-western digits to western digits end enum = name:match ('%d+$'); -- get parameter enumerator if it exists; nil else if not enum then -- no enumerator so looking for non-enumnerated parameters -- TODO: insert shortcut here? if params[name] then name holds the canonical parameter name; no need to search further for pname, aliases in pairs (params) do -- loop through each parameter the params table for _, alias in ipairs (aliases) do -- loop through each alias in the parameter's aliases table if name == alias then new_args[pname] = v; -- create a new entry in the new_args table origin [pname] = k; -- create an entry to make canonical parameter name to original local language parameter name found = true; -- flag so that we can break out of these nested for loops break; -- no need to search the rest of the aliases table for name so go on to the next k, v pair end end if found then -- true when we found an alias that matched name found = false; -- reset the flag break; -- go do next args k/v pair end end else -- enumerated parameters name = name:gsub ('%d$', '#'); -- replace enumeration digits with place holder for table search -- TODO: insert shortcut here? if num_params[name] then name holds the canonical parameter name; no need to search further for pname, aliases in pairs (enum_params) do -- loop through each parameter the num_params table for _, alias in ipairs (aliases) do -- loop through each alias in the parameter's aliases table if name == alias then pname = pname:gsub ('#$', enum); -- replace the '#' place holder with the actual enumerator new_args[pname] = v; -- create a new entry in the new_args table origin [pname] = k; -- create an entry to make canonical parameter name to original local language parameter name found = true; -- flag so that we can break out of these nested for loops break; -- no need to search the rest of the aliases table for name so go on to the next k, v pair end end if found then -- true when we found an alias that matched name found = false; -- reset the flag break; -- go do next args k/v pair end end end else unnamed_params = true; -- flag for unsupported positional parameters end end -- for k, v return new_args, origin, unnamed_params; end --[[--------------------------< W E B A R C H I V E >---------------------------------------------------------- template entry point ]] local function webarchive(frame) local args = getArgs (frame); local data = mw.loadData (table.concat ({ -- make a data module name; sandbox or live 'Module:Webarchive/data', frame:getTitle():find('sandbox', 1, true) and '/sandbox' or -- this instance is ./sandbox then append /sandbox })); categories = data.categories; -- fill in the forward declarations config = data.config; if data.digits.enable then digits = data.digits; -- for i18n; table of digits in the local wiki's language non_western_digits = true; -- use_non_western_digits end err_warn_msgs = data.err_warn_msgs; excepted_pages = data.excepted_pages; month_num = data.month_num; -- for i18n; table of month names in the local wiki's language prefixes = data.prefixes; services = data.services; s_text = data.s_text; uncategorized_namespaces = data.uncategorized_namespaces; uncategorized_subpages = data.uncategorized_subpages; local origin = {}; -- holds a map of English to local language parameter names used in the current template; not currently used local unnamed_params; -- boolean set to true when template call has unnamed parameters args, origin, unnamed_params = parameter_name_xlate (args, data.params, data.enum_params); -- translate parameter names in args to English local date, format, msg, udate, uri, url; local ldf = 'iso'; -- when there is no |date= parameter, render url dates in iso format if args.url and args.url1 then -- URL argument (first) return inlineError (data.crit_err_msgs.conflicting, {origin.url, origin.url1}); end url = args.url or args.url1; if not url then return inlineError (data.crit_err_msgs.empty); end -- these iabot bugs perportedly fixed; removing these causes lua script error --[[ -- at Template:Webarchive/testcases/Production; resolve that before deleting these tests if mw.ustring.find( url, "https://web.http", 1, true ) then -- track bug - TODO: IAbot bug; not known if the bug has been fixed; deferred track[categories.error] = 1; return inlineError (data.crit_err_msgs.iabot1); end if url == "https://web.archive.org/http:/" then -- track bug - TODO: IAbot bug; not known if the bug has been fixed; deferred track[categories.error] = 1; return inlineError (data.crit_err_msgs.iabot2); end ]] if not (url:lower():find ('^http') or url:find ('^//')) then return inlineError (data.crit_err_msgs.invalid_url ); end ulx.url1 = {} ulx.url1.url = url ulx.url1.extraurls = parseExtraArgs(args) local good = false; good, uri = pcall (mw.uri.new, ulx.url1.url); -- get a table of uri parts from this url; protected mode to prevent lua error when ulx.url1.url is malformed if not good or nil == uri.host then -- abandon when ulx.url1.url is malformed return inlineError (data.crit_err_msgs.invalid_url); end serviceName(uri.host, args.nolink) if args.date and args.date1 then -- Date argument return inlineError (data.crit_err_msgs.conflicting, {origin.date, origin.date1}); end date = args.date or args.date1; date = date and date:gsub (' +', ' '); -- replace multiple spaces with a single space if date and config.verifydates then if '*' == date then date = 'index'; ldf = 'iso'; -- set to default format elseif 'mdy' == date then date = nil; -- if date extracted from URL, ldf = 'mdy'; -- then |date=mdy overrides iso elseif 'dmy' == date then date = nil; -- if date extracted from URL, ldf = 'dmy'; -- then |date=dmy overrides iso elseif 'ymd' == date then date = nil; -- if date extracted from URL, ldf = 'ymd'; -- then |date=ymd overrides iso else date, ldf = decode_date (date); -- get an iso format date from date and get date's original format end end if 'wayback' == ulx.url1.service or 'locwebarchives' == ulx.url1.service or 'ukgwa' == ulx.url1.service then if date then if config.verifydates then if ldf then udate, msg = decodeWaybackDate (uri.path); -- get the url date in iso format and format of date in |date=; 'index' when wayback url date is * if not udate then -- this is the only 'fatal' error return return inlineError (data.crit_err_msgs[msg]); end if udate ~= date then -- date comparison using iso format dates date = udate; msg = table.concat ({ inlineRed (err_warn_msgs.mismatch, 'warning'), -- add warning message msg, -- add message if there is one }); end end end else -- no |date= udate, msg = decodeWaybackDate (uri.path); if not udate then -- this is the only 'fatal' error return return inlineError (data.crit_err_msgs[msg]); end if == udate then date = nil; -- unset else date = udate; end end elseif 'webcite' == ulx.url1.service then if date then if config.verifydates then if ldf then udate = decodeWebciteDate (uri.path); -- get the url date in iso format if 'query' ~= udate then -- skip if query if udate ~= date then -- date comparison using iso format dates date = udate; msg = table.concat ({ inlineRed (err_warn_msgs.mismatch, 'warning'), }); end end end end else date = decodeWebciteDate( uri.path, "iso" ) if date == "query" then date = nil; -- unset msg = inlineRed (err_warn_msgs.date_miss, 'warning'); elseif not date then -- invalid base62 string date = inlineRed (err_warn_msgs.date1, 'error'); end end elseif 'archiveis' == ulx.url1.service then if date then if config.verifydates then if ldf then udate, msg = decodeArchiveisDate (uri.path) -- get the url date in iso format if 'short link' ~= udate then -- skip if short link if udate ~= date then -- date comparison using iso format dates date = udate; msg = table.concat ({ inlineRed (err_warn_msgs.mismatch, 'warning'), -- add warning message msg, -- add message if there is one }); end end end end else -- no |date= udate, msg = decodeArchiveisDate( uri.path, "iso" ) if udate == "short link" then date = nil; -- unset msg = inlineRed (err_warn_msgs.date_miss, 'warning'); elseif == udate then date = nil; -- unset else date = udate; end end else -- some other service if not date then msg = inlineRed (err_warn_msgs.date_miss, 'warning'); end end if 'index' == date then ulx.url1.date = date .. (msg or ); -- create index + message (if there is one) elseif date then ulx.url1.date = makeDate (date, nil, nil, ldf) .. (msg or ); -- create a date in the wiki's local language + message (if there is one) else ulx.url1.date = msg; end format = args.format; -- Format argument if not format then format = "none" else for k, v in pairs (data.format_vals) do -- |format= accepts two specific values loop through a table of those values local found; -- declare a nil flag for _, p in ipairs (v) do -- loop through local language variants if format == p then -- when |format= value matches format = k; -- use name from table key found = true; -- declare found so that we can break out of outer for loop break; -- break out of inner for loop end end if found then break; end end if format == "addlpages" then if not ulx.url1.date then format = "none" end elseif format == "addlarchives" then format = "addlarchives" else format = "none" end end ulx.url1.format = format if args.title and args.title1 then -- Title argument return inlineError (data.crit_err_msgs.conflicting, {origin.title, origin.title1}); end ulx.url1.title = args.title or args.title1; local rend = createRendering() if not rend then return inlineError (data.crit_err_msgs.unknown); end return rend .. ((unnamed_params and inlineRed (err_warn_msgs.unnamed_params, 'warning')) or ) .. createTracking(); end --[[--------------------------< E X P O R T E D F U N C T I O N S >------------------------------------------ ]] return {webarchive = webarchive};, The Hollywood Reporter, May 19, 2014.
  32. ↑ Gardner, Eriq. "After Supreme Court, MGM Settles 'Raging Bull' Rights Dispute", The Hollywood Reporter, April 5, 2015. 
  33. ↑ Spy (Nov 1988). The Unstoppables. New York: Sussex Publishers, LLC, page 90. 
  34. ↑ 34.0 34.1 34.2 34.3 34.4 34.5 Evans, Mike The Making of Raging Bull pp. 124–129
  35. ↑ Raging Bull. Retrieved on October 17, 2008.
  36. ↑ The 53rd Academy Awards (1981) Nominees and Winners. oscars.org. Retrieved on October 7, 2011.
  37. ↑ Raging Bull – Academy Awards Database. AMPAS. Retrieved on June 28, 2007.
  38. ↑ John Hinckley, Jr (n.d.). Archived from the original on January 19, 2011. Retrieved on December 23, 2010.
  39. ↑ Parajanov-Vartanov Institute | Official site (en-US).[dead link]
  40. ↑ Walker, John Halliwell's Top 1000, The Ultimate Movie Countdown 2005, p. 561.
  41. ↑ "Ebert's 10 Best Lists: 1967–present". Retrieved on October 23, 2010. 
  42. ↑ "Roger Ebert's Ten Greatest Films of All Time". Retrieved on October 23, 2010. 
  43. ↑ Gamarekian, Barbara. "Library of Congress Adds 25 Titles to National Film Registry", The New York Times, October 19, 1990. Retrieved on July 22, 2009. 
  44. ↑ Raging Bull Movie Reviews, Pictures. Retrieved on October 23, 2010.
  45. ↑ Raging Bull Movie Reviews. Retrieved on October 23, 2010.
  46. ↑ Template:Cite magazine
  47. ↑ Thompson, Anne. "Lists: 50 Best Movies of All Time, Again", Variety. Retrieved on October 23, 2010. 
  48. ↑ Entertainment Weekly's 100 Greatest Movies of All Time. Entertainment Weekly. Published by AMC FilmSite.org. Retrieved on October 23, 2010.
  49. ↑ Top Ten Poll 2002 – Directors' Poll. Sight & Sound. British Film Institute. Retrieved on October 23, 2010.
  50. ↑ 2002 Sight & Sound Film Survey. cinemacom.com.
  51. ↑ FilmFour. Film Four's 100 Greatest Films of All Time. AMC FilmSite.org. Retrieved on October 23, 2010.
  52. ↑ Halliwell's Top 1000 Movies. Mindjack Film. Retrieved on October 23, 2010.
  53. ↑ "50 Greatest Movies (on TV and Video)", AMC's FilmSite. Retrieved on August 15, 2017. 
  54. ↑ The 100 Best Movies Ever Made. --[[ ---------------------------------- Lua module implementing the Template loop detected: Template:Webarchive template. A merger of the functionality of three templates: Template:Wayback, Template:Webcite and Template:Cite archives ]] --[[--------------------------< D E P E N D E N C I E S >------------------------------------------------------ ]] require('Module:No globals'); local getArgs = require ('Module:Arguments').getArgs; --[[--------------------------< F O R W A R D D E C L A R A T I O N S >-------------------------------------- ]] local categories = {}; -- category names local config = {}; -- global configuration settings local digits = {}; -- for i18n; table that translates local-wiki digits to western digits local err_warn_msgs = {}; -- error and warning messages local excepted_pages = {}; local month_num = {}; -- for i18n; table that translates local-wiki month names to western digits local prefixes = {}; -- service provider tail string prefixes local services = {}; -- archive service provider data from local s_text = {}; -- table of static text strings used to build final rendering local uncategorized_namespaces = {}; -- list of namespaces that we should not categorize local uncategorized_subpages = {}; -- list of subpages that should not be categorized --[[--------------------------< P A G E S C O P E I D E N T I F I E R S >---------------------------------- ]] local non_western_digits; -- boolean flag set true when data.digits.enable is true local this_page = mw.title.getCurrentTitle(); local track = {}; -- Associative array to hold tracking categories local ulx = {}; -- Associative array to hold template data --[[--------------------------< S U B S T I T U T E >---------------------------------------------------------- Populates numbered arguments in a message string using an argument table. ]] local function substitute (msg, args) return args and mw.message.newRawMessage (msg, args):plain() or msg; end --[[--------------------------< tableLength >----------------------- Given a 1-D table, return number of elements ]] local function tableLength(T) local count = 0 for _ in pairs(T) do count = count + 1 end return count end --[=[-------------------------< M A K E _ W I K I L I N K >---------------------------------------------------- Makes a wikilink; when both link and display text is provided, returns a wikilink in the form D; if only link is provided, returns a wikilink in the form L; if neither are provided or link is omitted, returns an empty string. ]=] local function make_wikilink (link, display, no_link) if nil == no_link then if link and ( ~= link) then if display and ( ~= display) then return table.concat ({'', display, ''}); else return table.concat ({'', link, ''}); end end return display or ; -- link not set so return the display text else -- no_link if display and ( ~= display) then -- if there is display text return display; -- return that else return link or ; -- return the target article name or empty string end end end --[[--------------------------< createTracking >----------------------- Return data in track[] ie. tracking categories ]] local function createTracking() if not excepted_pages[this_page.fullText] then -- namespace:title/fragment is allowed to be categorized (typically this module's / template's testcases page(s)) if uncategorized_namespaces[this_page.nsText] then return ; -- this page not to be categorized so return empty string end for _,v in ipairs (uncategorized_subpages) do -- cycle through page name patterns if this_page.text:match (v) then -- test page name against each pattern return ; -- this subpage type not to be categorized so return empty string end end end local out = {}; if tableLength(track) > 0 then for key, _ in pairs(track) do -- loop through table table.insert (out, make_wikilink (key)); -- and convert category names to links end end return table.concat (out); -- concat into one big string; empty string if table is empty end --[[--------------------------< inlineError >----------------------- Critical error. Render output completely in red. Add to tracking category. This function called as the last thing before abandoning this module ]] local function inlineError (msg, args) track[categories.error] = 1 return table.concat ({ 'Error in ', -- open the error message span config.tname, -- insert the local language template name ' template: ', substitute (msg, args), -- insert the formatted error message '.', -- close the span createTracking() -- add the category }) end --[[--------------------------< inlineRed >----------------------- Render a text fragment in red, such as a warning as part of the final output. Add tracking category. ]] local function inlineRed(msg, trackmsg) if trackmsg == "warning" then track[categories.warning] = 1; elseif trackmsg == "error" then track[categories.error] = 1; end return '' .. msg .. '' end --[[--------------------------< base62 >----------------------- Convert base-62 to base-10 Credit: https://de.wikipedia.org/wiki/Modul:Expr ]] local function base62( value ) local r = 1 -- default return value is input value is malformed if value:match ('%W') then -- value must only be in the set [0-9a-zA-Z] return; -- nil return when value contains extraneous characters end local n = #value -- number of characters in value local k = 1 local c r = 0 for i = n, 1, -1 do -- loop through all characters in value from ls digit to ms digit c = value:byte( i, i ) if c >= 48 and c <= 57 then -- character is digit 0-9 c = c - 48 elseif c >= 65 and c <= 90 then -- character is ascii a-z c = c - 55 else -- must be ascii A-Z c = c - 61 end r = r + c * k -- accumulate this base62 character's value k = k * 62 -- bump for next end -- for i return r end --[[--------------------------< D E C O D E _ D A T E >-------------------------------------------------------- Given a date string, return it in iso format along with an indicator of the date's format. Except that month names must be recognizable as legitimate month names with proper capitalization, and that the date string must match one of the recognized date formats, no error checking is done here; return nil else ]] local function decode_date (date_str) local patterns = { ['dmy'] = {'^(%d%d?) +([^%s%d]+) +(%d%d%d%d)$', 'd', 'm', 'y'}, -- %a does not recognize unicode combining characters used by some languages ['mdy'] = {'^([^%s%d]+) (%d%d?), +(%d%d%d%d)$', 'm', 'd', 'y'}, ['ymd'] = {'^(%d%d%d%d) +([^%s%d]+) (%d%d?)$', 'y', 'm', 'd'}, -- not mos compliant at en.wiki but may be acceptible at other wikis }; local t = {}; if non_western_digits then -- this wiki uses non-western digits? date_str = mw.ustring.gsub (date_str, '%d', digits); -- convert this wiki's non-western digits to western digits end if date_str:match ('^%d%d%d%d%-%d%d%-%d%d$') then -- already an iso format date, return western digits form return date_str, 'iso'; end for k, v in pairs (patterns) do local c1, c2, c3 = mw.ustring.match (date_str, patterns[k][1]); -- c1 .. c3 are captured but we don't know what they hold if c1 then -- set on match t = { -- translate unspecified captures to y, m, and d [patterns[k][2]] = c1, -- fill the table of captures with the captures [patterns[k][3]] = c2, -- take index names from src_pattern table and assign sequential captures [patterns[k][4]] = c3, }; if month_num[t.m] then -- when month not already a number t.m = month_num[t.m]; -- replace valid month name with a number else return nil, 'iso'; -- not a valid date form because month not valid end return mw.ustring.format ('%.4d-%.2d-%.2d', t.y, t.m, t.d), k; -- return date in iso format end end return nil, 'iso'; -- date could not be decoded; return nil and default iso date end --[[--------------------------< makeDate >----------------------- Given year, month, day numbers, (zero-padded or not) return a full date in df format where df may be one of: mdy, dmy, iso, ymd on entry, year, month, day are presumed to be correct for the date that they represent; all are required in this module, makeDate() is sometimes given an iso-format date in year: makeDate (2018-09-20, nil, nil, df) this works because table.concat() sees only one table member ]] local function makeDate (year, month, day, df) local format = { ['dmy'] = 'j F Y', ['mdy'] = 'F j, Y', ['ymd'] = 'Y F j', ['iso'] = 'Y-m-d', }; local date = table.concat ({year, month, day}, '-'); -- assemble year-initial numeric-format date (zero padding not required here) if non_western_digits then -- this wiki uses non-western digits? date = mw.ustring.gsub (date, '%d', digits); -- convert this wiki's non-western digits to western digits end return mw.getContentLanguage():formatDate (format[df], date); end --[[--------------------------< I S _ V A L I D _ D A T E >---------------------------------------------------- Returns true if date is after 31 December 1899 (why is 1900 the min year? shouldn't the internet's date-of-birth be min year?), not after today's date, and represents a valid date (29 February 2017 is not a valid date). Applies Gregorian leapyear rules. all arguments are required ]] local function is_valid_date (year, month, day) local days_in_month = {31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31}; local month_length; local y, m, d; local today = os.date ('*t'); -- fetch a table of current date parts if not year or == year or not month or == month or not day or == day then return false; -- something missing end y = tonumber (year); m = tonumber (month); d = tonumber (day); if 1900 > y or today.year < y or 1 > m or 12 < m then -- year and month are within bounds TODO: 1900? return false; end if (2==m) then -- if February month_length = 28; -- then 28 days unless if (0==(y%4) and (0~=(y%100) or 0==(y%400))) then -- is a leap year? month_length = 29; -- if leap year then 29 days in February end else month_length=days_in_month[m]; end if 1 > d or month_length < d then -- day is within bounds return false; end -- here when date parts represent a valid date return os.time({['year']=y, ['month']=m, ['day']=d, ['hour']=0}) <= os.time(); -- date at midnight must be less than or equal to current date/time end --[[--------------------------< decodeWebciteDate >----------------------- Given a URI-path to Webcite (eg. /67xHmVFWP) return the encoded date in df format returns date string in df format - webcite date is a unix timestamp encoded as bae62 or the string 'query' ]] local function decodeWebciteDate(path, df) local dt = {}; local decode; dt = mw.text.split(path, "/") -- valid URL formats that are not base62 -- http://www.webcitation.org/query?id=1138911916587475 -- http://www.webcitation.org/query?url=http..&date=2012-06-01+21:40:03 -- http://www.webcitation.org/1138911916587475 -- http://www.webcitation.org/cache/73e53dd1f16cf8c5da298418d2a6e452870cf50e -- http://www.webcitation.org/getfile.php?fileid=1c46e791d68e89e12d0c2532cc3cf629b8bc8c8e if dt[2]:find ('query', 1, true) or dt[2]:find ('cache', 1, true) or dt[2]:find ('getfile', 1, true) or tonumber(dt[2]) then return 'query'; end decode = base62(dt[2]); -- base62 string -> exponential number if not decode then return nil; -- nil return when dt[2] contains characters not in %w end dt = os.date('*t', string.format("%d", decode):sub(1,10)) -- exponential number -> text -> first 10 characters (a unix timestamp) -> a table of date parts decode = makeDate (dt.year, dt.month, dt.day, 'iso'); -- date comparisons are all done in iso format with western digits if non_western_digits then -- this wiki uses non-western digits? decode = mw.ustring.gsub (decode, '%d', digits); -- convert this wiki's non-western digits to western digits end return decode; end --[[--------------------------< decodeWaybackDate >----------------------- Given a URI-path to Wayback (eg. /web/20160901010101/http://example.com ) or Library of Congress Web Archives (eg. /all/20160901010101/http://example.com) or UK Government Web Archive (eg. /ukgwa/20160901010101/http://example.com or /tna/20160901010101/http://example.com) return the formatted date eg. "September 1, 2016" in df format Handle non-digits in snapshot ID such as "re_" and "-" and "*" returns two values: first value is one of these: valid date string in df format - wayback date is valid (including the text string 'index' when date is '/*/') empty string - wayback date is malformed (less than 8 digits, not a valid date) nil - wayback date is '/save/' or otherwise not a number second return value is an appropriate 'message' may or may not be formatted ]] local function decodeWaybackDate(path, df) local msg, snapdate; snapdate = path:gsub ('^/web/', ):gsub ('^/all/', ):gsub ('^/ukgwa/', ):gsub ('^/tna/', ):gsub ('^/', ); -- remove leading /web/, /all/, /ukgwa/, /tna/, or / snapdate = snapdate:match ('^[^/]+'); -- get timestamp if snapdate == "*" then -- eg. /web/*/http.., etc. return 'index'; -- return indicator that this url has an index date end snapdate = snapdate:gsub ('%a%a_%d?$', ):gsub ('%-', ); -- from date, remove any trailing "re_", dashes msg = ; if snapdate:match ('%*$') then -- a trailing '*' causes calendar display at archive .org snapdate = snapdate:gsub ('%*$', ); -- remove so not part of length calc later msg = inlineRed (err_warn_msgs.ts_cal, 'warning'); -- make a message end if not tonumber(snapdate) then return nil, 'ts_nan'; -- return nil (fatal error flag) and message selector end local dlen = snapdate:len(); if dlen < 8 then -- we need 8 digits TODO: but shouldn't this be testing for 14 digits? return , inlineRed (err_warn_msgs.ts_short, 'error'); -- return empty string and error message end local year, month, day = snapdate:match ('(%d%d%d%d)(%d%d)(%d%d)'); -- no need for snapdatelong here if not is_valid_date (year, month, day) then return , inlineRed (err_warn_msgs.ts_date, 'error'); -- return empty string and error message end snapdate = table.concat ({year, month, day}, '-'); -- date comparisons are all done in iso format if 14 == dlen then return snapdate, msg; -- return date with message if any else return snapdate, msg .. inlineRed (err_warn_msgs.ts_len, 'warning'); -- return date with warning message(s) end end --[[--------------------------< decodeArchiveisDate >----------------------- Given an Archive.is "long link" URI-path (e.g. /2016.08.28-144552/http://example.com) return the date in df format (e.g. if df = dmy, return 28 August 2016) Handles "." and "-" in snapshot date, so 2016.08.28-144552 is same as 20160828144552 returns two values: first value is one of these: valid date string in df format - archive.is date is valid (including the text string 'short link' when url is the short form) empty string - wayback date is malformed (not a number, less than 8 digits, not a valid date) nil - wayback date is '/save/' second return value is an appropriate 'message' may or may not be formatted ]] local function decodeArchiveisDate(path, df) local snapdate if path:match ('^/%w+$') then -- short form url path is '/' followed by some number of base 62 digits and nothing else return "short link" -- e.g. http://archive.is/hD1qz end snapdate = mw.text.split (path, '/')[2]:gsub('[%.%-]', ); -- get snapshot date, e.g. 2016.08.28-144552; remove periods and hyphens local dlen = string.len(snapdate) if dlen < 8 then -- we need 8 digits TODO: but shouldn't this be testing for 14 digits? return , inlineRed (err_warn_msgs.ts_short, 'error'); -- return empty string and error message end local year, month, day = snapdate:match ('(%d%d%d%d)(%d%d)(%d%d)'); -- no need for snapdatelong here if not is_valid_date (year, month, day) then return , inlineRed (err_warn_msgs.ts_date, 'error'); -- return empty string and error message end snapdate = table.concat ({year, month, day}, '-'); -- date comparisons are all done in iso format if 14 == dlen then return snapdate; -- return date else return snapdate, inlineRed (err_warn_msgs.ts_len, 'warning'); -- return date with warning message end end --[[--------------------------< serviceName >----------------------- Given a domain extracted by mw.uri.new() (eg. web.archive.org) set tail string and service ID ]] local function serviceName(host, no_link) local tracking; local index; host = host:lower():gsub ('^web%.(.+)', '%1'):gsub ('^www%.(.+)', '%1'); -- lowercase, remove web. and www. subdomains if services[host] then index = host; else for k, _ in pairs (services) do if host:find ('%f[%a]'..k:gsub ('([%.%-])', '%%%1')) then index = k; break; end end end if index then local out = {}; -- empty string in [1] so that concatenated result has leading single space ulx.url1.service = services[index][4] or 'other'; tracking = services[index][5] or categories.other; -- build tail string if false == services[index][1] then -- select prefix table.insert (out, prefixes.at); elseif true == services[index][1] then table.insert (out, prefixes.atthe); else table.insert (out, services[index][1]); end table.insert (out, make_wikilink (services[index][2], services[index][3], no_link)); -- add article wikilink if services[index][6] then -- add tail postfix if it exists table.insert (out, services[index][6]); end ulx.url1.tail = table.concat (out, ' '); -- put it all together; result has leading space character else -- here when unknown archive ulx.url1.service = 'other'; tracking = categories.unknown; ulx.url1.tail = table.concat ({, prefixes.at, host, inlineRed (err_warn_msgs.unknown_url, error)}, ' '); end track[tracking] = 1 end --[[--------------------------< parseExtraArgs >----------------------- Parse numbered arguments starting at 2, such as url2..url10, date2..date10, title2..title10 For example: Template loop detected: Template:Webarchive Three url arguments not in numeric sequence (1..4..7). Function only processes arguments numbered 2 or greater (in this case 4 and 7) It creates numeric sequenced table entries like: urlx.url2.url = <argument value for url4> urlx.url3.url = <argument value for url7> Returns the number of URL arguments found numbered 2 or greater (in this case returns "2") ]] local function parseExtraArgs(args) local i, j, argurl, argurl2, argdate, argtitle j = 2 for i = 2, config.maxurls do argurl = "url" .. i if args[argurl] then argurl2 = "url" .. j ulx[argurl2] = {} ulx[argurl2]["url"] = args[argurl] argdate = "date" .. i if args[argdate] then ulx[argurl2]["date"] = args[argdate] else ulx[argurl2]["date"] = inlineRed (err_warn_msgs.date_miss, 'warning'); end argtitle = "title" .. i if args[argtitle] then ulx[argurl2]["title"] = args[argtitle] else ulx[argurl2]["title"] = nil end j = j + 1 end end if j == 2 then return 0 else return j - 2 end end --[[--------------------------< comma >----------------------- Given a date string, return "," if it's MDY ]] local function comma(date) return (date and date:match ('%a+ +%d%d?(,) +%d%d%d%d')) or ; end --[[--------------------------< createRendering >----------------------- Return a rendering of the data in ulx[][] ]] local function createRendering() local displayfield local out = {}; local index_date, msg = ulx.url1.date:match ('(index)(.*)'); -- when ulx.url1.date extract 'index' text and message text (if there is a message) ulx.url1.date = ulx.url1.date:gsub ('index.*', 'index'); -- remove message if 'none' == ulx.url1.format then -- For Template:Wayback, Template:Webcite table.insert (out, '['); -- open extlink markup table.insert (out, ulx.url1.url); -- add url if ulx.url1.title then table.insert (out, ' ') -- the required space table.insert (out, ulx.url1.title) -- the title table.insert (out, ']'); -- close extlink markup table.insert (out, ulx.url1.tail); -- tail text if ulx.url1.date then table.insert (out, ' ('); -- open date text; TODO: why the html entity? replace with regular space? table.insert (out, 'index' == ulx.url1.date and s_text.archive or s_text.archived); -- add text table.insert (out, ' '); -- insert a space table.insert (out, ulx.url1.date); -- add date table.insert (out, ')'); -- close date text end else -- no title if index_date then -- when url date is 'index' table.insert (out, table.concat ({' ', s_text.Archive_index, ']'})); -- add the index link label table.insert (out, msg or ); -- add date mismatch message when url date is /*/ and |date= has valid date else table.insert (out, table.concat ({' ', s_text.Archived, '] '})); -- add link label for url has timestamp date (will include mismatch message if there is one) end if ulx.url1.date then if 'index' ~= ulx.url1.date then table.insert (out, ulx.url1.date); -- add date when data is not 'index' end table.insert (out, comma(ulx.url1.date)); -- add ',' if date format is mdy table.insert (out, ulx.url1.tail); -- add tail text else -- no date table.insert (out, ulx.url1.tail); -- add tail text end end if 0 < ulx.url1.extraurls then -- For multiple archive URLs local tot = ulx.url1.extraurls + 1 table.insert (out, '.') -- terminate first url table.insert (out, table.concat ({' ', s_text.addlarchives, ': '})); -- add header text for i=2, tot do -- loop through the additionals local index = table.concat ({'url', i}); -- make an index displayfield = ulx[index]['title'] and 'title' or 'date'; -- choose display text table.insert (out, '['); -- open extlink markup table.insert (out, ulx[index]['url']); -- add the url table.insert (out, ' '); -- the required space table.insert (out, ulx[index][displayfield]); -- add the label table.insert (out, ']'); -- close extlink markup table.insert (out, i==tot and '.' or ', '); -- add terminator end end return table.concat (out); -- make a big string and done else -- For Template:Cite archives if 'addlarchives' == ulx.url1.format then -- Multiple archive services table.insert (out, table.concat ({s_text.addlarchives, ': '})); -- add header text else -- Multiple pages from the same archive table.insert (out, table.concat ({s_text.addlpages, ' '})); -- add header text table.insert (out, ulx.url1.date); -- add date to header text table.insert (out, ': '); -- close header text end local tot = ulx.url1.extraurls + 1; for i=1, tot do -- loop through the additionals local index = table.concat ({'url', i}); -- make an index table.insert (out, '['); -- open extlink markup table.insert (out, ulx[index]['url']); -- add url table.insert (out, ' '); -- add required space displayfield = ulx[index]['title']; if 'addlarchives' == ulx.url1.format then if not displayfield then displayfield = ulx[index]['date'] end else -- must be addlpages if not displayfield then displayfield = table.concat ({s_text.Page, ' ', i}); end end table.insert (out, displayfield); -- add title, date, page label text table.insert (out, ']'); -- close extlink markup table.insert (out, (i==tot and '.' or ', ')); -- add terminator end return table.concat (out); -- make a big string and done end end --[[--------------------------< P A R A M E T E R _ N A M E _ X L A T E >-------------------------------------- for internaltionalization, translate local-language parameter names to their English equivalents TODO: return error message if multiple aliases of the same canonical parameter name are found? returns two tables: new_args - holds canonical form parameters and their values either from translation or because the parameter was already in canonical form origin - maps canonical-form parameter names to their untranslated (local language) form for error messaging in the local language unrecognized parameters are ignored ]] local function parameter_name_xlate (args, params, enum_params) local name; -- holds modifiable name of the parameter name during evaluation local enum; -- for enumerated parameters, holds the enumerator during evaluation local found = false; -- flag used to break out of nested for loops local new_args = {}; -- a table that holds canonical and translated parameter k/v pairs local origin = {}; -- a table that maps original (local language) parameter names to their canonical name for local language error messaging local unnamed_params; -- set true when unsupported positional parameters are detected for k, v in pairs (args) do -- loop through all of the arguments in the args table name = k; -- copy of original parameter name if 'string' == type (k) then if non_western_digits then -- true when non-western digits supported at this wiki name = mw.ustring.gsub (name, '%d', digits); -- convert this wiki's non-western digits to western digits end enum = name:match ('%d+$'); -- get parameter enumerator if it exists; nil else if not enum then -- no enumerator so looking for non-enumnerated parameters -- TODO: insert shortcut here? if params[name] then name holds the canonical parameter name; no need to search further for pname, aliases in pairs (params) do -- loop through each parameter the params table for _, alias in ipairs (aliases) do -- loop through each alias in the parameter's aliases table if name == alias then new_args[pname] = v; -- create a new entry in the new_args table origin [pname] = k; -- create an entry to make canonical parameter name to original local language parameter name found = true; -- flag so that we can break out of these nested for loops break; -- no need to search the rest of the aliases table for name so go on to the next k, v pair end end if found then -- true when we found an alias that matched name found = false; -- reset the flag break; -- go do next args k/v pair end end else -- enumerated parameters name = name:gsub ('%d$', '#'); -- replace enumeration digits with place holder for table search -- TODO: insert shortcut here? if num_params[name] then name holds the canonical parameter name; no need to search further for pname, aliases in pairs (enum_params) do -- loop through each parameter the num_params table for _, alias in ipairs (aliases) do -- loop through each alias in the parameter's aliases table if name == alias then pname = pname:gsub ('#$', enum); -- replace the '#' place holder with the actual enumerator new_args[pname] = v; -- create a new entry in the new_args table origin [pname] = k; -- create an entry to make canonical parameter name to original local language parameter name found = true; -- flag so that we can break out of these nested for loops break; -- no need to search the rest of the aliases table for name so go on to the next k, v pair end end if found then -- true when we found an alias that matched name found = false; -- reset the flag break; -- go do next args k/v pair end end end else unnamed_params = true; -- flag for unsupported positional parameters end end -- for k, v return new_args, origin, unnamed_params; end --[[--------------------------< W E B A R C H I V E >---------------------------------------------------------- template entry point ]] local function webarchive(frame) local args = getArgs (frame); local data = mw.loadData (table.concat ({ -- make a data module name; sandbox or live 'Module:Webarchive/data', frame:getTitle():find('sandbox', 1, true) and '/sandbox' or -- this instance is ./sandbox then append /sandbox })); categories = data.categories; -- fill in the forward declarations config = data.config; if data.digits.enable then digits = data.digits; -- for i18n; table of digits in the local wiki's language non_western_digits = true; -- use_non_western_digits end err_warn_msgs = data.err_warn_msgs; excepted_pages = data.excepted_pages; month_num = data.month_num; -- for i18n; table of month names in the local wiki's language prefixes = data.prefixes; services = data.services; s_text = data.s_text; uncategorized_namespaces = data.uncategorized_namespaces; uncategorized_subpages = data.uncategorized_subpages; local origin = {}; -- holds a map of English to local language parameter names used in the current template; not currently used local unnamed_params; -- boolean set to true when template call has unnamed parameters args, origin, unnamed_params = parameter_name_xlate (args, data.params, data.enum_params); -- translate parameter names in args to English local date, format, msg, udate, uri, url; local ldf = 'iso'; -- when there is no |date= parameter, render url dates in iso format if args.url and args.url1 then -- URL argument (first) return inlineError (data.crit_err_msgs.conflicting, {origin.url, origin.url1}); end url = args.url or args.url1; if not url then return inlineError (data.crit_err_msgs.empty); end -- these iabot bugs perportedly fixed; removing these causes lua script error --[[ -- at Template:Webarchive/testcases/Production; resolve that before deleting these tests if mw.ustring.find( url, "https://web.http", 1, true ) then -- track bug - TODO: IAbot bug; not known if the bug has been fixed; deferred track[categories.error] = 1; return inlineError (data.crit_err_msgs.iabot1); end if url == "https://web.archive.org/http:/" then -- track bug - TODO: IAbot bug; not known if the bug has been fixed; deferred track[categories.error] = 1; return inlineError (data.crit_err_msgs.iabot2); end ]] if not (url:lower():find ('^http') or url:find ('^//')) then return inlineError (data.crit_err_msgs.invalid_url ); end ulx.url1 = {} ulx.url1.url = url ulx.url1.extraurls = parseExtraArgs(args) local good = false; good, uri = pcall (mw.uri.new, ulx.url1.url); -- get a table of uri parts from this url; protected mode to prevent lua error when ulx.url1.url is malformed if not good or nil == uri.host then -- abandon when ulx.url1.url is malformed return inlineError (data.crit_err_msgs.invalid_url); end serviceName(uri.host, args.nolink) if args.date and args.date1 then -- Date argument return inlineError (data.crit_err_msgs.conflicting, {origin.date, origin.date1}); end date = args.date or args.date1; date = date and date:gsub (' +', ' '); -- replace multiple spaces with a single space if date and config.verifydates then if '*' == date then date = 'index'; ldf = 'iso'; -- set to default format elseif 'mdy' == date then date = nil; -- if date extracted from URL, ldf = 'mdy'; -- then |date=mdy overrides iso elseif 'dmy' == date then date = nil; -- if date extracted from URL, ldf = 'dmy'; -- then |date=dmy overrides iso elseif 'ymd' == date then date = nil; -- if date extracted from URL, ldf = 'ymd'; -- then |date=ymd overrides iso else date, ldf = decode_date (date); -- get an iso format date from date and get date's original format end end if 'wayback' == ulx.url1.service or 'locwebarchives' == ulx.url1.service or 'ukgwa' == ulx.url1.service then if date then if config.verifydates then if ldf then udate, msg = decodeWaybackDate (uri.path); -- get the url date in iso format and format of date in |date=; 'index' when wayback url date is * if not udate then -- this is the only 'fatal' error return return inlineError (data.crit_err_msgs[msg]); end if udate ~= date then -- date comparison using iso format dates date = udate; msg = table.concat ({ inlineRed (err_warn_msgs.mismatch, 'warning'), -- add warning message msg, -- add message if there is one }); end end end else -- no |date= udate, msg = decodeWaybackDate (uri.path); if not udate then -- this is the only 'fatal' error return return inlineError (data.crit_err_msgs[msg]); end if == udate then date = nil; -- unset else date = udate; end end elseif 'webcite' == ulx.url1.service then if date then if config.verifydates then if ldf then udate = decodeWebciteDate (uri.path); -- get the url date in iso format if 'query' ~= udate then -- skip if query if udate ~= date then -- date comparison using iso format dates date = udate; msg = table.concat ({ inlineRed (err_warn_msgs.mismatch, 'warning'), }); end end end end else date = decodeWebciteDate( uri.path, "iso" ) if date == "query" then date = nil; -- unset msg = inlineRed (err_warn_msgs.date_miss, 'warning'); elseif not date then -- invalid base62 string date = inlineRed (err_warn_msgs.date1, 'error'); end end elseif 'archiveis' == ulx.url1.service then if date then if config.verifydates then if ldf then udate, msg = decodeArchiveisDate (uri.path) -- get the url date in iso format if 'short link' ~= udate then -- skip if short link if udate ~= date then -- date comparison using iso format dates date = udate; msg = table.concat ({ inlineRed (err_warn_msgs.mismatch, 'warning'), -- add warning message msg, -- add message if there is one }); end end end end else -- no |date= udate, msg = decodeArchiveisDate( uri.path, "iso" ) if udate == "short link" then date = nil; -- unset msg = inlineRed (err_warn_msgs.date_miss, 'warning'); elseif == udate then date = nil; -- unset else date = udate; end end else -- some other service if not date then msg = inlineRed (err_warn_msgs.date_miss, 'warning'); end end if 'index' == date then ulx.url1.date = date .. (msg or ); -- create index + message (if there is one) elseif date then ulx.url1.date = makeDate (date, nil, nil, ldf) .. (msg or ); -- create a date in the wiki's local language + message (if there is one) else ulx.url1.date = msg; end format = args.format; -- Format argument if not format then format = "none" else for k, v in pairs (data.format_vals) do -- |format= accepts two specific values loop through a table of those values local found; -- declare a nil flag for _, p in ipairs (v) do -- loop through local language variants if format == p then -- when |format= value matches format = k; -- use name from table key found = true; -- declare found so that we can break out of outer for loop break; -- break out of inner for loop end end if found then break; end end if format == "addlpages" then if not ulx.url1.date then format = "none" end elseif format == "addlarchives" then format = "addlarchives" else format = "none" end end ulx.url1.format = format if args.title and args.title1 then -- Title argument return inlineError (data.crit_err_msgs.conflicting, {origin.title, origin.title1}); end ulx.url1.title = args.title or args.title1; local rend = createRendering() if not rend then return inlineError (data.crit_err_msgs.unknown); end return rend .. ((unnamed_params and inlineRed (err_warn_msgs.unnamed_params, 'warning')) or ) .. createTracking(); end --[[--------------------------< E X P O R T E D F U N C T I O N S >------------------------------------------ ]] return {webarchive = webarchive}; Movieline, published by AMC's Filmsite. Retrieved August 15, 2017.
  55. ↑ The 100 Must-See Films of the 20th Century. --[[ ---------------------------------- Lua module implementing the Template loop detected: Template:Webarchive template. A merger of the functionality of three templates: Template:Wayback, Template:Webcite and Template:Cite archives ]] --[[--------------------------< D E P E N D E N C I E S >------------------------------------------------------ ]] require('Module:No globals'); local getArgs = require ('Module:Arguments').getArgs; --[[--------------------------< F O R W A R D D E C L A R A T I O N S >-------------------------------------- ]] local categories = {}; -- category names local config = {}; -- global configuration settings local digits = {}; -- for i18n; table that translates local-wiki digits to western digits local err_warn_msgs = {}; -- error and warning messages local excepted_pages = {}; local month_num = {}; -- for i18n; table that translates local-wiki month names to western digits local prefixes = {}; -- service provider tail string prefixes local services = {}; -- archive service provider data from local s_text = {}; -- table of static text strings used to build final rendering local uncategorized_namespaces = {}; -- list of namespaces that we should not categorize local uncategorized_subpages = {}; -- list of subpages that should not be categorized --[[--------------------------< P A G E S C O P E I D E N T I F I E R S >---------------------------------- ]] local non_western_digits; -- boolean flag set true when data.digits.enable is true local this_page = mw.title.getCurrentTitle(); local track = {}; -- Associative array to hold tracking categories local ulx = {}; -- Associative array to hold template data --[[--------------------------< S U B S T I T U T E >---------------------------------------------------------- Populates numbered arguments in a message string using an argument table. ]] local function substitute (msg, args) return args and mw.message.newRawMessage (msg, args):plain() or msg; end --[[--------------------------< tableLength >----------------------- Given a 1-D table, return number of elements ]] local function tableLength(T) local count = 0 for _ in pairs(T) do count = count + 1 end return count end --[=[-------------------------< M A K E _ W I K I L I N K >---------------------------------------------------- Makes a wikilink; when both link and display text is provided, returns a wikilink in the form D; if only link is provided, returns a wikilink in the form L; if neither are provided or link is omitted, returns an empty string. ]=] local function make_wikilink (link, display, no_link) if nil == no_link then if link and ( ~= link) then if display and ( ~= display) then return table.concat ({'', display, ''}); else return table.concat ({'', link, ''}); end end return display or ; -- link not set so return the display text else -- no_link if display and ( ~= display) then -- if there is display text return display; -- return that else return link or ; -- return the target article name or empty string end end end --[[--------------------------< createTracking >----------------------- Return data in track[] ie. tracking categories ]] local function createTracking() if not excepted_pages[this_page.fullText] then -- namespace:title/fragment is allowed to be categorized (typically this module's / template's testcases page(s)) if uncategorized_namespaces[this_page.nsText] then return ; -- this page not to be categorized so return empty string end for _,v in ipairs (uncategorized_subpages) do -- cycle through page name patterns if this_page.text:match (v) then -- test page name against each pattern return ; -- this subpage type not to be categorized so return empty string end end end local out = {}; if tableLength(track) > 0 then for key, _ in pairs(track) do -- loop through table table.insert (out, make_wikilink (key)); -- and convert category names to links end end return table.concat (out); -- concat into one big string; empty string if table is empty end --[[--------------------------< inlineError >----------------------- Critical error. Render output completely in red. Add to tracking category. This function called as the last thing before abandoning this module ]] local function inlineError (msg, args) track[categories.error] = 1 return table.concat ({ 'Error in ', -- open the error message span config.tname, -- insert the local language template name ' template: ', substitute (msg, args), -- insert the formatted error message '.', -- close the span createTracking() -- add the category }) end --[[--------------------------< inlineRed >----------------------- Render a text fragment in red, such as a warning as part of the final output. Add tracking category. ]] local function inlineRed(msg, trackmsg) if trackmsg == "warning" then track[categories.warning] = 1; elseif trackmsg == "error" then track[categories.error] = 1; end return '' .. msg .. '' end --[[--------------------------< base62 >----------------------- Convert base-62 to base-10 Credit: https://de.wikipedia.org/wiki/Modul:Expr ]] local function base62( value ) local r = 1 -- default return value is input value is malformed if value:match ('%W') then -- value must only be in the set [0-9a-zA-Z] return; -- nil return when value contains extraneous characters end local n = #value -- number of characters in value local k = 1 local c r = 0 for i = n, 1, -1 do -- loop through all characters in value from ls digit to ms digit c = value:byte( i, i ) if c >= 48 and c <= 57 then -- character is digit 0-9 c = c - 48 elseif c >= 65 and c <= 90 then -- character is ascii a-z c = c - 55 else -- must be ascii A-Z c = c - 61 end r = r + c * k -- accumulate this base62 character's value k = k * 62 -- bump for next end -- for i return r end --[[--------------------------< D E C O D E _ D A T E >-------------------------------------------------------- Given a date string, return it in iso format along with an indicator of the date's format. Except that month names must be recognizable as legitimate month names with proper capitalization, and that the date string must match one of the recognized date formats, no error checking is done here; return nil else ]] local function decode_date (date_str) local patterns = { ['dmy'] = {'^(%d%d?) +([^%s%d]+) +(%d%d%d%d)$', 'd', 'm', 'y'}, -- %a does not recognize unicode combining characters used by some languages ['mdy'] = {'^([^%s%d]+) (%d%d?), +(%d%d%d%d)$', 'm', 'd', 'y'}, ['ymd'] = {'^(%d%d%d%d) +([^%s%d]+) (%d%d?)$', 'y', 'm', 'd'}, -- not mos compliant at en.wiki but may be acceptible at other wikis }; local t = {}; if non_western_digits then -- this wiki uses non-western digits? date_str = mw.ustring.gsub (date_str, '%d', digits); -- convert this wiki's non-western digits to western digits end if date_str:match ('^%d%d%d%d%-%d%d%-%d%d$') then -- already an iso format date, return western digits form return date_str, 'iso'; end for k, v in pairs (patterns) do local c1, c2, c3 = mw.ustring.match (date_str, patterns[k][1]); -- c1 .. c3 are captured but we don't know what they hold if c1 then -- set on match t = { -- translate unspecified captures to y, m, and d [patterns[k][2]] = c1, -- fill the table of captures with the captures [patterns[k][3]] = c2, -- take index names from src_pattern table and assign sequential captures [patterns[k][4]] = c3, }; if month_num[t.m] then -- when month not already a number t.m = month_num[t.m]; -- replace valid month name with a number else return nil, 'iso'; -- not a valid date form because month not valid end return mw.ustring.format ('%.4d-%.2d-%.2d', t.y, t.m, t.d), k; -- return date in iso format end end return nil, 'iso'; -- date could not be decoded; return nil and default iso date end --[[--------------------------< makeDate >----------------------- Given year, month, day numbers, (zero-padded or not) return a full date in df format where df may be one of: mdy, dmy, iso, ymd on entry, year, month, day are presumed to be correct for the date that they represent; all are required in this module, makeDate() is sometimes given an iso-format date in year: makeDate (2018-09-20, nil, nil, df) this works because table.concat() sees only one table member ]] local function makeDate (year, month, day, df) local format = { ['dmy'] = 'j F Y', ['mdy'] = 'F j, Y', ['ymd'] = 'Y F j', ['iso'] = 'Y-m-d', }; local date = table.concat ({year, month, day}, '-'); -- assemble year-initial numeric-format date (zero padding not required here) if non_western_digits then -- this wiki uses non-western digits? date = mw.ustring.gsub (date, '%d', digits); -- convert this wiki's non-western digits to western digits end return mw.getContentLanguage():formatDate (format[df], date); end --[[--------------------------< I S _ V A L I D _ D A T E >---------------------------------------------------- Returns true if date is after 31 December 1899 (why is 1900 the min year? shouldn't the internet's date-of-birth be min year?), not after today's date, and represents a valid date (29 February 2017 is not a valid date). Applies Gregorian leapyear rules. all arguments are required ]] local function is_valid_date (year, month, day) local days_in_month = {31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31}; local month_length; local y, m, d; local today = os.date ('*t'); -- fetch a table of current date parts if not year or == year or not month or == month or not day or == day then return false; -- something missing end y = tonumber (year); m = tonumber (month); d = tonumber (day); if 1900 > y or today.year < y or 1 > m or 12 < m then -- year and month are within bounds TODO: 1900? return false; end if (2==m) then -- if February month_length = 28; -- then 28 days unless if (0==(y%4) and (0~=(y%100) or 0==(y%400))) then -- is a leap year? month_length = 29; -- if leap year then 29 days in February end else month_length=days_in_month[m]; end if 1 > d or month_length < d then -- day is within bounds return false; end -- here when date parts represent a valid date return os.time({['year']=y, ['month']=m, ['day']=d, ['hour']=0}) <= os.time(); -- date at midnight must be less than or equal to current date/time end --[[--------------------------< decodeWebciteDate >----------------------- Given a URI-path to Webcite (eg. /67xHmVFWP) return the encoded date in df format returns date string in df format - webcite date is a unix timestamp encoded as bae62 or the string 'query' ]] local function decodeWebciteDate(path, df) local dt = {}; local decode; dt = mw.text.split(path, "/") -- valid URL formats that are not base62 -- http://www.webcitation.org/query?id=1138911916587475 -- http://www.webcitation.org/query?url=http..&date=2012-06-01+21:40:03 -- http://www.webcitation.org/1138911916587475 -- http://www.webcitation.org/cache/73e53dd1f16cf8c5da298418d2a6e452870cf50e -- http://www.webcitation.org/getfile.php?fileid=1c46e791d68e89e12d0c2532cc3cf629b8bc8c8e if dt[2]:find ('query', 1, true) or dt[2]:find ('cache', 1, true) or dt[2]:find ('getfile', 1, true) or tonumber(dt[2]) then return 'query'; end decode = base62(dt[2]); -- base62 string -> exponential number if not decode then return nil; -- nil return when dt[2] contains characters not in %w end dt = os.date('*t', string.format("%d", decode):sub(1,10)) -- exponential number -> text -> first 10 characters (a unix timestamp) -> a table of date parts decode = makeDate (dt.year, dt.month, dt.day, 'iso'); -- date comparisons are all done in iso format with western digits if non_western_digits then -- this wiki uses non-western digits? decode = mw.ustring.gsub (decode, '%d', digits); -- convert this wiki's non-western digits to western digits end return decode; end --[[--------------------------< decodeWaybackDate >----------------------- Given a URI-path to Wayback (eg. /web/20160901010101/http://example.com ) or Library of Congress Web Archives (eg. /all/20160901010101/http://example.com) or UK Government Web Archive (eg. /ukgwa/20160901010101/http://example.com or /tna/20160901010101/http://example.com) return the formatted date eg. "September 1, 2016" in df format Handle non-digits in snapshot ID such as "re_" and "-" and "*" returns two values: first value is one of these: valid date string in df format - wayback date is valid (including the text string 'index' when date is '/*/') empty string - wayback date is malformed (less than 8 digits, not a valid date) nil - wayback date is '/save/' or otherwise not a number second return value is an appropriate 'message' may or may not be formatted ]] local function decodeWaybackDate(path, df) local msg, snapdate; snapdate = path:gsub ('^/web/', ):gsub ('^/all/', ):gsub ('^/ukgwa/', ):gsub ('^/tna/', ):gsub ('^/', ); -- remove leading /web/, /all/, /ukgwa/, /tna/, or / snapdate = snapdate:match ('^[^/]+'); -- get timestamp if snapdate == "*" then -- eg. /web/*/http.., etc. return 'index'; -- return indicator that this url has an index date end snapdate = snapdate:gsub ('%a%a_%d?$', ):gsub ('%-', ); -- from date, remove any trailing "re_", dashes msg = ; if snapdate:match ('%*$') then -- a trailing '*' causes calendar display at archive .org snapdate = snapdate:gsub ('%*$', ); -- remove so not part of length calc later msg = inlineRed (err_warn_msgs.ts_cal, 'warning'); -- make a message end if not tonumber(snapdate) then return nil, 'ts_nan'; -- return nil (fatal error flag) and message selector end local dlen = snapdate:len(); if dlen < 8 then -- we need 8 digits TODO: but shouldn't this be testing for 14 digits? return , inlineRed (err_warn_msgs.ts_short, 'error'); -- return empty string and error message end local year, month, day = snapdate:match ('(%d%d%d%d)(%d%d)(%d%d)'); -- no need for snapdatelong here if not is_valid_date (year, month, day) then return , inlineRed (err_warn_msgs.ts_date, 'error'); -- return empty string and error message end snapdate = table.concat ({year, month, day}, '-'); -- date comparisons are all done in iso format if 14 == dlen then return snapdate, msg; -- return date with message if any else return snapdate, msg .. inlineRed (err_warn_msgs.ts_len, 'warning'); -- return date with warning message(s) end end --[[--------------------------< decodeArchiveisDate >----------------------- Given an Archive.is "long link" URI-path (e.g. /2016.08.28-144552/http://example.com) return the date in df format (e.g. if df = dmy, return 28 August 2016) Handles "." and "-" in snapshot date, so 2016.08.28-144552 is same as 20160828144552 returns two values: first value is one of these: valid date string in df format - archive.is date is valid (including the text string 'short link' when url is the short form) empty string - wayback date is malformed (not a number, less than 8 digits, not a valid date) nil - wayback date is '/save/' second return value is an appropriate 'message' may or may not be formatted ]] local function decodeArchiveisDate(path, df) local snapdate if path:match ('^/%w+$') then -- short form url path is '/' followed by some number of base 62 digits and nothing else return "short link" -- e.g. http://archive.is/hD1qz end snapdate = mw.text.split (path, '/')[2]:gsub('[%.%-]', ); -- get snapshot date, e.g. 2016.08.28-144552; remove periods and hyphens local dlen = string.len(snapdate) if dlen < 8 then -- we need 8 digits TODO: but shouldn't this be testing for 14 digits? return , inlineRed (err_warn_msgs.ts_short, 'error'); -- return empty string and error message end local year, month, day = snapdate:match ('(%d%d%d%d)(%d%d)(%d%d)'); -- no need for snapdatelong here if not is_valid_date (year, month, day) then return , inlineRed (err_warn_msgs.ts_date, 'error'); -- return empty string and error message end snapdate = table.concat ({year, month, day}, '-'); -- date comparisons are all done in iso format if 14 == dlen then return snapdate; -- return date else return snapdate, inlineRed (err_warn_msgs.ts_len, 'warning'); -- return date with warning message end end --[[--------------------------< serviceName >----------------------- Given a domain extracted by mw.uri.new() (eg. web.archive.org) set tail string and service ID ]] local function serviceName(host, no_link) local tracking; local index; host = host:lower():gsub ('^web%.(.+)', '%1'):gsub ('^www%.(.+)', '%1'); -- lowercase, remove web. and www. subdomains if services[host] then index = host; else for k, _ in pairs (services) do if host:find ('%f[%a]'..k:gsub ('([%.%-])', '%%%1')) then index = k; break; end end end if index then local out = {}; -- empty string in [1] so that concatenated result has leading single space ulx.url1.service = services[index][4] or 'other'; tracking = services[index][5] or categories.other; -- build tail string if false == services[index][1] then -- select prefix table.insert (out, prefixes.at); elseif true == services[index][1] then table.insert (out, prefixes.atthe); else table.insert (out, services[index][1]); end table.insert (out, make_wikilink (services[index][2], services[index][3], no_link)); -- add article wikilink if services[index][6] then -- add tail postfix if it exists table.insert (out, services[index][6]); end ulx.url1.tail = table.concat (out, ' '); -- put it all together; result has leading space character else -- here when unknown archive ulx.url1.service = 'other'; tracking = categories.unknown; ulx.url1.tail = table.concat ({, prefixes.at, host, inlineRed (err_warn_msgs.unknown_url, error)}, ' '); end track[tracking] = 1 end --[[--------------------------< parseExtraArgs >----------------------- Parse numbered arguments starting at 2, such as url2..url10, date2..date10, title2..title10 For example: Template loop detected: Template:Webarchive Three url arguments not in numeric sequence (1..4..7). Function only processes arguments numbered 2 or greater (in this case 4 and 7) It creates numeric sequenced table entries like: urlx.url2.url = <argument value for url4> urlx.url3.url = <argument value for url7> Returns the number of URL arguments found numbered 2 or greater (in this case returns "2") ]] local function parseExtraArgs(args) local i, j, argurl, argurl2, argdate, argtitle j = 2 for i = 2, config.maxurls do argurl = "url" .. i if args[argurl] then argurl2 = "url" .. j ulx[argurl2] = {} ulx[argurl2]["url"] = args[argurl] argdate = "date" .. i if args[argdate] then ulx[argurl2]["date"] = args[argdate] else ulx[argurl2]["date"] = inlineRed (err_warn_msgs.date_miss, 'warning'); end argtitle = "title" .. i if args[argtitle] then ulx[argurl2]["title"] = args[argtitle] else ulx[argurl2]["title"] = nil end j = j + 1 end end if j == 2 then return 0 else return j - 2 end end --[[--------------------------< comma >----------------------- Given a date string, return "," if it's MDY ]] local function comma(date) return (date and date:match ('%a+ +%d%d?(,) +%d%d%d%d')) or ; end --[[--------------------------< createRendering >----------------------- Return a rendering of the data in ulx[][] ]] local function createRendering() local displayfield local out = {}; local index_date, msg = ulx.url1.date:match ('(index)(.*)'); -- when ulx.url1.date extract 'index' text and message text (if there is a message) ulx.url1.date = ulx.url1.date:gsub ('index.*', 'index'); -- remove message if 'none' == ulx.url1.format then -- For Template:Wayback, Template:Webcite table.insert (out, '['); -- open extlink markup table.insert (out, ulx.url1.url); -- add url if ulx.url1.title then table.insert (out, ' ') -- the required space table.insert (out, ulx.url1.title) -- the title table.insert (out, ']'); -- close extlink markup table.insert (out, ulx.url1.tail); -- tail text if ulx.url1.date then table.insert (out, ' ('); -- open date text; TODO: why the html entity? replace with regular space? table.insert (out, 'index' == ulx.url1.date and s_text.archive or s_text.archived); -- add text table.insert (out, ' '); -- insert a space table.insert (out, ulx.url1.date); -- add date table.insert (out, ')'); -- close date text end else -- no title if index_date then -- when url date is 'index' table.insert (out, table.concat ({' ', s_text.Archive_index, ']'})); -- add the index link label table.insert (out, msg or ); -- add date mismatch message when url date is /*/ and |date= has valid date else table.insert (out, table.concat ({' ', s_text.Archived, '] '})); -- add link label for url has timestamp date (will include mismatch message if there is one) end if ulx.url1.date then if 'index' ~= ulx.url1.date then table.insert (out, ulx.url1.date); -- add date when data is not 'index' end table.insert (out, comma(ulx.url1.date)); -- add ',' if date format is mdy table.insert (out, ulx.url1.tail); -- add tail text else -- no date table.insert (out, ulx.url1.tail); -- add tail text end end if 0 < ulx.url1.extraurls then -- For multiple archive URLs local tot = ulx.url1.extraurls + 1 table.insert (out, '.') -- terminate first url table.insert (out, table.concat ({' ', s_text.addlarchives, ': '})); -- add header text for i=2, tot do -- loop through the additionals local index = table.concat ({'url', i}); -- make an index displayfield = ulx[index]['title'] and 'title' or 'date'; -- choose display text table.insert (out, '['); -- open extlink markup table.insert (out, ulx[index]['url']); -- add the url table.insert (out, ' '); -- the required space table.insert (out, ulx[index][displayfield]); -- add the label table.insert (out, ']'); -- close extlink markup table.insert (out, i==tot and '.' or ', '); -- add terminator end end return table.concat (out); -- make a big string and done else -- For Template:Cite archives if 'addlarchives' == ulx.url1.format then -- Multiple archive services table.insert (out, table.concat ({s_text.addlarchives, ': '})); -- add header text else -- Multiple pages from the same archive table.insert (out, table.concat ({s_text.addlpages, ' '})); -- add header text table.insert (out, ulx.url1.date); -- add date to header text table.insert (out, ': '); -- close header text end local tot = ulx.url1.extraurls + 1; for i=1, tot do -- loop through the additionals local index = table.concat ({'url', i}); -- make an index table.insert (out, '['); -- open extlink markup table.insert (out, ulx[index]['url']); -- add url table.insert (out, ' '); -- add required space displayfield = ulx[index]['title']; if 'addlarchives' == ulx.url1.format then if not displayfield then displayfield = ulx[index]['date'] end else -- must be addlpages if not displayfield then displayfield = table.concat ({s_text.Page, ' ', i}); end end table.insert (out, displayfield); -- add title, date, page label text table.insert (out, ']'); -- close extlink markup table.insert (out, (i==tot and '.' or ', ')); -- add terminator end return table.concat (out); -- make a big string and done end end --[[--------------------------< P A R A M E T E R _ N A M E _ X L A T E >-------------------------------------- for internaltionalization, translate local-language parameter names to their English equivalents TODO: return error message if multiple aliases of the same canonical parameter name are found? returns two tables: new_args - holds canonical form parameters and their values either from translation or because the parameter was already in canonical form origin - maps canonical-form parameter names to their untranslated (local language) form for error messaging in the local language unrecognized parameters are ignored ]] local function parameter_name_xlate (args, params, enum_params) local name; -- holds modifiable name of the parameter name during evaluation local enum; -- for enumerated parameters, holds the enumerator during evaluation local found = false; -- flag used to break out of nested for loops local new_args = {}; -- a table that holds canonical and translated parameter k/v pairs local origin = {}; -- a table that maps original (local language) parameter names to their canonical name for local language error messaging local unnamed_params; -- set true when unsupported positional parameters are detected for k, v in pairs (args) do -- loop through all of the arguments in the args table name = k; -- copy of original parameter name if 'string' == type (k) then if non_western_digits then -- true when non-western digits supported at this wiki name = mw.ustring.gsub (name, '%d', digits); -- convert this wiki's non-western digits to western digits end enum = name:match ('%d+$'); -- get parameter enumerator if it exists; nil else if not enum then -- no enumerator so looking for non-enumnerated parameters -- TODO: insert shortcut here? if params[name] then name holds the canonical parameter name; no need to search further for pname, aliases in pairs (params) do -- loop through each parameter the params table for _, alias in ipairs (aliases) do -- loop through each alias in the parameter's aliases table if name == alias then new_args[pname] = v; -- create a new entry in the new_args table origin [pname] = k; -- create an entry to make canonical parameter name to original local language parameter name found = true; -- flag so that we can break out of these nested for loops break; -- no need to search the rest of the aliases table for name so go on to the next k, v pair end end if found then -- true when we found an alias that matched name found = false; -- reset the flag break; -- go do next args k/v pair end end else -- enumerated parameters name = name:gsub ('%d$', '#'); -- replace enumeration digits with place holder for table search -- TODO: insert shortcut here? if num_params[name] then name holds the canonical parameter name; no need to search further for pname, aliases in pairs (enum_params) do -- loop through each parameter the num_params table for _, alias in ipairs (aliases) do -- loop through each alias in the parameter's aliases table if name == alias then pname = pname:gsub ('#$', enum); -- replace the '#' place holder with the actual enumerator new_args[pname] = v; -- create a new entry in the new_args table origin [pname] = k; -- create an entry to make canonical parameter name to original local language parameter name found = true; -- flag so that we can break out of these nested for loops break; -- no need to search the rest of the aliases table for name so go on to the next k, v pair end end if found then -- true when we found an alias that matched name found = false; -- reset the flag break; -- go do next args k/v pair end end end else unnamed_params = true; -- flag for unsupported positional parameters end end -- for k, v return new_args, origin, unnamed_params; end --[[--------------------------< W E B A R C H I V E >---------------------------------------------------------- template entry point ]] local function webarchive(frame) local args = getArgs (frame); local data = mw.loadData (table.concat ({ -- make a data module name; sandbox or live 'Module:Webarchive/data', frame:getTitle():find('sandbox', 1, true) and '/sandbox' or -- this instance is ./sandbox then append /sandbox })); categories = data.categories; -- fill in the forward declarations config = data.config; if data.digits.enable then digits = data.digits; -- for i18n; table of digits in the local wiki's language non_western_digits = true; -- use_non_western_digits end err_warn_msgs = data.err_warn_msgs; excepted_pages = data.excepted_pages; month_num = data.month_num; -- for i18n; table of month names in the local wiki's language prefixes = data.prefixes; services = data.services; s_text = data.s_text; uncategorized_namespaces = data.uncategorized_namespaces; uncategorized_subpages = data.uncategorized_subpages; local origin = {}; -- holds a map of English to local language parameter names used in the current template; not currently used local unnamed_params; -- boolean set to true when template call has unnamed parameters args, origin, unnamed_params = parameter_name_xlate (args, data.params, data.enum_params); -- translate parameter names in args to English local date, format, msg, udate, uri, url; local ldf = 'iso'; -- when there is no |date= parameter, render url dates in iso format if args.url and args.url1 then -- URL argument (first) return inlineError (data.crit_err_msgs.conflicting, {origin.url, origin.url1}); end url = args.url or args.url1; if not url then return inlineError (data.crit_err_msgs.empty); end -- these iabot bugs perportedly fixed; removing these causes lua script error --[[ -- at Template:Webarchive/testcases/Production; resolve that before deleting these tests if mw.ustring.find( url, "https://web.http", 1, true ) then -- track bug - TODO: IAbot bug; not known if the bug has been fixed; deferred track[categories.error] = 1; return inlineError (data.crit_err_msgs.iabot1); end if url == "https://web.archive.org/http:/" then -- track bug - TODO: IAbot bug; not known if the bug has been fixed; deferred track[categories.error] = 1; return inlineError (data.crit_err_msgs.iabot2); end ]] if not (url:lower():find ('^http') or url:find ('^//')) then return inlineError (data.crit_err_msgs.invalid_url ); end ulx.url1 = {} ulx.url1.url = url ulx.url1.extraurls = parseExtraArgs(args) local good = false; good, uri = pcall (mw.uri.new, ulx.url1.url); -- get a table of uri parts from this url; protected mode to prevent lua error when ulx.url1.url is malformed if not good or nil == uri.host then -- abandon when ulx.url1.url is malformed return inlineError (data.crit_err_msgs.invalid_url); end serviceName(uri.host, args.nolink) if args.date and args.date1 then -- Date argument return inlineError (data.crit_err_msgs.conflicting, {origin.date, origin.date1}); end date = args.date or args.date1; date = date and date:gsub (' +', ' '); -- replace multiple spaces with a single space if date and config.verifydates then if '*' == date then date = 'index'; ldf = 'iso'; -- set to default format elseif 'mdy' == date then date = nil; -- if date extracted from URL, ldf = 'mdy'; -- then |date=mdy overrides iso elseif 'dmy' == date then date = nil; -- if date extracted from URL, ldf = 'dmy'; -- then |date=dmy overrides iso elseif 'ymd' == date then date = nil; -- if date extracted from URL, ldf = 'ymd'; -- then |date=ymd overrides iso else date, ldf = decode_date (date); -- get an iso format date from date and get date's original format end end if 'wayback' == ulx.url1.service or 'locwebarchives' == ulx.url1.service or 'ukgwa' == ulx.url1.service then if date then if config.verifydates then if ldf then udate, msg = decodeWaybackDate (uri.path); -- get the url date in iso format and format of date in |date=; 'index' when wayback url date is * if not udate then -- this is the only 'fatal' error return return inlineError (data.crit_err_msgs[msg]); end if udate ~= date then -- date comparison using iso format dates date = udate; msg = table.concat ({ inlineRed (err_warn_msgs.mismatch, 'warning'), -- add warning message msg, -- add message if there is one }); end end end else -- no |date= udate, msg = decodeWaybackDate (uri.path); if not udate then -- this is the only 'fatal' error return return inlineError (data.crit_err_msgs[msg]); end if == udate then date = nil; -- unset else date = udate; end end elseif 'webcite' == ulx.url1.service then if date then if config.verifydates then if ldf then udate = decodeWebciteDate (uri.path); -- get the url date in iso format if 'query' ~= udate then -- skip if query if udate ~= date then -- date comparison using iso format dates date = udate; msg = table.concat ({ inlineRed (err_warn_msgs.mismatch, 'warning'), }); end end end end else date = decodeWebciteDate( uri.path, "iso" ) if date == "query" then date = nil; -- unset msg = inlineRed (err_warn_msgs.date_miss, 'warning'); elseif not date then -- invalid base62 string date = inlineRed (err_warn_msgs.date1, 'error'); end end elseif 'archiveis' == ulx.url1.service then if date then if config.verifydates then if ldf then udate, msg = decodeArchiveisDate (uri.path) -- get the url date in iso format if 'short link' ~= udate then -- skip if short link if udate ~= date then -- date comparison using iso format dates date = udate; msg = table.concat ({ inlineRed (err_warn_msgs.mismatch, 'warning'), -- add warning message msg, -- add message if there is one }); end end end end else -- no |date= udate, msg = decodeArchiveisDate( uri.path, "iso" ) if udate == "short link" then date = nil; -- unset msg = inlineRed (err_warn_msgs.date_miss, 'warning'); elseif == udate then date = nil; -- unset else date = udate; end end else -- some other service if not date then msg = inlineRed (err_warn_msgs.date_miss, 'warning'); end end if 'index' == date then ulx.url1.date = date .. (msg or ); -- create index + message (if there is one) elseif date then ulx.url1.date = makeDate (date, nil, nil, ldf) .. (msg or ); -- create a date in the wiki's local language + message (if there is one) else ulx.url1.date = msg; end format = args.format; -- Format argument if not format then format = "none" else for k, v in pairs (data.format_vals) do -- |format= accepts two specific values loop through a table of those values local found; -- declare a nil flag for _, p in ipairs (v) do -- loop through local language variants if format == p then -- when |format= value matches format = k; -- use name from table key found = true; -- declare found so that we can break out of outer for loop break; -- break out of inner for loop end end if found then break; end end if format == "addlpages" then if not ulx.url1.date then format = "none" end elseif format == "addlarchives" then format = "addlarchives" else format = "none" end end ulx.url1.format = format if args.title and args.title1 then -- Title argument return inlineError (data.crit_err_msgs.conflicting, {origin.title, origin.title1}); end ulx.url1.title = args.title or args.title1; local rend = createRendering() if not rend then return inlineError (data.crit_err_msgs.unknown); end return rend .. ((unnamed_params and inlineRed (err_warn_msgs.unnamed_params, 'warning')) or ) .. createTracking(); end --[[--------------------------< E X P O R T E D F U N C T I O N S >------------------------------------------ ]] return {webarchive = webarchive}; Leonard Maltin, published by AMC's Filmsite. Retrieved August 15, 2017.
  56. ↑ The Top 100 Films of All Time. --[[ ---------------------------------- Lua module implementing the Template loop detected: Template:Webarchive template. A merger of the functionality of three templates: Template:Wayback, Template:Webcite and Template:Cite archives ]] --[[--------------------------< D E P E N D E N C I E S >------------------------------------------------------ ]] require('Module:No globals'); local getArgs = require ('Module:Arguments').getArgs; --[[--------------------------< F O R W A R D D E C L A R A T I O N S >-------------------------------------- ]] local categories = {}; -- category names local config = {}; -- global configuration settings local digits = {}; -- for i18n; table that translates local-wiki digits to western digits local err_warn_msgs = {}; -- error and warning messages local excepted_pages = {}; local month_num = {}; -- for i18n; table that translates local-wiki month names to western digits local prefixes = {}; -- service provider tail string prefixes local services = {}; -- archive service provider data from local s_text = {}; -- table of static text strings used to build final rendering local uncategorized_namespaces = {}; -- list of namespaces that we should not categorize local uncategorized_subpages = {}; -- list of subpages that should not be categorized --[[--------------------------< P A G E S C O P E I D E N T I F I E R S >---------------------------------- ]] local non_western_digits; -- boolean flag set true when data.digits.enable is true local this_page = mw.title.getCurrentTitle(); local track = {}; -- Associative array to hold tracking categories local ulx = {}; -- Associative array to hold template data --[[--------------------------< S U B S T I T U T E >---------------------------------------------------------- Populates numbered arguments in a message string using an argument table. ]] local function substitute (msg, args) return args and mw.message.newRawMessage (msg, args):plain() or msg; end --[[--------------------------< tableLength >----------------------- Given a 1-D table, return number of elements ]] local function tableLength(T) local count = 0 for _ in pairs(T) do count = count + 1 end return count end --[=[-------------------------< M A K E _ W I K I L I N K >---------------------------------------------------- Makes a wikilink; when both link and display text is provided, returns a wikilink in the form D; if only link is provided, returns a wikilink in the form L; if neither are provided or link is omitted, returns an empty string. ]=] local function make_wikilink (link, display, no_link) if nil == no_link then if link and ( ~= link) then if display and ( ~= display) then return table.concat ({'', display, ''}); else return table.concat ({'', link, ''}); end end return display or ; -- link not set so return the display text else -- no_link if display and ( ~= display) then -- if there is display text return display; -- return that else return link or ; -- return the target article name or empty string end end end --[[--------------------------< createTracking >----------------------- Return data in track[] ie. tracking categories ]] local function createTracking() if not excepted_pages[this_page.fullText] then -- namespace:title/fragment is allowed to be categorized (typically this module's / template's testcases page(s)) if uncategorized_namespaces[this_page.nsText] then return ; -- this page not to be categorized so return empty string end for _,v in ipairs (uncategorized_subpages) do -- cycle through page name patterns if this_page.text:match (v) then -- test page name against each pattern return ; -- this subpage type not to be categorized so return empty string end end end local out = {}; if tableLength(track) > 0 then for key, _ in pairs(track) do -- loop through table table.insert (out, make_wikilink (key)); -- and convert category names to links end end return table.concat (out); -- concat into one big string; empty string if table is empty end --[[--------------------------< inlineError >----------------------- Critical error. Render output completely in red. Add to tracking category. This function called as the last thing before abandoning this module ]] local function inlineError (msg, args) track[categories.error] = 1 return table.concat ({ 'Error in ', -- open the error message span config.tname, -- insert the local language template name ' template: ', substitute (msg, args), -- insert the formatted error message '.', -- close the span createTracking() -- add the category }) end --[[--------------------------< inlineRed >----------------------- Render a text fragment in red, such as a warning as part of the final output. Add tracking category. ]] local function inlineRed(msg, trackmsg) if trackmsg == "warning" then track[categories.warning] = 1; elseif trackmsg == "error" then track[categories.error] = 1; end return '' .. msg .. '' end --[[--------------------------< base62 >----------------------- Convert base-62 to base-10 Credit: https://de.wikipedia.org/wiki/Modul:Expr ]] local function base62( value ) local r = 1 -- default return value is input value is malformed if value:match ('%W') then -- value must only be in the set [0-9a-zA-Z] return; -- nil return when value contains extraneous characters end local n = #value -- number of characters in value local k = 1 local c r = 0 for i = n, 1, -1 do -- loop through all characters in value from ls digit to ms digit c = value:byte( i, i ) if c >= 48 and c <= 57 then -- character is digit 0-9 c = c - 48 elseif c >= 65 and c <= 90 then -- character is ascii a-z c = c - 55 else -- must be ascii A-Z c = c - 61 end r = r + c * k -- accumulate this base62 character's value k = k * 62 -- bump for next end -- for i return r end --[[--------------------------< D E C O D E _ D A T E >-------------------------------------------------------- Given a date string, return it in iso format along with an indicator of the date's format. Except that month names must be recognizable as legitimate month names with proper capitalization, and that the date string must match one of the recognized date formats, no error checking is done here; return nil else ]] local function decode_date (date_str) local patterns = { ['dmy'] = {'^(%d%d?) +([^%s%d]+) +(%d%d%d%d)$', 'd', 'm', 'y'}, -- %a does not recognize unicode combining characters used by some languages ['mdy'] = {'^([^%s%d]+) (%d%d?), +(%d%d%d%d)$', 'm', 'd', 'y'}, ['ymd'] = {'^(%d%d%d%d) +([^%s%d]+) (%d%d?)$', 'y', 'm', 'd'}, -- not mos compliant at en.wiki but may be acceptible at other wikis }; local t = {}; if non_western_digits then -- this wiki uses non-western digits? date_str = mw.ustring.gsub (date_str, '%d', digits); -- convert this wiki's non-western digits to western digits end if date_str:match ('^%d%d%d%d%-%d%d%-%d%d$') then -- already an iso format date, return western digits form return date_str, 'iso'; end for k, v in pairs (patterns) do local c1, c2, c3 = mw.ustring.match (date_str, patterns[k][1]); -- c1 .. c3 are captured but we don't know what they hold if c1 then -- set on match t = { -- translate unspecified captures to y, m, and d [patterns[k][2]] = c1, -- fill the table of captures with the captures [patterns[k][3]] = c2, -- take index names from src_pattern table and assign sequential captures [patterns[k][4]] = c3, }; if month_num[t.m] then -- when month not already a number t.m = month_num[t.m]; -- replace valid month name with a number else return nil, 'iso'; -- not a valid date form because month not valid end return mw.ustring.format ('%.4d-%.2d-%.2d', t.y, t.m, t.d), k; -- return date in iso format end end return nil, 'iso'; -- date could not be decoded; return nil and default iso date end --[[--------------------------< makeDate >----------------------- Given year, month, day numbers, (zero-padded or not) return a full date in df format where df may be one of: mdy, dmy, iso, ymd on entry, year, month, day are presumed to be correct for the date that they represent; all are required in this module, makeDate() is sometimes given an iso-format date in year: makeDate (2018-09-20, nil, nil, df) this works because table.concat() sees only one table member ]] local function makeDate (year, month, day, df) local format = { ['dmy'] = 'j F Y', ['mdy'] = 'F j, Y', ['ymd'] = 'Y F j', ['iso'] = 'Y-m-d', }; local date = table.concat ({year, month, day}, '-'); -- assemble year-initial numeric-format date (zero padding not required here) if non_western_digits then -- this wiki uses non-western digits? date = mw.ustring.gsub (date, '%d', digits); -- convert this wiki's non-western digits to western digits end return mw.getContentLanguage():formatDate (format[df], date); end --[[--------------------------< I S _ V A L I D _ D A T E >---------------------------------------------------- Returns true if date is after 31 December 1899 (why is 1900 the min year? shouldn't the internet's date-of-birth be min year?), not after today's date, and represents a valid date (29 February 2017 is not a valid date). Applies Gregorian leapyear rules. all arguments are required ]] local function is_valid_date (year, month, day) local days_in_month = {31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31}; local month_length; local y, m, d; local today = os.date ('*t'); -- fetch a table of current date parts if not year or == year or not month or == month or not day or == day then return false; -- something missing end y = tonumber (year); m = tonumber (month); d = tonumber (day); if 1900 > y or today.year < y or 1 > m or 12 < m then -- year and month are within bounds TODO: 1900? return false; end if (2==m) then -- if February month_length = 28; -- then 28 days unless if (0==(y%4) and (0~=(y%100) or 0==(y%400))) then -- is a leap year? month_length = 29; -- if leap year then 29 days in February end else month_length=days_in_month[m]; end if 1 > d or month_length < d then -- day is within bounds return false; end -- here when date parts represent a valid date return os.time({['year']=y, ['month']=m, ['day']=d, ['hour']=0}) <= os.time(); -- date at midnight must be less than or equal to current date/time end --[[--------------------------< decodeWebciteDate >----------------------- Given a URI-path to Webcite (eg. /67xHmVFWP) return the encoded date in df format returns date string in df format - webcite date is a unix timestamp encoded as bae62 or the string 'query' ]] local function decodeWebciteDate(path, df) local dt = {}; local decode; dt = mw.text.split(path, "/") -- valid URL formats that are not base62 -- http://www.webcitation.org/query?id=1138911916587475 -- http://www.webcitation.org/query?url=http..&date=2012-06-01+21:40:03 -- http://www.webcitation.org/1138911916587475 -- http://www.webcitation.org/cache/73e53dd1f16cf8c5da298418d2a6e452870cf50e -- http://www.webcitation.org/getfile.php?fileid=1c46e791d68e89e12d0c2532cc3cf629b8bc8c8e if dt[2]:find ('query', 1, true) or dt[2]:find ('cache', 1, true) or dt[2]:find ('getfile', 1, true) or tonumber(dt[2]) then return 'query'; end decode = base62(dt[2]); -- base62 string -> exponential number if not decode then return nil; -- nil return when dt[2] contains characters not in %w end dt = os.date('*t', string.format("%d", decode):sub(1,10)) -- exponential number -> text -> first 10 characters (a unix timestamp) -> a table of date parts decode = makeDate (dt.year, dt.month, dt.day, 'iso'); -- date comparisons are all done in iso format with western digits if non_western_digits then -- this wiki uses non-western digits? decode = mw.ustring.gsub (decode, '%d', digits); -- convert this wiki's non-western digits to western digits end return decode; end --[[--------------------------< decodeWaybackDate >----------------------- Given a URI-path to Wayback (eg. /web/20160901010101/http://example.com ) or Library of Congress Web Archives (eg. /all/20160901010101/http://example.com) or UK Government Web Archive (eg. /ukgwa/20160901010101/http://example.com or /tna/20160901010101/http://example.com) return the formatted date eg. "September 1, 2016" in df format Handle non-digits in snapshot ID such as "re_" and "-" and "*" returns two values: first value is one of these: valid date string in df format - wayback date is valid (including the text string 'index' when date is '/*/') empty string - wayback date is malformed (less than 8 digits, not a valid date) nil - wayback date is '/save/' or otherwise not a number second return value is an appropriate 'message' may or may not be formatted ]] local function decodeWaybackDate(path, df) local msg, snapdate; snapdate = path:gsub ('^/web/', ):gsub ('^/all/', ):gsub ('^/ukgwa/', ):gsub ('^/tna/', ):gsub ('^/', ); -- remove leading /web/, /all/, /ukgwa/, /tna/, or / snapdate = snapdate:match ('^[^/]+'); -- get timestamp if snapdate == "*" then -- eg. /web/*/http.., etc. return 'index'; -- return indicator that this url has an index date end snapdate = snapdate:gsub ('%a%a_%d?$', ):gsub ('%-', ); -- from date, remove any trailing "re_", dashes msg = ; if snapdate:match ('%*$') then -- a trailing '*' causes calendar display at archive .org snapdate = snapdate:gsub ('%*$', ); -- remove so not part of length calc later msg = inlineRed (err_warn_msgs.ts_cal, 'warning'); -- make a message end if not tonumber(snapdate) then return nil, 'ts_nan'; -- return nil (fatal error flag) and message selector end local dlen = snapdate:len(); if dlen < 8 then -- we need 8 digits TODO: but shouldn't this be testing for 14 digits? return , inlineRed (err_warn_msgs.ts_short, 'error'); -- return empty string and error message end local year, month, day = snapdate:match ('(%d%d%d%d)(%d%d)(%d%d)'); -- no need for snapdatelong here if not is_valid_date (year, month, day) then return , inlineRed (err_warn_msgs.ts_date, 'error'); -- return empty string and error message end snapdate = table.concat ({year, month, day}, '-'); -- date comparisons are all done in iso format if 14 == dlen then return snapdate, msg; -- return date with message if any else return snapdate, msg .. inlineRed (err_warn_msgs.ts_len, 'warning'); -- return date with warning message(s) end end --[[--------------------------< decodeArchiveisDate >----------------------- Given an Archive.is "long link" URI-path (e.g. /2016.08.28-144552/http://example.com) return the date in df format (e.g. if df = dmy, return 28 August 2016) Handles "." and "-" in snapshot date, so 2016.08.28-144552 is same as 20160828144552 returns two values: first value is one of these: valid date string in df format - archive.is date is valid (including the text string 'short link' when url is the short form) empty string - wayback date is malformed (not a number, less than 8 digits, not a valid date) nil - wayback date is '/save/' second return value is an appropriate 'message' may or may not be formatted ]] local function decodeArchiveisDate(path, df) local snapdate if path:match ('^/%w+$') then -- short form url path is '/' followed by some number of base 62 digits and nothing else return "short link" -- e.g. http://archive.is/hD1qz end snapdate = mw.text.split (path, '/')[2]:gsub('[%.%-]', ); -- get snapshot date, e.g. 2016.08.28-144552; remove periods and hyphens local dlen = string.len(snapdate) if dlen < 8 then -- we need 8 digits TODO: but shouldn't this be testing for 14 digits? return , inlineRed (err_warn_msgs.ts_short, 'error'); -- return empty string and error message end local year, month, day = snapdate:match ('(%d%d%d%d)(%d%d)(%d%d)'); -- no need for snapdatelong here if not is_valid_date (year, month, day) then return , inlineRed (err_warn_msgs.ts_date, 'error'); -- return empty string and error message end snapdate = table.concat ({year, month, day}, '-'); -- date comparisons are all done in iso format if 14 == dlen then return snapdate; -- return date else return snapdate, inlineRed (err_warn_msgs.ts_len, 'warning'); -- return date with warning message end end --[[--------------------------< serviceName >----------------------- Given a domain extracted by mw.uri.new() (eg. web.archive.org) set tail string and service ID ]] local function serviceName(host, no_link) local tracking; local index; host = host:lower():gsub ('^web%.(.+)', '%1'):gsub ('^www%.(.+)', '%1'); -- lowercase, remove web. and www. subdomains if services[host] then index = host; else for k, _ in pairs (services) do if host:find ('%f[%a]'..k:gsub ('([%.%-])', '%%%1')) then index = k; break; end end end if index then local out = {}; -- empty string in [1] so that concatenated result has leading single space ulx.url1.service = services[index][4] or 'other'; tracking = services[index][5] or categories.other; -- build tail string if false == services[index][1] then -- select prefix table.insert (out, prefixes.at); elseif true == services[index][1] then table.insert (out, prefixes.atthe); else table.insert (out, services[index][1]); end table.insert (out, make_wikilink (services[index][2], services[index][3], no_link)); -- add article wikilink if services[index][6] then -- add tail postfix if it exists table.insert (out, services[index][6]); end ulx.url1.tail = table.concat (out, ' '); -- put it all together; result has leading space character else -- here when unknown archive ulx.url1.service = 'other'; tracking = categories.unknown; ulx.url1.tail = table.concat ({, prefixes.at, host, inlineRed (err_warn_msgs.unknown_url, error)}, ' '); end track[tracking] = 1 end --[[--------------------------< parseExtraArgs >----------------------- Parse numbered arguments starting at 2, such as url2..url10, date2..date10, title2..title10 For example: Template loop detected: Template:Webarchive Three url arguments not in numeric sequence (1..4..7). Function only processes arguments numbered 2 or greater (in this case 4 and 7) It creates numeric sequenced table entries like: urlx.url2.url = <argument value for url4> urlx.url3.url = <argument value for url7> Returns the number of URL arguments found numbered 2 or greater (in this case returns "2") ]] local function parseExtraArgs(args) local i, j, argurl, argurl2, argdate, argtitle j = 2 for i = 2, config.maxurls do argurl = "url" .. i if args[argurl] then argurl2 = "url" .. j ulx[argurl2] = {} ulx[argurl2]["url"] = args[argurl] argdate = "date" .. i if args[argdate] then ulx[argurl2]["date"] = args[argdate] else ulx[argurl2]["date"] = inlineRed (err_warn_msgs.date_miss, 'warning'); end argtitle = "title" .. i if args[argtitle] then ulx[argurl2]["title"] = args[argtitle] else ulx[argurl2]["title"] = nil end j = j + 1 end end if j == 2 then return 0 else return j - 2 end end --[[--------------------------< comma >----------------------- Given a date string, return "," if it's MDY ]] local function comma(date) return (date and date:match ('%a+ +%d%d?(,) +%d%d%d%d')) or ; end --[[--------------------------< createRendering >----------------------- Return a rendering of the data in ulx[][] ]] local function createRendering() local displayfield local out = {}; local index_date, msg = ulx.url1.date:match ('(index)(.*)'); -- when ulx.url1.date extract 'index' text and message text (if there is a message) ulx.url1.date = ulx.url1.date:gsub ('index.*', 'index'); -- remove message if 'none' == ulx.url1.format then -- For Template:Wayback, Template:Webcite table.insert (out, '['); -- open extlink markup table.insert (out, ulx.url1.url); -- add url if ulx.url1.title then table.insert (out, ' ') -- the required space table.insert (out, ulx.url1.title) -- the title table.insert (out, ']'); -- close extlink markup table.insert (out, ulx.url1.tail); -- tail text if ulx.url1.date then table.insert (out, ' ('); -- open date text; TODO: why the html entity? replace with regular space? table.insert (out, 'index' == ulx.url1.date and s_text.archive or s_text.archived); -- add text table.insert (out, ' '); -- insert a space table.insert (out, ulx.url1.date); -- add date table.insert (out, ')'); -- close date text end else -- no title if index_date then -- when url date is 'index' table.insert (out, table.concat ({' ', s_text.Archive_index, ']'})); -- add the index link label table.insert (out, msg or ); -- add date mismatch message when url date is /*/ and |date= has valid date else table.insert (out, table.concat ({' ', s_text.Archived, '] '})); -- add link label for url has timestamp date (will include mismatch message if there is one) end if ulx.url1.date then if 'index' ~= ulx.url1.date then table.insert (out, ulx.url1.date); -- add date when data is not 'index' end table.insert (out, comma(ulx.url1.date)); -- add ',' if date format is mdy table.insert (out, ulx.url1.tail); -- add tail text else -- no date table.insert (out, ulx.url1.tail); -- add tail text end end if 0 < ulx.url1.extraurls then -- For multiple archive URLs local tot = ulx.url1.extraurls + 1 table.insert (out, '.') -- terminate first url table.insert (out, table.concat ({' ', s_text.addlarchives, ': '})); -- add header text for i=2, tot do -- loop through the additionals local index = table.concat ({'url', i}); -- make an index displayfield = ulx[index]['title'] and 'title' or 'date'; -- choose display text table.insert (out, '['); -- open extlink markup table.insert (out, ulx[index]['url']); -- add the url table.insert (out, ' '); -- the required space table.insert (out, ulx[index][displayfield]); -- add the label table.insert (out, ']'); -- close extlink markup table.insert (out, i==tot and '.' or ', '); -- add terminator end end return table.concat (out); -- make a big string and done else -- For Template:Cite archives if 'addlarchives' == ulx.url1.format then -- Multiple archive services table.insert (out, table.concat ({s_text.addlarchives, ': '})); -- add header text else -- Multiple pages from the same archive table.insert (out, table.concat ({s_text.addlpages, ' '})); -- add header text table.insert (out, ulx.url1.date); -- add date to header text table.insert (out, ': '); -- close header text end local tot = ulx.url1.extraurls + 1; for i=1, tot do -- loop through the additionals local index = table.concat ({'url', i}); -- make an index table.insert (out, '['); -- open extlink markup table.insert (out, ulx[index]['url']); -- add url table.insert (out, ' '); -- add required space displayfield = ulx[index]['title']; if 'addlarchives' == ulx.url1.format then if not displayfield then displayfield = ulx[index]['date'] end else -- must be addlpages if not displayfield then displayfield = table.concat ({s_text.Page, ' ', i}); end end table.insert (out, displayfield); -- add title, date, page label text table.insert (out, ']'); -- close extlink markup table.insert (out, (i==tot and '.' or ', ')); -- add terminator end return table.concat (out); -- make a big string and done end end --[[--------------------------< P A R A M E T E R _ N A M E _ X L A T E >-------------------------------------- for internaltionalization, translate local-language parameter names to their English equivalents TODO: return error message if multiple aliases of the same canonical parameter name are found? returns two tables: new_args - holds canonical form parameters and their values either from translation or because the parameter was already in canonical form origin - maps canonical-form parameter names to their untranslated (local language) form for error messaging in the local language unrecognized parameters are ignored ]] local function parameter_name_xlate (args, params, enum_params) local name; -- holds modifiable name of the parameter name during evaluation local enum; -- for enumerated parameters, holds the enumerator during evaluation local found = false; -- flag used to break out of nested for loops local new_args = {}; -- a table that holds canonical and translated parameter k/v pairs local origin = {}; -- a table that maps original (local language) parameter names to their canonical name for local language error messaging local unnamed_params; -- set true when unsupported positional parameters are detected for k, v in pairs (args) do -- loop through all of the arguments in the args table name = k; -- copy of original parameter name if 'string' == type (k) then if non_western_digits then -- true when non-western digits supported at this wiki name = mw.ustring.gsub (name, '%d', digits); -- convert this wiki's non-western digits to western digits end enum = name:match ('%d+$'); -- get parameter enumerator if it exists; nil else if not enum then -- no enumerator so looking for non-enumnerated parameters -- TODO: insert shortcut here? if params[name] then name holds the canonical parameter name; no need to search further for pname, aliases in pairs (params) do -- loop through each parameter the params table for _, alias in ipairs (aliases) do -- loop through each alias in the parameter's aliases table if name == alias then new_args[pname] = v; -- create a new entry in the new_args table origin [pname] = k; -- create an entry to make canonical parameter name to original local language parameter name found = true; -- flag so that we can break out of these nested for loops break; -- no need to search the rest of the aliases table for name so go on to the next k, v pair end end if found then -- true when we found an alias that matched name found = false; -- reset the flag break; -- go do next args k/v pair end end else -- enumerated parameters name = name:gsub ('%d$', '#'); -- replace enumeration digits with place holder for table search -- TODO: insert shortcut here? if num_params[name] then name holds the canonical parameter name; no need to search further for pname, aliases in pairs (enum_params) do -- loop through each parameter the num_params table for _, alias in ipairs (aliases) do -- loop through each alias in the parameter's aliases table if name == alias then pname = pname:gsub ('#$', enum); -- replace the '#' place holder with the actual enumerator new_args[pname] = v; -- create a new entry in the new_args table origin [pname] = k; -- create an entry to make canonical parameter name to original local language parameter name found = true; -- flag so that we can break out of these nested for loops break; -- no need to search the rest of the aliases table for name so go on to the next k, v pair end end if found then -- true when we found an alias that matched name found = false; -- reset the flag break; -- go do next args k/v pair end end end else unnamed_params = true; -- flag for unsupported positional parameters end end -- for k, v return new_args, origin, unnamed_params; end --[[--------------------------< W E B A R C H I V E >---------------------------------------------------------- template entry point ]] local function webarchive(frame) local args = getArgs (frame); local data = mw.loadData (table.concat ({ -- make a data module name; sandbox or live 'Module:Webarchive/data', frame:getTitle():find('sandbox', 1, true) and '/sandbox' or -- this instance is ./sandbox then append /sandbox })); categories = data.categories; -- fill in the forward declarations config = data.config; if data.digits.enable then digits = data.digits; -- for i18n; table of digits in the local wiki's language non_western_digits = true; -- use_non_western_digits end err_warn_msgs = data.err_warn_msgs; excepted_pages = data.excepted_pages; month_num = data.month_num; -- for i18n; table of month names in the local wiki's language prefixes = data.prefixes; services = data.services; s_text = data.s_text; uncategorized_namespaces = data.uncategorized_namespaces; uncategorized_subpages = data.uncategorized_subpages; local origin = {}; -- holds a map of English to local language parameter names used in the current template; not currently used local unnamed_params; -- boolean set to true when template call has unnamed parameters args, origin, unnamed_params = parameter_name_xlate (args, data.params, data.enum_params); -- translate parameter names in args to English local date, format, msg, udate, uri, url; local ldf = 'iso'; -- when there is no |date= parameter, render url dates in iso format if args.url and args.url1 then -- URL argument (first) return inlineError (data.crit_err_msgs.conflicting, {origin.url, origin.url1}); end url = args.url or args.url1; if not url then return inlineError (data.crit_err_msgs.empty); end -- these iabot bugs perportedly fixed; removing these causes lua script error --[[ -- at Template:Webarchive/testcases/Production; resolve that before deleting these tests if mw.ustring.find( url, "https://web.http", 1, true ) then -- track bug - TODO: IAbot bug; not known if the bug has been fixed; deferred track[categories.error] = 1; return inlineError (data.crit_err_msgs.iabot1); end if url == "https://web.archive.org/http:/" then -- track bug - TODO: IAbot bug; not known if the bug has been fixed; deferred track[categories.error] = 1; return inlineError (data.crit_err_msgs.iabot2); end ]] if not (url:lower():find ('^http') or url:find ('^//')) then return inlineError (data.crit_err_msgs.invalid_url ); end ulx.url1 = {} ulx.url1.url = url ulx.url1.extraurls = parseExtraArgs(args) local good = false; good, uri = pcall (mw.uri.new, ulx.url1.url); -- get a table of uri parts from this url; protected mode to prevent lua error when ulx.url1.url is malformed if not good or nil == uri.host then -- abandon when ulx.url1.url is malformed return inlineError (data.crit_err_msgs.invalid_url); end serviceName(uri.host, args.nolink) if args.date and args.date1 then -- Date argument return inlineError (data.crit_err_msgs.conflicting, {origin.date, origin.date1}); end date = args.date or args.date1; date = date and date:gsub (' +', ' '); -- replace multiple spaces with a single space if date and config.verifydates then if '*' == date then date = 'index'; ldf = 'iso'; -- set to default format elseif 'mdy' == date then date = nil; -- if date extracted from URL, ldf = 'mdy'; -- then |date=mdy overrides iso elseif 'dmy' == date then date = nil; -- if date extracted from URL, ldf = 'dmy'; -- then |date=dmy overrides iso elseif 'ymd' == date then date = nil; -- if date extracted from URL, ldf = 'ymd'; -- then |date=ymd overrides iso else date, ldf = decode_date (date); -- get an iso format date from date and get date's original format end end if 'wayback' == ulx.url1.service or 'locwebarchives' == ulx.url1.service or 'ukgwa' == ulx.url1.service then if date then if config.verifydates then if ldf then udate, msg = decodeWaybackDate (uri.path); -- get the url date in iso format and format of date in |date=; 'index' when wayback url date is * if not udate then -- this is the only 'fatal' error return return inlineError (data.crit_err_msgs[msg]); end if udate ~= date then -- date comparison using iso format dates date = udate; msg = table.concat ({ inlineRed (err_warn_msgs.mismatch, 'warning'), -- add warning message msg, -- add message if there is one }); end end end else -- no |date= udate, msg = decodeWaybackDate (uri.path); if not udate then -- this is the only 'fatal' error return return inlineError (data.crit_err_msgs[msg]); end if == udate then date = nil; -- unset else date = udate; end end elseif 'webcite' == ulx.url1.service then if date then if config.verifydates then if ldf then udate = decodeWebciteDate (uri.path); -- get the url date in iso format if 'query' ~= udate then -- skip if query if udate ~= date then -- date comparison using iso format dates date = udate; msg = table.concat ({ inlineRed (err_warn_msgs.mismatch, 'warning'), }); end end end end else date = decodeWebciteDate( uri.path, "iso" ) if date == "query" then date = nil; -- unset msg = inlineRed (err_warn_msgs.date_miss, 'warning'); elseif not date then -- invalid base62 string date = inlineRed (err_warn_msgs.date1, 'error'); end end elseif 'archiveis' == ulx.url1.service then if date then if config.verifydates then if ldf then udate, msg = decodeArchiveisDate (uri.path) -- get the url date in iso format if 'short link' ~= udate then -- skip if short link if udate ~= date then -- date comparison using iso format dates date = udate; msg = table.concat ({ inlineRed (err_warn_msgs.mismatch, 'warning'), -- add warning message msg, -- add message if there is one }); end end end end else -- no |date= udate, msg = decodeArchiveisDate( uri.path, "iso" ) if udate == "short link" then date = nil; -- unset msg = inlineRed (err_warn_msgs.date_miss, 'warning'); elseif == udate then date = nil; -- unset else date = udate; end end else -- some other service if not date then msg = inlineRed (err_warn_msgs.date_miss, 'warning'); end end if 'index' == date then ulx.url1.date = date .. (msg or ); -- create index + message (if there is one) elseif date then ulx.url1.date = makeDate (date, nil, nil, ldf) .. (msg or ); -- create a date in the wiki's local language + message (if there is one) else ulx.url1.date = msg; end format = args.format; -- Format argument if not format then format = "none" else for k, v in pairs (data.format_vals) do -- |format= accepts two specific values loop through a table of those values local found; -- declare a nil flag for _, p in ipairs (v) do -- loop through local language variants if format == p then -- when |format= value matches format = k; -- use name from table key found = true; -- declare found so that we can break out of outer for loop break; -- break out of inner for loop end end if found then break; end end if format == "addlpages" then if not ulx.url1.date then format = "none" end elseif format == "addlarchives" then format = "addlarchives" else format = "none" end end ulx.url1.format = format if args.title and args.title1 then -- Title argument return inlineError (data.crit_err_msgs.conflicting, {origin.title, origin.title1}); end ulx.url1.title = args.title or args.title1; local rend = createRendering() if not rend then return inlineError (data.crit_err_msgs.unknown); end return rend .. ((unnamed_params and inlineRed (err_warn_msgs.unnamed_params, 'warning')) or ) .. createTracking(); end --[[--------------------------< E X P O R T E D F U N C T I O N S >------------------------------------------ ]] return {webarchive = webarchive}; AMC's Filmsite. Retrieved August 15, 2017.
  57. ↑ The 100 Greatest Movie Moments. --[[ ---------------------------------- Lua module implementing the Template loop detected: Template:Webarchive template. A merger of the functionality of three templates: Template:Wayback, Template:Webcite and Template:Cite archives ]] --[[--------------------------< D E P E N D E N C I E S >------------------------------------------------------ ]] require('Module:No globals'); local getArgs = require ('Module:Arguments').getArgs; --[[--------------------------< F O R W A R D D E C L A R A T I O N S >-------------------------------------- ]] local categories = {}; -- category names local config = {}; -- global configuration settings local digits = {}; -- for i18n; table that translates local-wiki digits to western digits local err_warn_msgs = {}; -- error and warning messages local excepted_pages = {}; local month_num = {}; -- for i18n; table that translates local-wiki month names to western digits local prefixes = {}; -- service provider tail string prefixes local services = {}; -- archive service provider data from local s_text = {}; -- table of static text strings used to build final rendering local uncategorized_namespaces = {}; -- list of namespaces that we should not categorize local uncategorized_subpages = {}; -- list of subpages that should not be categorized --[[--------------------------< P A G E S C O P E I D E N T I F I E R S >---------------------------------- ]] local non_western_digits; -- boolean flag set true when data.digits.enable is true local this_page = mw.title.getCurrentTitle(); local track = {}; -- Associative array to hold tracking categories local ulx = {}; -- Associative array to hold template data --[[--------------------------< S U B S T I T U T E >---------------------------------------------------------- Populates numbered arguments in a message string using an argument table. ]] local function substitute (msg, args) return args and mw.message.newRawMessage (msg, args):plain() or msg; end --[[--------------------------< tableLength >----------------------- Given a 1-D table, return number of elements ]] local function tableLength(T) local count = 0 for _ in pairs(T) do count = count + 1 end return count end --[=[-------------------------< M A K E _ W I K I L I N K >---------------------------------------------------- Makes a wikilink; when both link and display text is provided, returns a wikilink in the form D; if only link is provided, returns a wikilink in the form L; if neither are provided or link is omitted, returns an empty string. ]=] local function make_wikilink (link, display, no_link) if nil == no_link then if link and ( ~= link) then if display and ( ~= display) then return table.concat ({'', display, ''}); else return table.concat ({'', link, ''}); end end return display or ; -- link not set so return the display text else -- no_link if display and ( ~= display) then -- if there is display text return display; -- return that else return link or ; -- return the target article name or empty string end end end --[[--------------------------< createTracking >----------------------- Return data in track[] ie. tracking categories ]] local function createTracking() if not excepted_pages[this_page.fullText] then -- namespace:title/fragment is allowed to be categorized (typically this module's / template's testcases page(s)) if uncategorized_namespaces[this_page.nsText] then return ; -- this page not to be categorized so return empty string end for _,v in ipairs (uncategorized_subpages) do -- cycle through page name patterns if this_page.text:match (v) then -- test page name against each pattern return ; -- this subpage type not to be categorized so return empty string end end end local out = {}; if tableLength(track) > 0 then for key, _ in pairs(track) do -- loop through table table.insert (out, make_wikilink (key)); -- and convert category names to links end end return table.concat (out); -- concat into one big string; empty string if table is empty end --[[--------------------------< inlineError >----------------------- Critical error. Render output completely in red. Add to tracking category. This function called as the last thing before abandoning this module ]] local function inlineError (msg, args) track[categories.error] = 1 return table.concat ({ 'Error in ', -- open the error message span config.tname, -- insert the local language template name ' template: ', substitute (msg, args), -- insert the formatted error message '.', -- close the span createTracking() -- add the category }) end --[[--------------------------< inlineRed >----------------------- Render a text fragment in red, such as a warning as part of the final output. Add tracking category. ]] local function inlineRed(msg, trackmsg) if trackmsg == "warning" then track[categories.warning] = 1; elseif trackmsg == "error" then track[categories.error] = 1; end return '' .. msg .. '' end --[[--------------------------< base62 >----------------------- Convert base-62 to base-10 Credit: https://de.wikipedia.org/wiki/Modul:Expr ]] local function base62( value ) local r = 1 -- default return value is input value is malformed if value:match ('%W') then -- value must only be in the set [0-9a-zA-Z] return; -- nil return when value contains extraneous characters end local n = #value -- number of characters in value local k = 1 local c r = 0 for i = n, 1, -1 do -- loop through all characters in value from ls digit to ms digit c = value:byte( i, i ) if c >= 48 and c <= 57 then -- character is digit 0-9 c = c - 48 elseif c >= 65 and c <= 90 then -- character is ascii a-z c = c - 55 else -- must be ascii A-Z c = c - 61 end r = r + c * k -- accumulate this base62 character's value k = k * 62 -- bump for next end -- for i return r end --[[--------------------------< D E C O D E _ D A T E >-------------------------------------------------------- Given a date string, return it in iso format along with an indicator of the date's format. Except that month names must be recognizable as legitimate month names with proper capitalization, and that the date string must match one of the recognized date formats, no error checking is done here; return nil else ]] local function decode_date (date_str) local patterns = { ['dmy'] = {'^(%d%d?) +([^%s%d]+) +(%d%d%d%d)$', 'd', 'm', 'y'}, -- %a does not recognize unicode combining characters used by some languages ['mdy'] = {'^([^%s%d]+) (%d%d?), +(%d%d%d%d)$', 'm', 'd', 'y'}, ['ymd'] = {'^(%d%d%d%d) +([^%s%d]+) (%d%d?)$', 'y', 'm', 'd'}, -- not mos compliant at en.wiki but may be acceptible at other wikis }; local t = {}; if non_western_digits then -- this wiki uses non-western digits? date_str = mw.ustring.gsub (date_str, '%d', digits); -- convert this wiki's non-western digits to western digits end if date_str:match ('^%d%d%d%d%-%d%d%-%d%d$') then -- already an iso format date, return western digits form return date_str, 'iso'; end for k, v in pairs (patterns) do local c1, c2, c3 = mw.ustring.match (date_str, patterns[k][1]); -- c1 .. c3 are captured but we don't know what they hold if c1 then -- set on match t = { -- translate unspecified captures to y, m, and d [patterns[k][2]] = c1, -- fill the table of captures with the captures [patterns[k][3]] = c2, -- take index names from src_pattern table and assign sequential captures [patterns[k][4]] = c3, }; if month_num[t.m] then -- when month not already a number t.m = month_num[t.m]; -- replace valid month name with a number else return nil, 'iso'; -- not a valid date form because month not valid end return mw.ustring.format ('%.4d-%.2d-%.2d', t.y, t.m, t.d), k; -- return date in iso format end end return nil, 'iso'; -- date could not be decoded; return nil and default iso date end --[[--------------------------< makeDate >----------------------- Given year, month, day numbers, (zero-padded or not) return a full date in df format where df may be one of: mdy, dmy, iso, ymd on entry, year, month, day are presumed to be correct for the date that they represent; all are required in this module, makeDate() is sometimes given an iso-format date in year: makeDate (2018-09-20, nil, nil, df) this works because table.concat() sees only one table member ]] local function makeDate (year, month, day, df) local format = { ['dmy'] = 'j F Y', ['mdy'] = 'F j, Y', ['ymd'] = 'Y F j', ['iso'] = 'Y-m-d', }; local date = table.concat ({year, month, day}, '-'); -- assemble year-initial numeric-format date (zero padding not required here) if non_western_digits then -- this wiki uses non-western digits? date = mw.ustring.gsub (date, '%d', digits); -- convert this wiki's non-western digits to western digits end return mw.getContentLanguage():formatDate (format[df], date); end --[[--------------------------< I S _ V A L I D _ D A T E >---------------------------------------------------- Returns true if date is after 31 December 1899 (why is 1900 the min year? shouldn't the internet's date-of-birth be min year?), not after today's date, and represents a valid date (29 February 2017 is not a valid date). Applies Gregorian leapyear rules. all arguments are required ]] local function is_valid_date (year, month, day) local days_in_month = {31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31}; local month_length; local y, m, d; local today = os.date ('*t'); -- fetch a table of current date parts if not year or == year or not month or == month or not day or == day then return false; -- something missing end y = tonumber (year); m = tonumber (month); d = tonumber (day); if 1900 > y or today.year < y or 1 > m or 12 < m then -- year and month are within bounds TODO: 1900? return false; end if (2==m) then -- if February month_length = 28; -- then 28 days unless if (0==(y%4) and (0~=(y%100) or 0==(y%400))) then -- is a leap year? month_length = 29; -- if leap year then 29 days in February end else month_length=days_in_month[m]; end if 1 > d or month_length < d then -- day is within bounds return false; end -- here when date parts represent a valid date return os.time({['year']=y, ['month']=m, ['day']=d, ['hour']=0}) <= os.time(); -- date at midnight must be less than or equal to current date/time end --[[--------------------------< decodeWebciteDate >----------------------- Given a URI-path to Webcite (eg. /67xHmVFWP) return the encoded date in df format returns date string in df format - webcite date is a unix timestamp encoded as bae62 or the string 'query' ]] local function decodeWebciteDate(path, df) local dt = {}; local decode; dt = mw.text.split(path, "/") -- valid URL formats that are not base62 -- http://www.webcitation.org/query?id=1138911916587475 -- http://www.webcitation.org/query?url=http..&date=2012-06-01+21:40:03 -- http://www.webcitation.org/1138911916587475 -- http://www.webcitation.org/cache/73e53dd1f16cf8c5da298418d2a6e452870cf50e -- http://www.webcitation.org/getfile.php?fileid=1c46e791d68e89e12d0c2532cc3cf629b8bc8c8e if dt[2]:find ('query', 1, true) or dt[2]:find ('cache', 1, true) or dt[2]:find ('getfile', 1, true) or tonumber(dt[2]) then return 'query'; end decode = base62(dt[2]); -- base62 string -> exponential number if not decode then return nil; -- nil return when dt[2] contains characters not in %w end dt = os.date('*t', string.format("%d", decode):sub(1,10)) -- exponential number -> text -> first 10 characters (a unix timestamp) -> a table of date parts decode = makeDate (dt.year, dt.month, dt.day, 'iso'); -- date comparisons are all done in iso format with western digits if non_western_digits then -- this wiki uses non-western digits? decode = mw.ustring.gsub (decode, '%d', digits); -- convert this wiki's non-western digits to western digits end return decode; end --[[--------------------------< decodeWaybackDate >----------------------- Given a URI-path to Wayback (eg. /web/20160901010101/http://example.com ) or Library of Congress Web Archives (eg. /all/20160901010101/http://example.com) or UK Government Web Archive (eg. /ukgwa/20160901010101/http://example.com or /tna/20160901010101/http://example.com) return the formatted date eg. "September 1, 2016" in df format Handle non-digits in snapshot ID such as "re_" and "-" and "*" returns two values: first value is one of these: valid date string in df format - wayback date is valid (including the text string 'index' when date is '/*/') empty string - wayback date is malformed (less than 8 digits, not a valid date) nil - wayback date is '/save/' or otherwise not a number second return value is an appropriate 'message' may or may not be formatted ]] local function decodeWaybackDate(path, df) local msg, snapdate; snapdate = path:gsub ('^/web/', ):gsub ('^/all/', ):gsub ('^/ukgwa/', ):gsub ('^/tna/', ):gsub ('^/', ); -- remove leading /web/, /all/, /ukgwa/, /tna/, or / snapdate = snapdate:match ('^[^/]+'); -- get timestamp if snapdate == "*" then -- eg. /web/*/http.., etc. return 'index'; -- return indicator that this url has an index date end snapdate = snapdate:gsub ('%a%a_%d?$', ):gsub ('%-', ); -- from date, remove any trailing "re_", dashes msg = ; if snapdate:match ('%*$') then -- a trailing '*' causes calendar display at archive .org snapdate = snapdate:gsub ('%*$', ); -- remove so not part of length calc later msg = inlineRed (err_warn_msgs.ts_cal, 'warning'); -- make a message end if not tonumber(snapdate) then return nil, 'ts_nan'; -- return nil (fatal error flag) and message selector end local dlen = snapdate:len(); if dlen < 8 then -- we need 8 digits TODO: but shouldn't this be testing for 14 digits? return , inlineRed (err_warn_msgs.ts_short, 'error'); -- return empty string and error message end local year, month, day = snapdate:match ('(%d%d%d%d)(%d%d)(%d%d)'); -- no need for snapdatelong here if not is_valid_date (year, month, day) then return , inlineRed (err_warn_msgs.ts_date, 'error'); -- return empty string and error message end snapdate = table.concat ({year, month, day}, '-'); -- date comparisons are all done in iso format if 14 == dlen then return snapdate, msg; -- return date with message if any else return snapdate, msg .. inlineRed (err_warn_msgs.ts_len, 'warning'); -- return date with warning message(s) end end --[[--------------------------< decodeArchiveisDate >----------------------- Given an Archive.is "long link" URI-path (e.g. /2016.08.28-144552/http://example.com) return the date in df format (e.g. if df = dmy, return 28 August 2016) Handles "." and "-" in snapshot date, so 2016.08.28-144552 is same as 20160828144552 returns two values: first value is one of these: valid date string in df format - archive.is date is valid (including the text string 'short link' when url is the short form) empty string - wayback date is malformed (not a number, less than 8 digits, not a valid date) nil - wayback date is '/save/' second return value is an appropriate 'message' may or may not be formatted ]] local function decodeArchiveisDate(path, df) local snapdate if path:match ('^/%w+$') then -- short form url path is '/' followed by some number of base 62 digits and nothing else return "short link" -- e.g. http://archive.is/hD1qz end snapdate = mw.text.split (path, '/')[2]:gsub('[%.%-]', ); -- get snapshot date, e.g. 2016.08.28-144552; remove periods and hyphens local dlen = string.len(snapdate) if dlen < 8 then -- we need 8 digits TODO: but shouldn't this be testing for 14 digits? return , inlineRed (err_warn_msgs.ts_short, 'error'); -- return empty string and error message end local year, month, day = snapdate:match ('(%d%d%d%d)(%d%d)(%d%d)'); -- no need for snapdatelong here if not is_valid_date (year, month, day) then return , inlineRed (err_warn_msgs.ts_date, 'error'); -- return empty string and error message end snapdate = table.concat ({year, month, day}, '-'); -- date comparisons are all done in iso format if 14 == dlen then return snapdate; -- return date else return snapdate, inlineRed (err_warn_msgs.ts_len, 'warning'); -- return date with warning message end end --[[--------------------------< serviceName >----------------------- Given a domain extracted by mw.uri.new() (eg. web.archive.org) set tail string and service ID ]] local function serviceName(host, no_link) local tracking; local index; host = host:lower():gsub ('^web%.(.+)', '%1'):gsub ('^www%.(.+)', '%1'); -- lowercase, remove web. and www. subdomains if services[host] then index = host; else for k, _ in pairs (services) do if host:find ('%f[%a]'..k:gsub ('([%.%-])', '%%%1')) then index = k; break; end end end if index then local out = {}; -- empty string in [1] so that concatenated result has leading single space ulx.url1.service = services[index][4] or 'other'; tracking = services[index][5] or categories.other; -- build tail string if false == services[index][1] then -- select prefix table.insert (out, prefixes.at); elseif true == services[index][1] then table.insert (out, prefixes.atthe); else table.insert (out, services[index][1]); end table.insert (out, make_wikilink (services[index][2], services[index][3], no_link)); -- add article wikilink if services[index][6] then -- add tail postfix if it exists table.insert (out, services[index][6]); end ulx.url1.tail = table.concat (out, ' '); -- put it all together; result has leading space character else -- here when unknown archive ulx.url1.service = 'other'; tracking = categories.unknown; ulx.url1.tail = table.concat ({, prefixes.at, host, inlineRed (err_warn_msgs.unknown_url, error)}, ' '); end track[tracking] = 1 end --[[--------------------------< parseExtraArgs >----------------------- Parse numbered arguments starting at 2, such as url2..url10, date2..date10, title2..title10 For example: Template loop detected: Template:Webarchive Three url arguments not in numeric sequence (1..4..7). Function only processes arguments numbered 2 or greater (in this case 4 and 7) It creates numeric sequenced table entries like: urlx.url2.url = <argument value for url4> urlx.url3.url = <argument value for url7> Returns the number of URL arguments found numbered 2 or greater (in this case returns "2") ]] local function parseExtraArgs(args) local i, j, argurl, argurl2, argdate, argtitle j = 2 for i = 2, config.maxurls do argurl = "url" .. i if args[argurl] then argurl2 = "url" .. j ulx[argurl2] = {} ulx[argurl2]["url"] = args[argurl] argdate = "date" .. i if args[argdate] then ulx[argurl2]["date"] = args[argdate] else ulx[argurl2]["date"] = inlineRed (err_warn_msgs.date_miss, 'warning'); end argtitle = "title" .. i if args[argtitle] then ulx[argurl2]["title"] = args[argtitle] else ulx[argurl2]["title"] = nil end j = j + 1 end end if j == 2 then return 0 else return j - 2 end end --[[--------------------------< comma >----------------------- Given a date string, return "," if it's MDY ]] local function comma(date) return (date and date:match ('%a+ +%d%d?(,) +%d%d%d%d')) or ; end --[[--------------------------< createRendering >----------------------- Return a rendering of the data in ulx[][] ]] local function createRendering() local displayfield local out = {}; local index_date, msg = ulx.url1.date:match ('(index)(.*)'); -- when ulx.url1.date extract 'index' text and message text (if there is a message) ulx.url1.date = ulx.url1.date:gsub ('index.*', 'index'); -- remove message if 'none' == ulx.url1.format then -- For Template:Wayback, Template:Webcite table.insert (out, '['); -- open extlink markup table.insert (out, ulx.url1.url); -- add url if ulx.url1.title then table.insert (out, ' ') -- the required space table.insert (out, ulx.url1.title) -- the title table.insert (out, ']'); -- close extlink markup table.insert (out, ulx.url1.tail); -- tail text if ulx.url1.date then table.insert (out, ' ('); -- open date text; TODO: why the html entity? replace with regular space? table.insert (out, 'index' == ulx.url1.date and s_text.archive or s_text.archived); -- add text table.insert (out, ' '); -- insert a space table.insert (out, ulx.url1.date); -- add date table.insert (out, ')'); -- close date text end else -- no title if index_date then -- when url date is 'index' table.insert (out, table.concat ({' ', s_text.Archive_index, ']'})); -- add the index link label table.insert (out, msg or ); -- add date mismatch message when url date is /*/ and |date= has valid date else table.insert (out, table.concat ({' ', s_text.Archived, '] '})); -- add link label for url has timestamp date (will include mismatch message if there is one) end if ulx.url1.date then if 'index' ~= ulx.url1.date then table.insert (out, ulx.url1.date); -- add date when data is not 'index' end table.insert (out, comma(ulx.url1.date)); -- add ',' if date format is mdy table.insert (out, ulx.url1.tail); -- add tail text else -- no date table.insert (out, ulx.url1.tail); -- add tail text end end if 0 < ulx.url1.extraurls then -- For multiple archive URLs local tot = ulx.url1.extraurls + 1 table.insert (out, '.') -- terminate first url table.insert (out, table.concat ({' ', s_text.addlarchives, ': '})); -- add header text for i=2, tot do -- loop through the additionals local index = table.concat ({'url', i}); -- make an index displayfield = ulx[index]['title'] and 'title' or 'date'; -- choose display text table.insert (out, '['); -- open extlink markup table.insert (out, ulx[index]['url']); -- add the url table.insert (out, ' '); -- the required space table.insert (out, ulx[index][displayfield]); -- add the label table.insert (out, ']'); -- close extlink markup table.insert (out, i==tot and '.' or ', '); -- add terminator end end return table.concat (out); -- make a big string and done else -- For Template:Cite archives if 'addlarchives' == ulx.url1.format then -- Multiple archive services table.insert (out, table.concat ({s_text.addlarchives, ': '})); -- add header text else -- Multiple pages from the same archive table.insert (out, table.concat ({s_text.addlpages, ' '})); -- add header text table.insert (out, ulx.url1.date); -- add date to header text table.insert (out, ': '); -- close header text end local tot = ulx.url1.extraurls + 1; for i=1, tot do -- loop through the additionals local index = table.concat ({'url', i}); -- make an index table.insert (out, '['); -- open extlink markup table.insert (out, ulx[index]['url']); -- add url table.insert (out, ' '); -- add required space displayfield = ulx[index]['title']; if 'addlarchives' == ulx.url1.format then if not displayfield then displayfield = ulx[index]['date'] end else -- must be addlpages if not displayfield then displayfield = table.concat ({s_text.Page, ' ', i}); end end table.insert (out, displayfield); -- add title, date, page label text table.insert (out, ']'); -- close extlink markup table.insert (out, (i==tot and '.' or ', ')); -- add terminator end return table.concat (out); -- make a big string and done end end --[[--------------------------< P A R A M E T E R _ N A M E _ X L A T E >-------------------------------------- for internaltionalization, translate local-language parameter names to their English equivalents TODO: return error message if multiple aliases of the same canonical parameter name are found? returns two tables: new_args - holds canonical form parameters and their values either from translation or because the parameter was already in canonical form origin - maps canonical-form parameter names to their untranslated (local language) form for error messaging in the local language unrecognized parameters are ignored ]] local function parameter_name_xlate (args, params, enum_params) local name; -- holds modifiable name of the parameter name during evaluation local enum; -- for enumerated parameters, holds the enumerator during evaluation local found = false; -- flag used to break out of nested for loops local new_args = {}; -- a table that holds canonical and translated parameter k/v pairs local origin = {}; -- a table that maps original (local language) parameter names to their canonical name for local language error messaging local unnamed_params; -- set true when unsupported positional parameters are detected for k, v in pairs (args) do -- loop through all of the arguments in the args table name = k; -- copy of original parameter name if 'string' == type (k) then if non_western_digits then -- true when non-western digits supported at this wiki name = mw.ustring.gsub (name, '%d', digits); -- convert this wiki's non-western digits to western digits end enum = name:match ('%d+$'); -- get parameter enumerator if it exists; nil else if not enum then -- no enumerator so looking for non-enumnerated parameters -- TODO: insert shortcut here? if params[name] then name holds the canonical parameter name; no need to search further for pname, aliases in pairs (params) do -- loop through each parameter the params table for _, alias in ipairs (aliases) do -- loop through each alias in the parameter's aliases table if name == alias then new_args[pname] = v; -- create a new entry in the new_args table origin [pname] = k; -- create an entry to make canonical parameter name to original local language parameter name found = true; -- flag so that we can break out of these nested for loops break; -- no need to search the rest of the aliases table for name so go on to the next k, v pair end end if found then -- true when we found an alias that matched name found = false; -- reset the flag break; -- go do next args k/v pair end end else -- enumerated parameters name = name:gsub ('%d$', '#'); -- replace enumeration digits with place holder for table search -- TODO: insert shortcut here? if num_params[name] then name holds the canonical parameter name; no need to search further for pname, aliases in pairs (enum_params) do -- loop through each parameter the num_params table for _, alias in ipairs (aliases) do -- loop through each alias in the parameter's aliases table if name == alias then pname = pname:gsub ('#$', enum); -- replace the '#' place holder with the actual enumerator new_args[pname] = v; -- create a new entry in the new_args table origin [pname] = k; -- create an entry to make canonical parameter name to original local language parameter name found = true; -- flag so that we can break out of these nested for loops break; -- no need to search the rest of the aliases table for name so go on to the next k, v pair end end if found then -- true when we found an alias that matched name found = false; -- reset the flag break; -- go do next args k/v pair end end end else unnamed_params = true; -- flag for unsupported positional parameters end end -- for k, v return new_args, origin, unnamed_params; end --[[--------------------------< W E B A R C H I V E >---------------------------------------------------------- template entry point ]] local function webarchive(frame) local args = getArgs (frame); local data = mw.loadData (table.concat ({ -- make a data module name; sandbox or live 'Module:Webarchive/data', frame:getTitle():find('sandbox', 1, true) and '/sandbox' or -- this instance is ./sandbox then append /sandbox })); categories = data.categories; -- fill in the forward declarations config = data.config; if data.digits.enable then digits = data.digits; -- for i18n; table of digits in the local wiki's language non_western_digits = true; -- use_non_western_digits end err_warn_msgs = data.err_warn_msgs; excepted_pages = data.excepted_pages; month_num = data.month_num; -- for i18n; table of month names in the local wiki's language prefixes = data.prefixes; services = data.services; s_text = data.s_text; uncategorized_namespaces = data.uncategorized_namespaces; uncategorized_subpages = data.uncategorized_subpages; local origin = {}; -- holds a map of English to local language parameter names used in the current template; not currently used local unnamed_params; -- boolean set to true when template call has unnamed parameters args, origin, unnamed_params = parameter_name_xlate (args, data.params, data.enum_params); -- translate parameter names in args to English local date, format, msg, udate, uri, url; local ldf = 'iso'; -- when there is no |date= parameter, render url dates in iso format if args.url and args.url1 then -- URL argument (first) return inlineError (data.crit_err_msgs.conflicting, {origin.url, origin.url1}); end url = args.url or args.url1; if not url then return inlineError (data.crit_err_msgs.empty); end -- these iabot bugs perportedly fixed; removing these causes lua script error --[[ -- at Template:Webarchive/testcases/Production; resolve that before deleting these tests if mw.ustring.find( url, "https://web.http", 1, true ) then -- track bug - TODO: IAbot bug; not known if the bug has been fixed; deferred track[categories.error] = 1; return inlineError (data.crit_err_msgs.iabot1); end if url == "https://web.archive.org/http:/" then -- track bug - TODO: IAbot bug; not known if the bug has been fixed; deferred track[categories.error] = 1; return inlineError (data.crit_err_msgs.iabot2); end ]] if not (url:lower():find ('^http') or url:find ('^//')) then return inlineError (data.crit_err_msgs.invalid_url ); end ulx.url1 = {} ulx.url1.url = url ulx.url1.extraurls = parseExtraArgs(args) local good = false; good, uri = pcall (mw.uri.new, ulx.url1.url); -- get a table of uri parts from this url; protected mode to prevent lua error when ulx.url1.url is malformed if not good or nil == uri.host then -- abandon when ulx.url1.url is malformed return inlineError (data.crit_err_msgs.invalid_url); end serviceName(uri.host, args.nolink) if args.date and args.date1 then -- Date argument return inlineError (data.crit_err_msgs.conflicting, {origin.date, origin.date1}); end date = args.date or args.date1; date = date and date:gsub (' +', ' '); -- replace multiple spaces with a single space if date and config.verifydates then if '*' == date then date = 'index'; ldf = 'iso'; -- set to default format elseif 'mdy' == date then date = nil; -- if date extracted from URL, ldf = 'mdy'; -- then |date=mdy overrides iso elseif 'dmy' == date then date = nil; -- if date extracted from URL, ldf = 'dmy'; -- then |date=dmy overrides iso elseif 'ymd' == date then date = nil; -- if date extracted from URL, ldf = 'ymd'; -- then |date=ymd overrides iso else date, ldf = decode_date (date); -- get an iso format date from date and get date's original format end end if 'wayback' == ulx.url1.service or 'locwebarchives' == ulx.url1.service or 'ukgwa' == ulx.url1.service then if date then if config.verifydates then if ldf then udate, msg = decodeWaybackDate (uri.path); -- get the url date in iso format and format of date in |date=; 'index' when wayback url date is * if not udate then -- this is the only 'fatal' error return return inlineError (data.crit_err_msgs[msg]); end if udate ~= date then -- date comparison using iso format dates date = udate; msg = table.concat ({ inlineRed (err_warn_msgs.mismatch, 'warning'), -- add warning message msg, -- add message if there is one }); end end end else -- no |date= udate, msg = decodeWaybackDate (uri.path); if not udate then -- this is the only 'fatal' error return return inlineError (data.crit_err_msgs[msg]); end if == udate then date = nil; -- unset else date = udate; end end elseif 'webcite' == ulx.url1.service then if date then if config.verifydates then if ldf then udate = decodeWebciteDate (uri.path); -- get the url date in iso format if 'query' ~= udate then -- skip if query if udate ~= date then -- date comparison using iso format dates date = udate; msg = table.concat ({ inlineRed (err_warn_msgs.mismatch, 'warning'), }); end end end end else date = decodeWebciteDate( uri.path, "iso" ) if date == "query" then date = nil; -- unset msg = inlineRed (err_warn_msgs.date_miss, 'warning'); elseif not date then -- invalid base62 string date = inlineRed (err_warn_msgs.date1, 'error'); end end elseif 'archiveis' == ulx.url1.service then if date then if config.verifydates then if ldf then udate, msg = decodeArchiveisDate (uri.path) -- get the url date in iso format if 'short link' ~= udate then -- skip if short link if udate ~= date then -- date comparison using iso format dates date = udate; msg = table.concat ({ inlineRed (err_warn_msgs.mismatch, 'warning'), -- add warning message msg, -- add message if there is one }); end end end end else -- no |date= udate, msg = decodeArchiveisDate( uri.path, "iso" ) if udate == "short link" then date = nil; -- unset msg = inlineRed (err_warn_msgs.date_miss, 'warning'); elseif == udate then date = nil; -- unset else date = udate; end end else -- some other service if not date then msg = inlineRed (err_warn_msgs.date_miss, 'warning'); end end if 'index' == date then ulx.url1.date = date .. (msg or ); -- create index + message (if there is one) elseif date then ulx.url1.date = makeDate (date, nil, nil, ldf) .. (msg or ); -- create a date in the wiki's local language + message (if there is one) else ulx.url1.date = msg; end format = args.format; -- Format argument if not format then format = "none" else for k, v in pairs (data.format_vals) do -- |format= accepts two specific values loop through a table of those values local found; -- declare a nil flag for _, p in ipairs (v) do -- loop through local language variants if format == p then -- when |format= value matches format = k; -- use name from table key found = true; -- declare found so that we can break out of outer for loop break; -- break out of inner for loop end end if found then break; end end if format == "addlpages" then if not ulx.url1.date then format = "none" end elseif format == "addlarchives" then format = "addlarchives" else format = "none" end end ulx.url1.format = format if args.title and args.title1 then -- Title argument return inlineError (data.crit_err_msgs.conflicting, {origin.title, origin.title1}); end ulx.url1.title = args.title or args.title1; local rend = createRendering() if not rend then return inlineError (data.crit_err_msgs.unknown); end return rend .. ((unnamed_params and inlineRed (err_warn_msgs.unnamed_params, 'warning')) or ) .. createTracking(); end --[[--------------------------< E X P O R T E D F U N C T I O N S >------------------------------------------ ]] return {webarchive = webarchive}; AMC's Filmsite. Retrieved August 15, 2017.
  58. ↑ 100 Essential Films. --[[ ---------------------------------- Lua module implementing the Template loop detected: Template:Webarchive template. A merger of the functionality of three templates: Template:Wayback, Template:Webcite and Template:Cite archives ]] --[[--------------------------< D E P E N D E N C I E S >------------------------------------------------------ ]] require('Module:No globals'); local getArgs = require ('Module:Arguments').getArgs; --[[--------------------------< F O R W A R D D E C L A R A T I O N S >-------------------------------------- ]] local categories = {}; -- category names local config = {}; -- global configuration settings local digits = {}; -- for i18n; table that translates local-wiki digits to western digits local err_warn_msgs = {}; -- error and warning messages local excepted_pages = {}; local month_num = {}; -- for i18n; table that translates local-wiki month names to western digits local prefixes = {}; -- service provider tail string prefixes local services = {}; -- archive service provider data from local s_text = {}; -- table of static text strings used to build final rendering local uncategorized_namespaces = {}; -- list of namespaces that we should not categorize local uncategorized_subpages = {}; -- list of subpages that should not be categorized --[[--------------------------< P A G E S C O P E I D E N T I F I E R S >---------------------------------- ]] local non_western_digits; -- boolean flag set true when data.digits.enable is true local this_page = mw.title.getCurrentTitle(); local track = {}; -- Associative array to hold tracking categories local ulx = {}; -- Associative array to hold template data --[[--------------------------< S U B S T I T U T E >---------------------------------------------------------- Populates numbered arguments in a message string using an argument table. ]] local function substitute (msg, args) return args and mw.message.newRawMessage (msg, args):plain() or msg; end --[[--------------------------< tableLength >----------------------- Given a 1-D table, return number of elements ]] local function tableLength(T) local count = 0 for _ in pairs(T) do count = count + 1 end return count end --[=[-------------------------< M A K E _ W I K I L I N K >---------------------------------------------------- Makes a wikilink; when both link and display text is provided, returns a wikilink in the form D; if only link is provided, returns a wikilink in the form L; if neither are provided or link is omitted, returns an empty string. ]=] local function make_wikilink (link, display, no_link) if nil == no_link then if link and ( ~= link) then if display and ( ~= display) then return table.concat ({'', display, ''}); else return table.concat ({'', link, ''}); end end return display or ; -- link not set so return the display text else -- no_link if display and ( ~= display) then -- if there is display text return display; -- return that else return link or ; -- return the target article name or empty string end end end --[[--------------------------< createTracking >----------------------- Return data in track[] ie. tracking categories ]] local function createTracking() if not excepted_pages[this_page.fullText] then -- namespace:title/fragment is allowed to be categorized (typically this module's / template's testcases page(s)) if uncategorized_namespaces[this_page.nsText] then return ; -- this page not to be categorized so return empty string end for _,v in ipairs (uncategorized_subpages) do -- cycle through page name patterns if this_page.text:match (v) then -- test page name against each pattern return ; -- this subpage type not to be categorized so return empty string end end end local out = {}; if tableLength(track) > 0 then for key, _ in pairs(track) do -- loop through table table.insert (out, make_wikilink (key)); -- and convert category names to links end end return table.concat (out); -- concat into one big string; empty string if table is empty end --[[--------------------------< inlineError >----------------------- Critical error. Render output completely in red. Add to tracking category. This function called as the last thing before abandoning this module ]] local function inlineError (msg, args) track[categories.error] = 1 return table.concat ({ 'Error in ', -- open the error message span config.tname, -- insert the local language template name ' template: ', substitute (msg, args), -- insert the formatted error message '.', -- close the span createTracking() -- add the category }) end --[[--------------------------< inlineRed >----------------------- Render a text fragment in red, such as a warning as part of the final output. Add tracking category. ]] local function inlineRed(msg, trackmsg) if trackmsg == "warning" then track[categories.warning] = 1; elseif trackmsg == "error" then track[categories.error] = 1; end return '' .. msg .. '' end --[[--------------------------< base62 >----------------------- Convert base-62 to base-10 Credit: https://de.wikipedia.org/wiki/Modul:Expr ]] local function base62( value ) local r = 1 -- default return value is input value is malformed if value:match ('%W') then -- value must only be in the set [0-9a-zA-Z] return; -- nil return when value contains extraneous characters end local n = #value -- number of characters in value local k = 1 local c r = 0 for i = n, 1, -1 do -- loop through all characters in value from ls digit to ms digit c = value:byte( i, i ) if c >= 48 and c <= 57 then -- character is digit 0-9 c = c - 48 elseif c >= 65 and c <= 90 then -- character is ascii a-z c = c - 55 else -- must be ascii A-Z c = c - 61 end r = r + c * k -- accumulate this base62 character's value k = k * 62 -- bump for next end -- for i return r end --[[--------------------------< D E C O D E _ D A T E >-------------------------------------------------------- Given a date string, return it in iso format along with an indicator of the date's format. Except that month names must be recognizable as legitimate month names with proper capitalization, and that the date string must match one of the recognized date formats, no error checking is done here; return nil else ]] local function decode_date (date_str) local patterns = { ['dmy'] = {'^(%d%d?) +([^%s%d]+) +(%d%d%d%d)$', 'd', 'm', 'y'}, -- %a does not recognize unicode combining characters used by some languages ['mdy'] = {'^([^%s%d]+) (%d%d?), +(%d%d%d%d)$', 'm', 'd', 'y'}, ['ymd'] = {'^(%d%d%d%d) +([^%s%d]+) (%d%d?)$', 'y', 'm', 'd'}, -- not mos compliant at en.wiki but may be acceptible at other wikis }; local t = {}; if non_western_digits then -- this wiki uses non-western digits? date_str = mw.ustring.gsub (date_str, '%d', digits); -- convert this wiki's non-western digits to western digits end if date_str:match ('^%d%d%d%d%-%d%d%-%d%d$') then -- already an iso format date, return western digits form return date_str, 'iso'; end for k, v in pairs (patterns) do local c1, c2, c3 = mw.ustring.match (date_str, patterns[k][1]); -- c1 .. c3 are captured but we don't know what they hold if c1 then -- set on match t = { -- translate unspecified captures to y, m, and d [patterns[k][2]] = c1, -- fill the table of captures with the captures [patterns[k][3]] = c2, -- take index names from src_pattern table and assign sequential captures [patterns[k][4]] = c3, }; if month_num[t.m] then -- when month not already a number t.m = month_num[t.m]; -- replace valid month name with a number else return nil, 'iso'; -- not a valid date form because month not valid end return mw.ustring.format ('%.4d-%.2d-%.2d', t.y, t.m, t.d), k; -- return date in iso format end end return nil, 'iso'; -- date could not be decoded; return nil and default iso date end --[[--------------------------< makeDate >----------------------- Given year, month, day numbers, (zero-padded or not) return a full date in df format where df may be one of: mdy, dmy, iso, ymd on entry, year, month, day are presumed to be correct for the date that they represent; all are required in this module, makeDate() is sometimes given an iso-format date in year: makeDate (2018-09-20, nil, nil, df) this works because table.concat() sees only one table member ]] local function makeDate (year, month, day, df) local format = { ['dmy'] = 'j F Y', ['mdy'] = 'F j, Y', ['ymd'] = 'Y F j', ['iso'] = 'Y-m-d', }; local date = table.concat ({year, month, day}, '-'); -- assemble year-initial numeric-format date (zero padding not required here) if non_western_digits then -- this wiki uses non-western digits? date = mw.ustring.gsub (date, '%d', digits); -- convert this wiki's non-western digits to western digits end return mw.getContentLanguage():formatDate (format[df], date); end --[[--------------------------< I S _ V A L I D _ D A T E >---------------------------------------------------- Returns true if date is after 31 December 1899 (why is 1900 the min year? shouldn't the internet's date-of-birth be min year?), not after today's date, and represents a valid date (29 February 2017 is not a valid date). Applies Gregorian leapyear rules. all arguments are required ]] local function is_valid_date (year, month, day) local days_in_month = {31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31}; local month_length; local y, m, d; local today = os.date ('*t'); -- fetch a table of current date parts if not year or == year or not month or == month or not day or == day then return false; -- something missing end y = tonumber (year); m = tonumber (month); d = tonumber (day); if 1900 > y or today.year < y or 1 > m or 12 < m then -- year and month are within bounds TODO: 1900? return false; end if (2==m) then -- if February month_length = 28; -- then 28 days unless if (0==(y%4) and (0~=(y%100) or 0==(y%400))) then -- is a leap year? month_length = 29; -- if leap year then 29 days in February end else month_length=days_in_month[m]; end if 1 > d or month_length < d then -- day is within bounds return false; end -- here when date parts represent a valid date return os.time({['year']=y, ['month']=m, ['day']=d, ['hour']=0}) <= os.time(); -- date at midnight must be less than or equal to current date/time end --[[--------------------------< decodeWebciteDate >----------------------- Given a URI-path to Webcite (eg. /67xHmVFWP) return the encoded date in df format returns date string in df format - webcite date is a unix timestamp encoded as bae62 or the string 'query' ]] local function decodeWebciteDate(path, df) local dt = {}; local decode; dt = mw.text.split(path, "/") -- valid URL formats that are not base62 -- http://www.webcitation.org/query?id=1138911916587475 -- http://www.webcitation.org/query?url=http..&date=2012-06-01+21:40:03 -- http://www.webcitation.org/1138911916587475 -- http://www.webcitation.org/cache/73e53dd1f16cf8c5da298418d2a6e452870cf50e -- http://www.webcitation.org/getfile.php?fileid=1c46e791d68e89e12d0c2532cc3cf629b8bc8c8e if dt[2]:find ('query', 1, true) or dt[2]:find ('cache', 1, true) or dt[2]:find ('getfile', 1, true) or tonumber(dt[2]) then return 'query'; end decode = base62(dt[2]); -- base62 string -> exponential number if not decode then return nil; -- nil return when dt[2] contains characters not in %w end dt = os.date('*t', string.format("%d", decode):sub(1,10)) -- exponential number -> text -> first 10 characters (a unix timestamp) -> a table of date parts decode = makeDate (dt.year, dt.month, dt.day, 'iso'); -- date comparisons are all done in iso format with western digits if non_western_digits then -- this wiki uses non-western digits? decode = mw.ustring.gsub (decode, '%d', digits); -- convert this wiki's non-western digits to western digits end return decode; end --[[--------------------------< decodeWaybackDate >----------------------- Given a URI-path to Wayback (eg. /web/20160901010101/http://example.com ) or Library of Congress Web Archives (eg. /all/20160901010101/http://example.com) or UK Government Web Archive (eg. /ukgwa/20160901010101/http://example.com or /tna/20160901010101/http://example.com) return the formatted date eg. "September 1, 2016" in df format Handle non-digits in snapshot ID such as "re_" and "-" and "*" returns two values: first value is one of these: valid date string in df format - wayback date is valid (including the text string 'index' when date is '/*/') empty string - wayback date is malformed (less than 8 digits, not a valid date) nil - wayback date is '/save/' or otherwise not a number second return value is an appropriate 'message' may or may not be formatted ]] local function decodeWaybackDate(path, df) local msg, snapdate; snapdate = path:gsub ('^/web/', ):gsub ('^/all/', ):gsub ('^/ukgwa/', ):gsub ('^/tna/', ):gsub ('^/', ); -- remove leading /web/, /all/, /ukgwa/, /tna/, or / snapdate = snapdate:match ('^[^/]+'); -- get timestamp if snapdate == "*" then -- eg. /web/*/http.., etc. return 'index'; -- return indicator that this url has an index date end snapdate = snapdate:gsub ('%a%a_%d?$', ):gsub ('%-', ); -- from date, remove any trailing "re_", dashes msg = ; if snapdate:match ('%*$') then -- a trailing '*' causes calendar display at archive .org snapdate = snapdate:gsub ('%*$', ); -- remove so not part of length calc later msg = inlineRed (err_warn_msgs.ts_cal, 'warning'); -- make a message end if not tonumber(snapdate) then return nil, 'ts_nan'; -- return nil (fatal error flag) and message selector end local dlen = snapdate:len(); if dlen < 8 then -- we need 8 digits TODO: but shouldn't this be testing for 14 digits? return , inlineRed (err_warn_msgs.ts_short, 'error'); -- return empty string and error message end local year, month, day = snapdate:match ('(%d%d%d%d)(%d%d)(%d%d)'); -- no need for snapdatelong here if not is_valid_date (year, month, day) then return , inlineRed (err_warn_msgs.ts_date, 'error'); -- return empty string and error message end snapdate = table.concat ({year, month, day}, '-'); -- date comparisons are all done in iso format if 14 == dlen then return snapdate, msg; -- return date with message if any else return snapdate, msg .. inlineRed (err_warn_msgs.ts_len, 'warning'); -- return date with warning message(s) end end --[[--------------------------< decodeArchiveisDate >----------------------- Given an Archive.is "long link" URI-path (e.g. /2016.08.28-144552/http://example.com) return the date in df format (e.g. if df = dmy, return 28 August 2016) Handles "." and "-" in snapshot date, so 2016.08.28-144552 is same as 20160828144552 returns two values: first value is one of these: valid date string in df format - archive.is date is valid (including the text string 'short link' when url is the short form) empty string - wayback date is malformed (not a number, less than 8 digits, not a valid date) nil - wayback date is '/save/' second return value is an appropriate 'message' may or may not be formatted ]] local function decodeArchiveisDate(path, df) local snapdate if path:match ('^/%w+$') then -- short form url path is '/' followed by some number of base 62 digits and nothing else return "short link" -- e.g. http://archive.is/hD1qz end snapdate = mw.text.split (path, '/')[2]:gsub('[%.%-]', ); -- get snapshot date, e.g. 2016.08.28-144552; remove periods and hyphens local dlen = string.len(snapdate) if dlen < 8 then -- we need 8 digits TODO: but shouldn't this be testing for 14 digits? return , inlineRed (err_warn_msgs.ts_short, 'error'); -- return empty string and error message end local year, month, day = snapdate:match ('(%d%d%d%d)(%d%d)(%d%d)'); -- no need for snapdatelong here if not is_valid_date (year, month, day) then return , inlineRed (err_warn_msgs.ts_date, 'error'); -- return empty string and error message end snapdate = table.concat ({year, month, day}, '-'); -- date comparisons are all done in iso format if 14 == dlen then return snapdate; -- return date else return snapdate, inlineRed (err_warn_msgs.ts_len, 'warning'); -- return date with warning message end end --[[--------------------------< serviceName >----------------------- Given a domain extracted by mw.uri.new() (eg. web.archive.org) set tail string and service ID ]] local function serviceName(host, no_link) local tracking; local index; host = host:lower():gsub ('^web%.(.+)', '%1'):gsub ('^www%.(.+)', '%1'); -- lowercase, remove web. and www. subdomains if services[host] then index = host; else for k, _ in pairs (services) do if host:find ('%f[%a]'..k:gsub ('([%.%-])', '%%%1')) then index = k; break; end end end if index then local out = {}; -- empty string in [1] so that concatenated result has leading single space ulx.url1.service = services[index][4] or 'other'; tracking = services[index][5] or categories.other; -- build tail string if false == services[index][1] then -- select prefix table.insert (out, prefixes.at); elseif true == services[index][1] then table.insert (out, prefixes.atthe); else table.insert (out, services[index][1]); end table.insert (out, make_wikilink (services[index][2], services[index][3], no_link)); -- add article wikilink if services[index][6] then -- add tail postfix if it exists table.insert (out, services[index][6]); end ulx.url1.tail = table.concat (out, ' '); -- put it all together; result has leading space character else -- here when unknown archive ulx.url1.service = 'other'; tracking = categories.unknown; ulx.url1.tail = table.concat ({, prefixes.at, host, inlineRed (err_warn_msgs.unknown_url, error)}, ' '); end track[tracking] = 1 end --[[--------------------------< parseExtraArgs >----------------------- Parse numbered arguments starting at 2, such as url2..url10, date2..date10, title2..title10 For example: Template loop detected: Template:Webarchive Three url arguments not in numeric sequence (1..4..7). Function only processes arguments numbered 2 or greater (in this case 4 and 7) It creates numeric sequenced table entries like: urlx.url2.url = <argument value for url4> urlx.url3.url = <argument value for url7> Returns the number of URL arguments found numbered 2 or greater (in this case returns "2") ]] local function parseExtraArgs(args) local i, j, argurl, argurl2, argdate, argtitle j = 2 for i = 2, config.maxurls do argurl = "url" .. i if args[argurl] then argurl2 = "url" .. j ulx[argurl2] = {} ulx[argurl2]["url"] = args[argurl] argdate = "date" .. i if args[argdate] then ulx[argurl2]["date"] = args[argdate] else ulx[argurl2]["date"] = inlineRed (err_warn_msgs.date_miss, 'warning'); end argtitle = "title" .. i if args[argtitle] then ulx[argurl2]["title"] = args[argtitle] else ulx[argurl2]["title"] = nil end j = j + 1 end end if j == 2 then return 0 else return j - 2 end end --[[--------------------------< comma >----------------------- Given a date string, return "," if it's MDY ]] local function comma(date) return (date and date:match ('%a+ +%d%d?(,) +%d%d%d%d')) or ; end --[[--------------------------< createRendering >----------------------- Return a rendering of the data in ulx[][] ]] local function createRendering() local displayfield local out = {}; local index_date, msg = ulx.url1.date:match ('(index)(.*)'); -- when ulx.url1.date extract 'index' text and message text (if there is a message) ulx.url1.date = ulx.url1.date:gsub ('index.*', 'index'); -- remove message if 'none' == ulx.url1.format then -- For Template:Wayback, Template:Webcite table.insert (out, '['); -- open extlink markup table.insert (out, ulx.url1.url); -- add url if ulx.url1.title then table.insert (out, ' ') -- the required space table.insert (out, ulx.url1.title) -- the title table.insert (out, ']'); -- close extlink markup table.insert (out, ulx.url1.tail); -- tail text if ulx.url1.date then table.insert (out, ' ('); -- open date text; TODO: why the html entity? replace with regular space? table.insert (out, 'index' == ulx.url1.date and s_text.archive or s_text.archived); -- add text table.insert (out, ' '); -- insert a space table.insert (out, ulx.url1.date); -- add date table.insert (out, ')'); -- close date text end else -- no title if index_date then -- when url date is 'index' table.insert (out, table.concat ({' ', s_text.Archive_index, ']'})); -- add the index link label table.insert (out, msg or ); -- add date mismatch message when url date is /*/ and |date= has valid date else table.insert (out, table.concat ({' ', s_text.Archived, '] '})); -- add link label for url has timestamp date (will include mismatch message if there is one) end if ulx.url1.date then if 'index' ~= ulx.url1.date then table.insert (out, ulx.url1.date); -- add date when data is not 'index' end table.insert (out, comma(ulx.url1.date)); -- add ',' if date format is mdy table.insert (out, ulx.url1.tail); -- add tail text else -- no date table.insert (out, ulx.url1.tail); -- add tail text end end if 0 < ulx.url1.extraurls then -- For multiple archive URLs local tot = ulx.url1.extraurls + 1 table.insert (out, '.') -- terminate first url table.insert (out, table.concat ({' ', s_text.addlarchives, ': '})); -- add header text for i=2, tot do -- loop through the additionals local index = table.concat ({'url', i}); -- make an index displayfield = ulx[index]['title'] and 'title' or 'date'; -- choose display text table.insert (out, '['); -- open extlink markup table.insert (out, ulx[index]['url']); -- add the url table.insert (out, ' '); -- the required space table.insert (out, ulx[index][displayfield]); -- add the label table.insert (out, ']'); -- close extlink markup table.insert (out, i==tot and '.' or ', '); -- add terminator end end return table.concat (out); -- make a big string and done else -- For Template:Cite archives if 'addlarchives' == ulx.url1.format then -- Multiple archive services table.insert (out, table.concat ({s_text.addlarchives, ': '})); -- add header text else -- Multiple pages from the same archive table.insert (out, table.concat ({s_text.addlpages, ' '})); -- add header text table.insert (out, ulx.url1.date); -- add date to header text table.insert (out, ': '); -- close header text end local tot = ulx.url1.extraurls + 1; for i=1, tot do -- loop through the additionals local index = table.concat ({'url', i}); -- make an index table.insert (out, '['); -- open extlink markup table.insert (out, ulx[index]['url']); -- add url table.insert (out, ' '); -- add required space displayfield = ulx[index]['title']; if 'addlarchives' == ulx.url1.format then if not displayfield then displayfield = ulx[index]['date'] end else -- must be addlpages if not displayfield then displayfield = table.concat ({s_text.Page, ' ', i}); end end table.insert (out, displayfield); -- add title, date, page label text table.insert (out, ']'); -- close extlink markup table.insert (out, (i==tot and '.' or ', ')); -- add terminator end return table.concat (out); -- make a big string and done end end --[[--------------------------< P A R A M E T E R _ N A M E _ X L A T E >-------------------------------------- for internaltionalization, translate local-language parameter names to their English equivalents TODO: return error message if multiple aliases of the same canonical parameter name are found? returns two tables: new_args - holds canonical form parameters and their values either from translation or because the parameter was already in canonical form origin - maps canonical-form parameter names to their untranslated (local language) form for error messaging in the local language unrecognized parameters are ignored ]] local function parameter_name_xlate (args, params, enum_params) local name; -- holds modifiable name of the parameter name during evaluation local enum; -- for enumerated parameters, holds the enumerator during evaluation local found = false; -- flag used to break out of nested for loops local new_args = {}; -- a table that holds canonical and translated parameter k/v pairs local origin = {}; -- a table that maps original (local language) parameter names to their canonical name for local language error messaging local unnamed_params; -- set true when unsupported positional parameters are detected for k, v in pairs (args) do -- loop through all of the arguments in the args table name = k; -- copy of original parameter name if 'string' == type (k) then if non_western_digits then -- true when non-western digits supported at this wiki name = mw.ustring.gsub (name, '%d', digits); -- convert this wiki's non-western digits to western digits end enum = name:match ('%d+$'); -- get parameter enumerator if it exists; nil else if not enum then -- no enumerator so looking for non-enumnerated parameters -- TODO: insert shortcut here? if params[name] then name holds the canonical parameter name; no need to search further for pname, aliases in pairs (params) do -- loop through each parameter the params table for _, alias in ipairs (aliases) do -- loop through each alias in the parameter's aliases table if name == alias then new_args[pname] = v; -- create a new entry in the new_args table origin [pname] = k; -- create an entry to make canonical parameter name to original local language parameter name found = true; -- flag so that we can break out of these nested for loops break; -- no need to search the rest of the aliases table for name so go on to the next k, v pair end end if found then -- true when we found an alias that matched name found = false; -- reset the flag break; -- go do next args k/v pair end end else -- enumerated parameters name = name:gsub ('%d$', '#'); -- replace enumeration digits with place holder for table search -- TODO: insert shortcut here? if num_params[name] then name holds the canonical parameter name; no need to search further for pname, aliases in pairs (enum_params) do -- loop through each parameter the num_params table for _, alias in ipairs (aliases) do -- loop through each alias in the parameter's aliases table if name == alias then pname = pname:gsub ('#$', enum); -- replace the '#' place holder with the actual enumerator new_args[pname] = v; -- create a new entry in the new_args table origin [pname] = k; -- create an entry to make canonical parameter name to original local language parameter name found = true; -- flag so that we can break out of these nested for loops break; -- no need to search the rest of the aliases table for name so go on to the next k, v pair end end if found then -- true when we found an alias that matched name found = false; -- reset the flag break; -- go do next args k/v pair end end end else unnamed_params = true; -- flag for unsupported positional parameters end end -- for k, v return new_args, origin, unnamed_params; end --[[--------------------------< W E B A R C H I V E >---------------------------------------------------------- template entry point ]] local function webarchive(frame) local args = getArgs (frame); local data = mw.loadData (table.concat ({ -- make a data module name; sandbox or live 'Module:Webarchive/data', frame:getTitle():find('sandbox', 1, true) and '/sandbox' or -- this instance is ./sandbox then append /sandbox })); categories = data.categories; -- fill in the forward declarations config = data.config; if data.digits.enable then digits = data.digits; -- for i18n; table of digits in the local wiki's language non_western_digits = true; -- use_non_western_digits end err_warn_msgs = data.err_warn_msgs; excepted_pages = data.excepted_pages; month_num = data.month_num; -- for i18n; table of month names in the local wiki's language prefixes = data.prefixes; services = data.services; s_text = data.s_text; uncategorized_namespaces = data.uncategorized_namespaces; uncategorized_subpages = data.uncategorized_subpages; local origin = {}; -- holds a map of English to local language parameter names used in the current template; not currently used local unnamed_params; -- boolean set to true when template call has unnamed parameters args, origin, unnamed_params = parameter_name_xlate (args, data.params, data.enum_params); -- translate parameter names in args to English local date, format, msg, udate, uri, url; local ldf = 'iso'; -- when there is no |date= parameter, render url dates in iso format if args.url and args.url1 then -- URL argument (first) return inlineError (data.crit_err_msgs.conflicting, {origin.url, origin.url1}); end url = args.url or args.url1; if not url then return inlineError (data.crit_err_msgs.empty); end -- these iabot bugs perportedly fixed; removing these causes lua script error --[[ -- at Template:Webarchive/testcases/Production; resolve that before deleting these tests if mw.ustring.find( url, "https://web.http", 1, true ) then -- track bug - TODO: IAbot bug; not known if the bug has been fixed; deferred track[categories.error] = 1; return inlineError (data.crit_err_msgs.iabot1); end if url == "https://web.archive.org/http:/" then -- track bug - TODO: IAbot bug; not known if the bug has been fixed; deferred track[categories.error] = 1; return inlineError (data.crit_err_msgs.iabot2); end ]] if not (url:lower():find ('^http') or url:find ('^//')) then return inlineError (data.crit_err_msgs.invalid_url ); end ulx.url1 = {} ulx.url1.url = url ulx.url1.extraurls = parseExtraArgs(args) local good = false; good, uri = pcall (mw.uri.new, ulx.url1.url); -- get a table of uri parts from this url; protected mode to prevent lua error when ulx.url1.url is malformed if not good or nil == uri.host then -- abandon when ulx.url1.url is malformed return inlineError (data.crit_err_msgs.invalid_url); end serviceName(uri.host, args.nolink) if args.date and args.date1 then -- Date argument return inlineError (data.crit_err_msgs.conflicting, {origin.date, origin.date1}); end date = args.date or args.date1; date = date and date:gsub (' +', ' '); -- replace multiple spaces with a single space if date and config.verifydates then if '*' == date then date = 'index'; ldf = 'iso'; -- set to default format elseif 'mdy' == date then date = nil; -- if date extracted from URL, ldf = 'mdy'; -- then |date=mdy overrides iso elseif 'dmy' == date then date = nil; -- if date extracted from URL, ldf = 'dmy'; -- then |date=dmy overrides iso elseif 'ymd' == date then date = nil; -- if date extracted from URL, ldf = 'ymd'; -- then |date=ymd overrides iso else date, ldf = decode_date (date); -- get an iso format date from date and get date's original format end end if 'wayback' == ulx.url1.service or 'locwebarchives' == ulx.url1.service or 'ukgwa' == ulx.url1.service then if date then if config.verifydates then if ldf then udate, msg = decodeWaybackDate (uri.path); -- get the url date in iso format and format of date in |date=; 'index' when wayback url date is * if not udate then -- this is the only 'fatal' error return return inlineError (data.crit_err_msgs[msg]); end if udate ~= date then -- date comparison using iso format dates date = udate; msg = table.concat ({ inlineRed (err_warn_msgs.mismatch, 'warning'), -- add warning message msg, -- add message if there is one }); end end end else -- no |date= udate, msg = decodeWaybackDate (uri.path); if not udate then -- this is the only 'fatal' error return return inlineError (data.crit_err_msgs[msg]); end if == udate then date = nil; -- unset else date = udate; end end elseif 'webcite' == ulx.url1.service then if date then if config.verifydates then if ldf then udate = decodeWebciteDate (uri.path); -- get the url date in iso format if 'query' ~= udate then -- skip if query if udate ~= date then -- date comparison using iso format dates date = udate; msg = table.concat ({ inlineRed (err_warn_msgs.mismatch, 'warning'), }); end end end end else date = decodeWebciteDate( uri.path, "iso" ) if date == "query" then date = nil; -- unset msg = inlineRed (err_warn_msgs.date_miss, 'warning'); elseif not date then -- invalid base62 string date = inlineRed (err_warn_msgs.date1, 'error'); end end elseif 'archiveis' == ulx.url1.service then if date then if config.verifydates then if ldf then udate, msg = decodeArchiveisDate (uri.path) -- get the url date in iso format if 'short link' ~= udate then -- skip if short link if udate ~= date then -- date comparison using iso format dates date = udate; msg = table.concat ({ inlineRed (err_warn_msgs.mismatch, 'warning'), -- add warning message msg, -- add message if there is one }); end end end end else -- no |date= udate, msg = decodeArchiveisDate( uri.path, "iso" ) if udate == "short link" then date = nil; -- unset msg = inlineRed (err_warn_msgs.date_miss, 'warning'); elseif == udate then date = nil; -- unset else date = udate; end end else -- some other service if not date then msg = inlineRed (err_warn_msgs.date_miss, 'warning'); end end if 'index' == date then ulx.url1.date = date .. (msg or ); -- create index + message (if there is one) elseif date then ulx.url1.date = makeDate (date, nil, nil, ldf) .. (msg or ); -- create a date in the wiki's local language + message (if there is one) else ulx.url1.date = msg; end format = args.format; -- Format argument if not format then format = "none" else for k, v in pairs (data.format_vals) do -- |format= accepts two specific values loop through a table of those values local found; -- declare a nil flag for _, p in ipairs (v) do -- loop through local language variants if format == p then -- when |format= value matches format = k; -- use name from table key found = true; -- declare found so that we can break out of outer for loop break; -- break out of inner for loop end end if found then break; end end if format == "addlpages" then if not ulx.url1.date then format = "none" end elseif format == "addlarchives" then format = "addlarchives" else format = "none" end end ulx.url1.format = format if args.title and args.title1 then -- Title argument return inlineError (data.crit_err_msgs.conflicting, {origin.title, origin.title1}); end ulx.url1.title = args.title or args.title1; local rend = createRendering() if not rend then return inlineError (data.crit_err_msgs.unknown); end return rend .. ((unnamed_params and inlineRed (err_warn_msgs.unnamed_params, 'warning')) or ) .. createTracking(); end --[[--------------------------< E X P O R T E D F U N C T I O N S >------------------------------------------ ]] return {webarchive = webarchive}; National Society of Film Critics, published by AMC's Filmsite. Retrieved August 15, 2017.
  59. ↑ 100 Maverick Movies in the Last 100 Years. --[[ ---------------------------------- Lua module implementing the Template loop detected: Template:Webarchive template. A merger of the functionality of three templates: Template:Wayback, Template:Webcite and Template:Cite archives ]] --[[--------------------------< D E P E N D E N C I E S >------------------------------------------------------ ]] require('Module:No globals'); local getArgs = require ('Module:Arguments').getArgs; --[[--------------------------< F O R W A R D D E C L A R A T I O N S >-------------------------------------- ]] local categories = {}; -- category names local config = {}; -- global configuration settings local digits = {}; -- for i18n; table that translates local-wiki digits to western digits local err_warn_msgs = {}; -- error and warning messages local excepted_pages = {}; local month_num = {}; -- for i18n; table that translates local-wiki month names to western digits local prefixes = {}; -- service provider tail string prefixes local services = {}; -- archive service provider data from local s_text = {}; -- table of static text strings used to build final rendering local uncategorized_namespaces = {}; -- list of namespaces that we should not categorize local uncategorized_subpages = {}; -- list of subpages that should not be categorized --[[--------------------------< P A G E S C O P E I D E N T I F I E R S >---------------------------------- ]] local non_western_digits; -- boolean flag set true when data.digits.enable is true local this_page = mw.title.getCurrentTitle(); local track = {}; -- Associative array to hold tracking categories local ulx = {}; -- Associative array to hold template data --[[--------------------------< S U B S T I T U T E >---------------------------------------------------------- Populates numbered arguments in a message string using an argument table. ]] local function substitute (msg, args) return args and mw.message.newRawMessage (msg, args):plain() or msg; end --[[--------------------------< tableLength >----------------------- Given a 1-D table, return number of elements ]] local function tableLength(T) local count = 0 for _ in pairs(T) do count = count + 1 end return count end --[=[-------------------------< M A K E _ W I K I L I N K >---------------------------------------------------- Makes a wikilink; when both link and display text is provided, returns a wikilink in the form D; if only link is provided, returns a wikilink in the form L; if neither are provided or link is omitted, returns an empty string. ]=] local function make_wikilink (link, display, no_link) if nil == no_link then if link and ( ~= link) then if display and ( ~= display) then return table.concat ({'', display, ''}); else return table.concat ({'', link, ''}); end end return display or ; -- link not set so return the display text else -- no_link if display and ( ~= display) then -- if there is display text return display; -- return that else return link or ; -- return the target article name or empty string end end end --[[--------------------------< createTracking >----------------------- Return data in track[] ie. tracking categories ]] local function createTracking() if not excepted_pages[this_page.fullText] then -- namespace:title/fragment is allowed to be categorized (typically this module's / template's testcases page(s)) if uncategorized_namespaces[this_page.nsText] then return ; -- this page not to be categorized so return empty string end for _,v in ipairs (uncategorized_subpages) do -- cycle through page name patterns if this_page.text:match (v) then -- test page name against each pattern return ; -- this subpage type not to be categorized so return empty string end end end local out = {}; if tableLength(track) > 0 then for key, _ in pairs(track) do -- loop through table table.insert (out, make_wikilink (key)); -- and convert category names to links end end return table.concat (out); -- concat into one big string; empty string if table is empty end --[[--------------------------< inlineError >----------------------- Critical error. Render output completely in red. Add to tracking category. This function called as the last thing before abandoning this module ]] local function inlineError (msg, args) track[categories.error] = 1 return table.concat ({ 'Error in ', -- open the error message span config.tname, -- insert the local language template name ' template: ', substitute (msg, args), -- insert the formatted error message '.', -- close the span createTracking() -- add the category }) end --[[--------------------------< inlineRed >----------------------- Render a text fragment in red, such as a warning as part of the final output. Add tracking category. ]] local function inlineRed(msg, trackmsg) if trackmsg == "warning" then track[categories.warning] = 1; elseif trackmsg == "error" then track[categories.error] = 1; end return '' .. msg .. '' end --[[--------------------------< base62 >----------------------- Convert base-62 to base-10 Credit: https://de.wikipedia.org/wiki/Modul:Expr ]] local function base62( value ) local r = 1 -- default return value is input value is malformed if value:match ('%W') then -- value must only be in the set [0-9a-zA-Z] return; -- nil return when value contains extraneous characters end local n = #value -- number of characters in value local k = 1 local c r = 0 for i = n, 1, -1 do -- loop through all characters in value from ls digit to ms digit c = value:byte( i, i ) if c >= 48 and c <= 57 then -- character is digit 0-9 c = c - 48 elseif c >= 65 and c <= 90 then -- character is ascii a-z c = c - 55 else -- must be ascii A-Z c = c - 61 end r = r + c * k -- accumulate this base62 character's value k = k * 62 -- bump for next end -- for i return r end --[[--------------------------< D E C O D E _ D A T E >-------------------------------------------------------- Given a date string, return it in iso format along with an indicator of the date's format. Except that month names must be recognizable as legitimate month names with proper capitalization, and that the date string must match one of the recognized date formats, no error checking is done here; return nil else ]] local function decode_date (date_str) local patterns = { ['dmy'] = {'^(%d%d?) +([^%s%d]+) +(%d%d%d%d)$', 'd', 'm', 'y'}, -- %a does not recognize unicode combining characters used by some languages ['mdy'] = {'^([^%s%d]+) (%d%d?), +(%d%d%d%d)$', 'm', 'd', 'y'}, ['ymd'] = {'^(%d%d%d%d) +([^%s%d]+) (%d%d?)$', 'y', 'm', 'd'}, -- not mos compliant at en.wiki but may be acceptible at other wikis }; local t = {}; if non_western_digits then -- this wiki uses non-western digits? date_str = mw.ustring.gsub (date_str, '%d', digits); -- convert this wiki's non-western digits to western digits end if date_str:match ('^%d%d%d%d%-%d%d%-%d%d$') then -- already an iso format date, return western digits form return date_str, 'iso'; end for k, v in pairs (patterns) do local c1, c2, c3 = mw.ustring.match (date_str, patterns[k][1]); -- c1 .. c3 are captured but we don't know what they hold if c1 then -- set on match t = { -- translate unspecified captures to y, m, and d [patterns[k][2]] = c1, -- fill the table of captures with the captures [patterns[k][3]] = c2, -- take index names from src_pattern table and assign sequential captures [patterns[k][4]] = c3, }; if month_num[t.m] then -- when month not already a number t.m = month_num[t.m]; -- replace valid month name with a number else return nil, 'iso'; -- not a valid date form because month not valid end return mw.ustring.format ('%.4d-%.2d-%.2d', t.y, t.m, t.d), k; -- return date in iso format end end return nil, 'iso'; -- date could not be decoded; return nil and default iso date end --[[--------------------------< makeDate >----------------------- Given year, month, day numbers, (zero-padded or not) return a full date in df format where df may be one of: mdy, dmy, iso, ymd on entry, year, month, day are presumed to be correct for the date that they represent; all are required in this module, makeDate() is sometimes given an iso-format date in year: makeDate (2018-09-20, nil, nil, df) this works because table.concat() sees only one table member ]] local function makeDate (year, month, day, df) local format = { ['dmy'] = 'j F Y', ['mdy'] = 'F j, Y', ['ymd'] = 'Y F j', ['iso'] = 'Y-m-d', }; local date = table.concat ({year, month, day}, '-'); -- assemble year-initial numeric-format date (zero padding not required here) if non_western_digits then -- this wiki uses non-western digits? date = mw.ustring.gsub (date, '%d', digits); -- convert this wiki's non-western digits to western digits end return mw.getContentLanguage():formatDate (format[df], date); end --[[--------------------------< I S _ V A L I D _ D A T E >---------------------------------------------------- Returns true if date is after 31 December 1899 (why is 1900 the min year? shouldn't the internet's date-of-birth be min year?), not after today's date, and represents a valid date (29 February 2017 is not a valid date). Applies Gregorian leapyear rules. all arguments are required ]] local function is_valid_date (year, month, day) local days_in_month = {31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31}; local month_length; local y, m, d; local today = os.date ('*t'); -- fetch a table of current date parts if not year or == year or not month or == month or not day or == day then return false; -- something missing end y = tonumber (year); m = tonumber (month); d = tonumber (day); if 1900 > y or today.year < y or 1 > m or 12 < m then -- year and month are within bounds TODO: 1900? return false; end if (2==m) then -- if February month_length = 28; -- then 28 days unless if (0==(y%4) and (0~=(y%100) or 0==(y%400))) then -- is a leap year? month_length = 29; -- if leap year then 29 days in February end else month_length=days_in_month[m]; end if 1 > d or month_length < d then -- day is within bounds return false; end -- here when date parts represent a valid date return os.time({['year']=y, ['month']=m, ['day']=d, ['hour']=0}) <= os.time(); -- date at midnight must be less than or equal to current date/time end --[[--------------------------< decodeWebciteDate >----------------------- Given a URI-path to Webcite (eg. /67xHmVFWP) return the encoded date in df format returns date string in df format - webcite date is a unix timestamp encoded as bae62 or the string 'query' ]] local function decodeWebciteDate(path, df) local dt = {}; local decode; dt = mw.text.split(path, "/") -- valid URL formats that are not base62 -- http://www.webcitation.org/query?id=1138911916587475 -- http://www.webcitation.org/query?url=http..&date=2012-06-01+21:40:03 -- http://www.webcitation.org/1138911916587475 -- http://www.webcitation.org/cache/73e53dd1f16cf8c5da298418d2a6e452870cf50e -- http://www.webcitation.org/getfile.php?fileid=1c46e791d68e89e12d0c2532cc3cf629b8bc8c8e if dt[2]:find ('query', 1, true) or dt[2]:find ('cache', 1, true) or dt[2]:find ('getfile', 1, true) or tonumber(dt[2]) then return 'query'; end decode = base62(dt[2]); -- base62 string -> exponential number if not decode then return nil; -- nil return when dt[2] contains characters not in %w end dt = os.date('*t', string.format("%d", decode):sub(1,10)) -- exponential number -> text -> first 10 characters (a unix timestamp) -> a table of date parts decode = makeDate (dt.year, dt.month, dt.day, 'iso'); -- date comparisons are all done in iso format with western digits if non_western_digits then -- this wiki uses non-western digits? decode = mw.ustring.gsub (decode, '%d', digits); -- convert this wiki's non-western digits to western digits end return decode; end --[[--------------------------< decodeWaybackDate >----------------------- Given a URI-path to Wayback (eg. /web/20160901010101/http://example.com ) or Library of Congress Web Archives (eg. /all/20160901010101/http://example.com) or UK Government Web Archive (eg. /ukgwa/20160901010101/http://example.com or /tna/20160901010101/http://example.com) return the formatted date eg. "September 1, 2016" in df format Handle non-digits in snapshot ID such as "re_" and "-" and "*" returns two values: first value is one of these: valid date string in df format - wayback date is valid (including the text string 'index' when date is '/*/') empty string - wayback date is malformed (less than 8 digits, not a valid date) nil - wayback date is '/save/' or otherwise not a number second return value is an appropriate 'message' may or may not be formatted ]] local function decodeWaybackDate(path, df) local msg, snapdate; snapdate = path:gsub ('^/web/', ):gsub ('^/all/', ):gsub ('^/ukgwa/', ):gsub ('^/tna/', ):gsub ('^/', ); -- remove leading /web/, /all/, /ukgwa/, /tna/, or / snapdate = snapdate:match ('^[^/]+'); -- get timestamp if snapdate == "*" then -- eg. /web/*/http.., etc. return 'index'; -- return indicator that this url has an index date end snapdate = snapdate:gsub ('%a%a_%d?$', ):gsub ('%-', ); -- from date, remove any trailing "re_", dashes msg = ; if snapdate:match ('%*$') then -- a trailing '*' causes calendar display at archive .org snapdate = snapdate:gsub ('%*$', ); -- remove so not part of length calc later msg = inlineRed (err_warn_msgs.ts_cal, 'warning'); -- make a message end if not tonumber(snapdate) then return nil, 'ts_nan'; -- return nil (fatal error flag) and message selector end local dlen = snapdate:len(); if dlen < 8 then -- we need 8 digits TODO: but shouldn't this be testing for 14 digits? return , inlineRed (err_warn_msgs.ts_short, 'error'); -- return empty string and error message end local year, month, day = snapdate:match ('(%d%d%d%d)(%d%d)(%d%d)'); -- no need for snapdatelong here if not is_valid_date (year, month, day) then return , inlineRed (err_warn_msgs.ts_date, 'error'); -- return empty string and error message end snapdate = table.concat ({year, month, day}, '-'); -- date comparisons are all done in iso format if 14 == dlen then return snapdate, msg; -- return date with message if any else return snapdate, msg .. inlineRed (err_warn_msgs.ts_len, 'warning'); -- return date with warning message(s) end end --[[--------------------------< decodeArchiveisDate >----------------------- Given an Archive.is "long link" URI-path (e.g. /2016.08.28-144552/http://example.com) return the date in df format (e.g. if df = dmy, return 28 August 2016) Handles "." and "-" in snapshot date, so 2016.08.28-144552 is same as 20160828144552 returns two values: first value is one of these: valid date string in df format - archive.is date is valid (including the text string 'short link' when url is the short form) empty string - wayback date is malformed (not a number, less than 8 digits, not a valid date) nil - wayback date is '/save/' second return value is an appropriate 'message' may or may not be formatted ]] local function decodeArchiveisDate(path, df) local snapdate if path:match ('^/%w+$') then -- short form url path is '/' followed by some number of base 62 digits and nothing else return "short link" -- e.g. http://archive.is/hD1qz end snapdate = mw.text.split (path, '/')[2]:gsub('[%.%-]', ); -- get snapshot date, e.g. 2016.08.28-144552; remove periods and hyphens local dlen = string.len(snapdate) if dlen < 8 then -- we need 8 digits TODO: but shouldn't this be testing for 14 digits? return , inlineRed (err_warn_msgs.ts_short, 'error'); -- return empty string and error message end local year, month, day = snapdate:match ('(%d%d%d%d)(%d%d)(%d%d)'); -- no need for snapdatelong here if not is_valid_date (year, month, day) then return , inlineRed (err_warn_msgs.ts_date, 'error'); -- return empty string and error message end snapdate = table.concat ({year, month, day}, '-'); -- date comparisons are all done in iso format if 14 == dlen then return snapdate; -- return date else return snapdate, inlineRed (err_warn_msgs.ts_len, 'warning'); -- return date with warning message end end --[[--------------------------< serviceName >----------------------- Given a domain extracted by mw.uri.new() (eg. web.archive.org) set tail string and service ID ]] local function serviceName(host, no_link) local tracking; local index; host = host:lower():gsub ('^web%.(.+)', '%1'):gsub ('^www%.(.+)', '%1'); -- lowercase, remove web. and www. subdomains if services[host] then index = host; else for k, _ in pairs (services) do if host:find ('%f[%a]'..k:gsub ('([%.%-])', '%%%1')) then index = k; break; end end end if index then local out = {}; -- empty string in [1] so that concatenated result has leading single space ulx.url1.service = services[index][4] or 'other'; tracking = services[index][5] or categories.other; -- build tail string if false == services[index][1] then -- select prefix table.insert (out, prefixes.at); elseif true == services[index][1] then table.insert (out, prefixes.atthe); else table.insert (out, services[index][1]); end table.insert (out, make_wikilink (services[index][2], services[index][3], no_link)); -- add article wikilink if services[index][6] then -- add tail postfix if it exists table.insert (out, services[index][6]); end ulx.url1.tail = table.concat (out, ' '); -- put it all together; result has leading space character else -- here when unknown archive ulx.url1.service = 'other'; tracking = categories.unknown; ulx.url1.tail = table.concat ({, prefixes.at, host, inlineRed (err_warn_msgs.unknown_url, error)}, ' '); end track[tracking] = 1 end --[[--------------------------< parseExtraArgs >----------------------- Parse numbered arguments starting at 2, such as url2..url10, date2..date10, title2..title10 For example: Template loop detected: Template:Webarchive Three url arguments not in numeric sequence (1..4..7). Function only processes arguments numbered 2 or greater (in this case 4 and 7) It creates numeric sequenced table entries like: urlx.url2.url = <argument value for url4> urlx.url3.url = <argument value for url7> Returns the number of URL arguments found numbered 2 or greater (in this case returns "2") ]] local function parseExtraArgs(args) local i, j, argurl, argurl2, argdate, argtitle j = 2 for i = 2, config.maxurls do argurl = "url" .. i if args[argurl] then argurl2 = "url" .. j ulx[argurl2] = {} ulx[argurl2]["url"] = args[argurl] argdate = "date" .. i if args[argdate] then ulx[argurl2]["date"] = args[argdate] else ulx[argurl2]["date"] = inlineRed (err_warn_msgs.date_miss, 'warning'); end argtitle = "title" .. i if args[argtitle] then ulx[argurl2]["title"] = args[argtitle] else ulx[argurl2]["title"] = nil end j = j + 1 end end if j == 2 then return 0 else return j - 2 end end --[[--------------------------< comma >----------------------- Given a date string, return "," if it's MDY ]] local function comma(date) return (date and date:match ('%a+ +%d%d?(,) +%d%d%d%d')) or ; end --[[--------------------------< createRendering >----------------------- Return a rendering of the data in ulx[][] ]] local function createRendering() local displayfield local out = {}; local index_date, msg = ulx.url1.date:match ('(index)(.*)'); -- when ulx.url1.date extract 'index' text and message text (if there is a message) ulx.url1.date = ulx.url1.date:gsub ('index.*', 'index'); -- remove message if 'none' == ulx.url1.format then -- For Template:Wayback, Template:Webcite table.insert (out, '['); -- open extlink markup table.insert (out, ulx.url1.url); -- add url if ulx.url1.title then table.insert (out, ' ') -- the required space table.insert (out, ulx.url1.title) -- the title table.insert (out, ']'); -- close extlink markup table.insert (out, ulx.url1.tail); -- tail text if ulx.url1.date then table.insert (out, ' ('); -- open date text; TODO: why the html entity? replace with regular space? table.insert (out, 'index' == ulx.url1.date and s_text.archive or s_text.archived); -- add text table.insert (out, ' '); -- insert a space table.insert (out, ulx.url1.date); -- add date table.insert (out, ')'); -- close date text end else -- no title if index_date then -- when url date is 'index' table.insert (out, table.concat ({' ', s_text.Archive_index, ']'})); -- add the index link label table.insert (out, msg or ); -- add date mismatch message when url date is /*/ and |date= has valid date else table.insert (out, table.concat ({' ', s_text.Archived, '] '})); -- add link label for url has timestamp date (will include mismatch message if there is one) end if ulx.url1.date then if 'index' ~= ulx.url1.date then table.insert (out, ulx.url1.date); -- add date when data is not 'index' end table.insert (out, comma(ulx.url1.date)); -- add ',' if date format is mdy table.insert (out, ulx.url1.tail); -- add tail text else -- no date table.insert (out, ulx.url1.tail); -- add tail text end end if 0 < ulx.url1.extraurls then -- For multiple archive URLs local tot = ulx.url1.extraurls + 1 table.insert (out, '.') -- terminate first url table.insert (out, table.concat ({' ', s_text.addlarchives, ': '})); -- add header text for i=2, tot do -- loop through the additionals local index = table.concat ({'url', i}); -- make an index displayfield = ulx[index]['title'] and 'title' or 'date'; -- choose display text table.insert (out, '['); -- open extlink markup table.insert (out, ulx[index]['url']); -- add the url table.insert (out, ' '); -- the required space table.insert (out, ulx[index][displayfield]); -- add the label table.insert (out, ']'); -- close extlink markup table.insert (out, i==tot and '.' or ', '); -- add terminator end end return table.concat (out); -- make a big string and done else -- For Template:Cite archives if 'addlarchives' == ulx.url1.format then -- Multiple archive services table.insert (out, table.concat ({s_text.addlarchives, ': '})); -- add header text else -- Multiple pages from the same archive table.insert (out, table.concat ({s_text.addlpages, ' '})); -- add header text table.insert (out, ulx.url1.date); -- add date to header text table.insert (out, ': '); -- close header text end local tot = ulx.url1.extraurls + 1; for i=1, tot do -- loop through the additionals local index = table.concat ({'url', i}); -- make an index table.insert (out, '['); -- open extlink markup table.insert (out, ulx[index]['url']); -- add url table.insert (out, ' '); -- add required space displayfield = ulx[index]['title']; if 'addlarchives' == ulx.url1.format then if not displayfield then displayfield = ulx[index]['date'] end else -- must be addlpages if not displayfield then displayfield = table.concat ({s_text.Page, ' ', i}); end end table.insert (out, displayfield); -- add title, date, page label text table.insert (out, ']'); -- close extlink markup table.insert (out, (i==tot and '.' or ', ')); -- add terminator end return table.concat (out); -- make a big string and done end end --[[--------------------------< P A R A M E T E R _ N A M E _ X L A T E >-------------------------------------- for internaltionalization, translate local-language parameter names to their English equivalents TODO: return error message if multiple aliases of the same canonical parameter name are found? returns two tables: new_args - holds canonical form parameters and their values either from translation or because the parameter was already in canonical form origin - maps canonical-form parameter names to their untranslated (local language) form for error messaging in the local language unrecognized parameters are ignored ]] local function parameter_name_xlate (args, params, enum_params) local name; -- holds modifiable name of the parameter name during evaluation local enum; -- for enumerated parameters, holds the enumerator during evaluation local found = false; -- flag used to break out of nested for loops local new_args = {}; -- a table that holds canonical and translated parameter k/v pairs local origin = {}; -- a table that maps original (local language) parameter names to their canonical name for local language error messaging local unnamed_params; -- set true when unsupported positional parameters are detected for k, v in pairs (args) do -- loop through all of the arguments in the args table name = k; -- copy of original parameter name if 'string' == type (k) then if non_western_digits then -- true when non-western digits supported at this wiki name = mw.ustring.gsub (name, '%d', digits); -- convert this wiki's non-western digits to western digits end enum = name:match ('%d+$'); -- get parameter enumerator if it exists; nil else if not enum then -- no enumerator so looking for non-enumnerated parameters -- TODO: insert shortcut here? if params[name] then name holds the canonical parameter name; no need to search further for pname, aliases in pairs (params) do -- loop through each parameter the params table for _, alias in ipairs (aliases) do -- loop through each alias in the parameter's aliases table if name == alias then new_args[pname] = v; -- create a new entry in the new_args table origin [pname] = k; -- create an entry to make canonical parameter name to original local language parameter name found = true; -- flag so that we can break out of these nested for loops break; -- no need to search the rest of the aliases table for name so go on to the next k, v pair end end if found then -- true when we found an alias that matched name found = false; -- reset the flag break; -- go do next args k/v pair end end else -- enumerated parameters name = name:gsub ('%d$', '#'); -- replace enumeration digits with place holder for table search -- TODO: insert shortcut here? if num_params[name] then name holds the canonical parameter name; no need to search further for pname, aliases in pairs (enum_params) do -- loop through each parameter the num_params table for _, alias in ipairs (aliases) do -- loop through each alias in the parameter's aliases table if name == alias then pname = pname:gsub ('#$', enum); -- replace the '#' place holder with the actual enumerator new_args[pname] = v; -- create a new entry in the new_args table origin [pname] = k; -- create an entry to make canonical parameter name to original local language parameter name found = true; -- flag so that we can break out of these nested for loops break; -- no need to search the rest of the aliases table for name so go on to the next k, v pair end end if found then -- true when we found an alias that matched name found = false; -- reset the flag break; -- go do next args k/v pair end end end else unnamed_params = true; -- flag for unsupported positional parameters end end -- for k, v return new_args, origin, unnamed_params; end --[[--------------------------< W E B A R C H I V E >---------------------------------------------------------- template entry point ]] local function webarchive(frame) local args = getArgs (frame); local data = mw.loadData (table.concat ({ -- make a data module name; sandbox or live 'Module:Webarchive/data', frame:getTitle():find('sandbox', 1, true) and '/sandbox' or -- this instance is ./sandbox then append /sandbox })); categories = data.categories; -- fill in the forward declarations config = data.config; if data.digits.enable then digits = data.digits; -- for i18n; table of digits in the local wiki's language non_western_digits = true; -- use_non_western_digits end err_warn_msgs = data.err_warn_msgs; excepted_pages = data.excepted_pages; month_num = data.month_num; -- for i18n; table of month names in the local wiki's language prefixes = data.prefixes; services = data.services; s_text = data.s_text; uncategorized_namespaces = data.uncategorized_namespaces; uncategorized_subpages = data.uncategorized_subpages; local origin = {}; -- holds a map of English to local language parameter names used in the current template; not currently used local unnamed_params; -- boolean set to true when template call has unnamed parameters args, origin, unnamed_params = parameter_name_xlate (args, data.params, data.enum_params); -- translate parameter names in args to English local date, format, msg, udate, uri, url; local ldf = 'iso'; -- when there is no |date= parameter, render url dates in iso format if args.url and args.url1 then -- URL argument (first) return inlineError (data.crit_err_msgs.conflicting, {origin.url, origin.url1}); end url = args.url or args.url1; if not url then return inlineError (data.crit_err_msgs.empty); end -- these iabot bugs perportedly fixed; removing these causes lua script error --[[ -- at Template:Webarchive/testcases/Production; resolve that before deleting these tests if mw.ustring.find( url, "https://web.http", 1, true ) then -- track bug - TODO: IAbot bug; not known if the bug has been fixed; deferred track[categories.error] = 1; return inlineError (data.crit_err_msgs.iabot1); end if url == "https://web.archive.org/http:/" then -- track bug - TODO: IAbot bug; not known if the bug has been fixed; deferred track[categories.error] = 1; return inlineError (data.crit_err_msgs.iabot2); end ]] if not (url:lower():find ('^http') or url:find ('^//')) then return inlineError (data.crit_err_msgs.invalid_url ); end ulx.url1 = {} ulx.url1.url = url ulx.url1.extraurls = parseExtraArgs(args) local good = false; good, uri = pcall (mw.uri.new, ulx.url1.url); -- get a table of uri parts from this url; protected mode to prevent lua error when ulx.url1.url is malformed if not good or nil == uri.host then -- abandon when ulx.url1.url is malformed return inlineError (data.crit_err_msgs.invalid_url); end serviceName(uri.host, args.nolink) if args.date and args.date1 then -- Date argument return inlineError (data.crit_err_msgs.conflicting, {origin.date, origin.date1}); end date = args.date or args.date1; date = date and date:gsub (' +', ' '); -- replace multiple spaces with a single space if date and config.verifydates then if '*' == date then date = 'index'; ldf = 'iso'; -- set to default format elseif 'mdy' == date then date = nil; -- if date extracted from URL, ldf = 'mdy'; -- then |date=mdy overrides iso elseif 'dmy' == date then date = nil; -- if date extracted from URL, ldf = 'dmy'; -- then |date=dmy overrides iso elseif 'ymd' == date then date = nil; -- if date extracted from URL, ldf = 'ymd'; -- then |date=ymd overrides iso else date, ldf = decode_date (date); -- get an iso format date from date and get date's original format end end if 'wayback' == ulx.url1.service or 'locwebarchives' == ulx.url1.service or 'ukgwa' == ulx.url1.service then if date then if config.verifydates then if ldf then udate, msg = decodeWaybackDate (uri.path); -- get the url date in iso format and format of date in |date=; 'index' when wayback url date is * if not udate then -- this is the only 'fatal' error return return inlineError (data.crit_err_msgs[msg]); end if udate ~= date then -- date comparison using iso format dates date = udate; msg = table.concat ({ inlineRed (err_warn_msgs.mismatch, 'warning'), -- add warning message msg, -- add message if there is one }); end end end else -- no |date= udate, msg = decodeWaybackDate (uri.path); if not udate then -- this is the only 'fatal' error return return inlineError (data.crit_err_msgs[msg]); end if == udate then date = nil; -- unset else date = udate; end end elseif 'webcite' == ulx.url1.service then if date then if config.verifydates then if ldf then udate = decodeWebciteDate (uri.path); -- get the url date in iso format if 'query' ~= udate then -- skip if query if udate ~= date then -- date comparison using iso format dates date = udate; msg = table.concat ({ inlineRed (err_warn_msgs.mismatch, 'warning'), }); end end end end else date = decodeWebciteDate( uri.path, "iso" ) if date == "query" then date = nil; -- unset msg = inlineRed (err_warn_msgs.date_miss, 'warning'); elseif not date then -- invalid base62 string date = inlineRed (err_warn_msgs.date1, 'error'); end end elseif 'archiveis' == ulx.url1.service then if date then if config.verifydates then if ldf then udate, msg = decodeArchiveisDate (uri.path) -- get the url date in iso format if 'short link' ~= udate then -- skip if short link if udate ~= date then -- date comparison using iso format dates date = udate; msg = table.concat ({ inlineRed (err_warn_msgs.mismatch, 'warning'), -- add warning message msg, -- add message if there is one }); end end end end else -- no |date= udate, msg = decodeArchiveisDate( uri.path, "iso" ) if udate == "short link" then date = nil; -- unset msg = inlineRed (err_warn_msgs.date_miss, 'warning'); elseif == udate then date = nil; -- unset else date = udate; end end else -- some other service if not date then msg = inlineRed (err_warn_msgs.date_miss, 'warning'); end end if 'index' == date then ulx.url1.date = date .. (msg or ); -- create index + message (if there is one) elseif date then ulx.url1.date = makeDate (date, nil, nil, ldf) .. (msg or ); -- create a date in the wiki's local language + message (if there is one) else ulx.url1.date = msg; end format = args.format; -- Format argument if not format then format = "none" else for k, v in pairs (data.format_vals) do -- |format= accepts two specific values loop through a table of those values local found; -- declare a nil flag for _, p in ipairs (v) do -- loop through local language variants if format == p then -- when |format= value matches format = k; -- use name from table key found = true; -- declare found so that we can break out of outer for loop break; -- break out of inner for loop end end if found then break; end end if format == "addlpages" then if not ulx.url1.date then format = "none" end elseif format == "addlarchives" then format = "addlarchives" else format = "none" end end ulx.url1.format = format if args.title and args.title1 then -- Title argument return inlineError (data.crit_err_msgs.conflicting, {origin.title, origin.title1}); end ulx.url1.title = args.title or args.title1; local rend = createRendering() if not rend then return inlineError (data.crit_err_msgs.unknown); end return rend .. ((unnamed_params and inlineRed (err_warn_msgs.unnamed_params, 'warning')) or ) .. createTracking(); end --[[--------------------------< E X P O R T E D F U N C T I O N S >------------------------------------------ ]] return {webarchive = webarchive}; Rolling Stone, published by AMC's Filmsite. Retrieved August 15, 2017.
  60. ↑ Greatest American Films. --[[ ---------------------------------- Lua module implementing the Template loop detected: Template:Webarchive template. A merger of the functionality of three templates: Template:Wayback, Template:Webcite and Template:Cite archives ]] --[[--------------------------< D E P E N D E N C I E S >------------------------------------------------------ ]] require('Module:No globals'); local getArgs = require ('Module:Arguments').getArgs; --[[--------------------------< F O R W A R D D E C L A R A T I O N S >-------------------------------------- ]] local categories = {}; -- category names local config = {}; -- global configuration settings local digits = {}; -- for i18n; table that translates local-wiki digits to western digits local err_warn_msgs = {}; -- error and warning messages local excepted_pages = {}; local month_num = {}; -- for i18n; table that translates local-wiki month names to western digits local prefixes = {}; -- service provider tail string prefixes local services = {}; -- archive service provider data from local s_text = {}; -- table of static text strings used to build final rendering local uncategorized_namespaces = {}; -- list of namespaces that we should not categorize local uncategorized_subpages = {}; -- list of subpages that should not be categorized --[[--------------------------< P A G E S C O P E I D E N T I F I E R S >---------------------------------- ]] local non_western_digits; -- boolean flag set true when data.digits.enable is true local this_page = mw.title.getCurrentTitle(); local track = {}; -- Associative array to hold tracking categories local ulx = {}; -- Associative array to hold template data --[[--------------------------< S U B S T I T U T E >---------------------------------------------------------- Populates numbered arguments in a message string using an argument table. ]] local function substitute (msg, args) return args and mw.message.newRawMessage (msg, args):plain() or msg; end --[[--------------------------< tableLength >----------------------- Given a 1-D table, return number of elements ]] local function tableLength(T) local count = 0 for _ in pairs(T) do count = count + 1 end return count end --[=[-------------------------< M A K E _ W I K I L I N K >---------------------------------------------------- Makes a wikilink; when both link and display text is provided, returns a wikilink in the form D; if only link is provided, returns a wikilink in the form L; if neither are provided or link is omitted, returns an empty string. ]=] local function make_wikilink (link, display, no_link) if nil == no_link then if link and ( ~= link) then if display and ( ~= display) then return table.concat ({'', display, ''}); else return table.concat ({'', link, ''}); end end return display or ; -- link not set so return the display text else -- no_link if display and ( ~= display) then -- if there is display text return display; -- return that else return link or ; -- return the target article name or empty string end end end --[[--------------------------< createTracking >----------------------- Return data in track[] ie. tracking categories ]] local function createTracking() if not excepted_pages[this_page.fullText] then -- namespace:title/fragment is allowed to be categorized (typically this module's / template's testcases page(s)) if uncategorized_namespaces[this_page.nsText] then return ; -- this page not to be categorized so return empty string end for _,v in ipairs (uncategorized_subpages) do -- cycle through page name patterns if this_page.text:match (v) then -- test page name against each pattern return ; -- this subpage type not to be categorized so return empty string end end end local out = {}; if tableLength(track) > 0 then for key, _ in pairs(track) do -- loop through table table.insert (out, make_wikilink (key)); -- and convert category names to links end end return table.concat (out); -- concat into one big string; empty string if table is empty end --[[--------------------------< inlineError >----------------------- Critical error. Render output completely in red. Add to tracking category. This function called as the last thing before abandoning this module ]] local function inlineError (msg, args) track[categories.error] = 1 return table.concat ({ 'Error in ', -- open the error message span config.tname, -- insert the local language template name ' template: ', substitute (msg, args), -- insert the formatted error message '.', -- close the span createTracking() -- add the category }) end --[[--------------------------< inlineRed >----------------------- Render a text fragment in red, such as a warning as part of the final output. Add tracking category. ]] local function inlineRed(msg, trackmsg) if trackmsg == "warning" then track[categories.warning] = 1; elseif trackmsg == "error" then track[categories.error] = 1; end return '' .. msg .. '' end --[[--------------------------< base62 >----------------------- Convert base-62 to base-10 Credit: https://de.wikipedia.org/wiki/Modul:Expr ]] local function base62( value ) local r = 1 -- default return value is input value is malformed if value:match ('%W') then -- value must only be in the set [0-9a-zA-Z] return; -- nil return when value contains extraneous characters end local n = #value -- number of characters in value local k = 1 local c r = 0 for i = n, 1, -1 do -- loop through all characters in value from ls digit to ms digit c = value:byte( i, i ) if c >= 48 and c <= 57 then -- character is digit 0-9 c = c - 48 elseif c >= 65 and c <= 90 then -- character is ascii a-z c = c - 55 else -- must be ascii A-Z c = c - 61 end r = r + c * k -- accumulate this base62 character's value k = k * 62 -- bump for next end -- for i return r end --[[--------------------------< D E C O D E _ D A T E >-------------------------------------------------------- Given a date string, return it in iso format along with an indicator of the date's format. Except that month names must be recognizable as legitimate month names with proper capitalization, and that the date string must match one of the recognized date formats, no error checking is done here; return nil else ]] local function decode_date (date_str) local patterns = { ['dmy'] = {'^(%d%d?) +([^%s%d]+) +(%d%d%d%d)$', 'd', 'm', 'y'}, -- %a does not recognize unicode combining characters used by some languages ['mdy'] = {'^([^%s%d]+) (%d%d?), +(%d%d%d%d)$', 'm', 'd', 'y'}, ['ymd'] = {'^(%d%d%d%d) +([^%s%d]+) (%d%d?)$', 'y', 'm', 'd'}, -- not mos compliant at en.wiki but may be acceptible at other wikis }; local t = {}; if non_western_digits then -- this wiki uses non-western digits? date_str = mw.ustring.gsub (date_str, '%d', digits); -- convert this wiki's non-western digits to western digits end if date_str:match ('^%d%d%d%d%-%d%d%-%d%d$') then -- already an iso format date, return western digits form return date_str, 'iso'; end for k, v in pairs (patterns) do local c1, c2, c3 = mw.ustring.match (date_str, patterns[k][1]); -- c1 .. c3 are captured but we don't know what they hold if c1 then -- set on match t = { -- translate unspecified captures to y, m, and d [patterns[k][2]] = c1, -- fill the table of captures with the captures [patterns[k][3]] = c2, -- take index names from src_pattern table and assign sequential captures [patterns[k][4]] = c3, }; if month_num[t.m] then -- when month not already a number t.m = month_num[t.m]; -- replace valid month name with a number else return nil, 'iso'; -- not a valid date form because month not valid end return mw.ustring.format ('%.4d-%.2d-%.2d', t.y, t.m, t.d), k; -- return date in iso format end end return nil, 'iso'; -- date could not be decoded; return nil and default iso date end --[[--------------------------< makeDate >----------------------- Given year, month, day numbers, (zero-padded or not) return a full date in df format where df may be one of: mdy, dmy, iso, ymd on entry, year, month, day are presumed to be correct for the date that they represent; all are required in this module, makeDate() is sometimes given an iso-format date in year: makeDate (2018-09-20, nil, nil, df) this works because table.concat() sees only one table member ]] local function makeDate (year, month, day, df) local format = { ['dmy'] = 'j F Y', ['mdy'] = 'F j, Y', ['ymd'] = 'Y F j', ['iso'] = 'Y-m-d', }; local date = table.concat ({year, month, day}, '-'); -- assemble year-initial numeric-format date (zero padding not required here) if non_western_digits then -- this wiki uses non-western digits? date = mw.ustring.gsub (date, '%d', digits); -- convert this wiki's non-western digits to western digits end return mw.getContentLanguage():formatDate (format[df], date); end --[[--------------------------< I S _ V A L I D _ D A T E >---------------------------------------------------- Returns true if date is after 31 December 1899 (why is 1900 the min year? shouldn't the internet's date-of-birth be min year?), not after today's date, and represents a valid date (29 February 2017 is not a valid date). Applies Gregorian leapyear rules. all arguments are required ]] local function is_valid_date (year, month, day) local days_in_month = {31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31}; local month_length; local y, m, d; local today = os.date ('*t'); -- fetch a table of current date parts if not year or == year or not month or == month or not day or == day then return false; -- something missing end y = tonumber (year); m = tonumber (month); d = tonumber (day); if 1900 > y or today.year < y or 1 > m or 12 < m then -- year and month are within bounds TODO: 1900? return false; end if (2==m) then -- if February month_length = 28; -- then 28 days unless if (0==(y%4) and (0~=(y%100) or 0==(y%400))) then -- is a leap year? month_length = 29; -- if leap year then 29 days in February end else month_length=days_in_month[m]; end if 1 > d or month_length < d then -- day is within bounds return false; end -- here when date parts represent a valid date return os.time({['year']=y, ['month']=m, ['day']=d, ['hour']=0}) <= os.time(); -- date at midnight must be less than or equal to current date/time end --[[--------------------------< decodeWebciteDate >----------------------- Given a URI-path to Webcite (eg. /67xHmVFWP) return the encoded date in df format returns date string in df format - webcite date is a unix timestamp encoded as bae62 or the string 'query' ]] local function decodeWebciteDate(path, df) local dt = {}; local decode; dt = mw.text.split(path, "/") -- valid URL formats that are not base62 -- http://www.webcitation.org/query?id=1138911916587475 -- http://www.webcitation.org/query?url=http..&date=2012-06-01+21:40:03 -- http://www.webcitation.org/1138911916587475 -- http://www.webcitation.org/cache/73e53dd1f16cf8c5da298418d2a6e452870cf50e -- http://www.webcitation.org/getfile.php?fileid=1c46e791d68e89e12d0c2532cc3cf629b8bc8c8e if dt[2]:find ('query', 1, true) or dt[2]:find ('cache', 1, true) or dt[2]:find ('getfile', 1, true) or tonumber(dt[2]) then return 'query'; end decode = base62(dt[2]); -- base62 string -> exponential number if not decode then return nil; -- nil return when dt[2] contains characters not in %w end dt = os.date('*t', string.format("%d", decode):sub(1,10)) -- exponential number -> text -> first 10 characters (a unix timestamp) -> a table of date parts decode = makeDate (dt.year, dt.month, dt.day, 'iso'); -- date comparisons are all done in iso format with western digits if non_western_digits then -- this wiki uses non-western digits? decode = mw.ustring.gsub (decode, '%d', digits); -- convert this wiki's non-western digits to western digits end return decode; end --[[--------------------------< decodeWaybackDate >----------------------- Given a URI-path to Wayback (eg. /web/20160901010101/http://example.com ) or Library of Congress Web Archives (eg. /all/20160901010101/http://example.com) or UK Government Web Archive (eg. /ukgwa/20160901010101/http://example.com or /tna/20160901010101/http://example.com) return the formatted date eg. "September 1, 2016" in df format Handle non-digits in snapshot ID such as "re_" and "-" and "*" returns two values: first value is one of these: valid date string in df format - wayback date is valid (including the text string 'index' when date is '/*/') empty string - wayback date is malformed (less than 8 digits, not a valid date) nil - wayback date is '/save/' or otherwise not a number second return value is an appropriate 'message' may or may not be formatted ]] local function decodeWaybackDate(path, df) local msg, snapdate; snapdate = path:gsub ('^/web/', ):gsub ('^/all/', ):gsub ('^/ukgwa/', ):gsub ('^/tna/', ):gsub ('^/', ); -- remove leading /web/, /all/, /ukgwa/, /tna/, or / snapdate = snapdate:match ('^[^/]+'); -- get timestamp if snapdate == "*" then -- eg. /web/*/http.., etc. return 'index'; -- return indicator that this url has an index date end snapdate = snapdate:gsub ('%a%a_%d?$', ):gsub ('%-', ); -- from date, remove any trailing "re_", dashes msg = ; if snapdate:match ('%*$') then -- a trailing '*' causes calendar display at archive .org snapdate = snapdate:gsub ('%*$', ); -- remove so not part of length calc later msg = inlineRed (err_warn_msgs.ts_cal, 'warning'); -- make a message end if not tonumber(snapdate) then return nil, 'ts_nan'; -- return nil (fatal error flag) and message selector end local dlen = snapdate:len(); if dlen < 8 then -- we need 8 digits TODO: but shouldn't this be testing for 14 digits? return , inlineRed (err_warn_msgs.ts_short, 'error'); -- return empty string and error message end local year, month, day = snapdate:match ('(%d%d%d%d)(%d%d)(%d%d)'); -- no need for snapdatelong here if not is_valid_date (year, month, day) then return , inlineRed (err_warn_msgs.ts_date, 'error'); -- return empty string and error message end snapdate = table.concat ({year, month, day}, '-'); -- date comparisons are all done in iso format if 14 == dlen then return snapdate, msg; -- return date with message if any else return snapdate, msg .. inlineRed (err_warn_msgs.ts_len, 'warning'); -- return date with warning message(s) end end --[[--------------------------< decodeArchiveisDate >----------------------- Given an Archive.is "long link" URI-path (e.g. /2016.08.28-144552/http://example.com) return the date in df format (e.g. if df = dmy, return 28 August 2016) Handles "." and "-" in snapshot date, so 2016.08.28-144552 is same as 20160828144552 returns two values: first value is one of these: valid date string in df format - archive.is date is valid (including the text string 'short link' when url is the short form) empty string - wayback date is malformed (not a number, less than 8 digits, not a valid date) nil - wayback date is '/save/' second return value is an appropriate 'message' may or may not be formatted ]] local function decodeArchiveisDate(path, df) local snapdate if path:match ('^/%w+$') then -- short form url path is '/' followed by some number of base 62 digits and nothing else return "short link" -- e.g. http://archive.is/hD1qz end snapdate = mw.text.split (path, '/')[2]:gsub('[%.%-]', ); -- get snapshot date, e.g. 2016.08.28-144552; remove periods and hyphens local dlen = string.len(snapdate) if dlen < 8 then -- we need 8 digits TODO: but shouldn't this be testing for 14 digits? return , inlineRed (err_warn_msgs.ts_short, 'error'); -- return empty string and error message end local year, month, day = snapdate:match ('(%d%d%d%d)(%d%d)(%d%d)'); -- no need for snapdatelong here if not is_valid_date (year, month, day) then return , inlineRed (err_warn_msgs.ts_date, 'error'); -- return empty string and error message end snapdate = table.concat ({year, month, day}, '-'); -- date comparisons are all done in iso format if 14 == dlen then return snapdate; -- return date else return snapdate, inlineRed (err_warn_msgs.ts_len, 'warning'); -- return date with warning message end end --[[--------------------------< serviceName >----------------------- Given a domain extracted by mw.uri.new() (eg. web.archive.org) set tail string and service ID ]] local function serviceName(host, no_link) local tracking; local index; host = host:lower():gsub ('^web%.(.+)', '%1'):gsub ('^www%.(.+)', '%1'); -- lowercase, remove web. and www. subdomains if services[host] then index = host; else for k, _ in pairs (services) do if host:find ('%f[%a]'..k:gsub ('([%.%-])', '%%%1')) then index = k; break; end end end if index then local out = {}; -- empty string in [1] so that concatenated result has leading single space ulx.url1.service = services[index][4] or 'other'; tracking = services[index][5] or categories.other; -- build tail string if false == services[index][1] then -- select prefix table.insert (out, prefixes.at); elseif true == services[index][1] then table.insert (out, prefixes.atthe); else table.insert (out, services[index][1]); end table.insert (out, make_wikilink (services[index][2], services[index][3], no_link)); -- add article wikilink if services[index][6] then -- add tail postfix if it exists table.insert (out, services[index][6]); end ulx.url1.tail = table.concat (out, ' '); -- put it all together; result has leading space character else -- here when unknown archive ulx.url1.service = 'other'; tracking = categories.unknown; ulx.url1.tail = table.concat ({, prefixes.at, host, inlineRed (err_warn_msgs.unknown_url, error)}, ' '); end track[tracking] = 1 end --[[--------------------------< parseExtraArgs >----------------------- Parse numbered arguments starting at 2, such as url2..url10, date2..date10, title2..title10 For example: Template loop detected: Template:Webarchive Three url arguments not in numeric sequence (1..4..7). Function only processes arguments numbered 2 or greater (in this case 4 and 7) It creates numeric sequenced table entries like: urlx.url2.url = <argument value for url4> urlx.url3.url = <argument value for url7> Returns the number of URL arguments found numbered 2 or greater (in this case returns "2") ]] local function parseExtraArgs(args) local i, j, argurl, argurl2, argdate, argtitle j = 2 for i = 2, config.maxurls do argurl = "url" .. i if args[argurl] then argurl2 = "url" .. j ulx[argurl2] = {} ulx[argurl2]["url"] = args[argurl] argdate = "date" .. i if args[argdate] then ulx[argurl2]["date"] = args[argdate] else ulx[argurl2]["date"] = inlineRed (err_warn_msgs.date_miss, 'warning'); end argtitle = "title" .. i if args[argtitle] then ulx[argurl2]["title"] = args[argtitle] else ulx[argurl2]["title"] = nil end j = j + 1 end end if j == 2 then return 0 else return j - 2 end end --[[--------------------------< comma >----------------------- Given a date string, return "," if it's MDY ]] local function comma(date) return (date and date:match ('%a+ +%d%d?(,) +%d%d%d%d')) or ; end --[[--------------------------< createRendering >----------------------- Return a rendering of the data in ulx[][] ]] local function createRendering() local displayfield local out = {}; local index_date, msg = ulx.url1.date:match ('(index)(.*)'); -- when ulx.url1.date extract 'index' text and message text (if there is a message) ulx.url1.date = ulx.url1.date:gsub ('index.*', 'index'); -- remove message if 'none' == ulx.url1.format then -- For Template:Wayback, Template:Webcite table.insert (out, '['); -- open extlink markup table.insert (out, ulx.url1.url); -- add url if ulx.url1.title then table.insert (out, ' ') -- the required space table.insert (out, ulx.url1.title) -- the title table.insert (out, ']'); -- close extlink markup table.insert (out, ulx.url1.tail); -- tail text if ulx.url1.date then table.insert (out, ' ('); -- open date text; TODO: why the html entity? replace with regular space? table.insert (out, 'index' == ulx.url1.date and s_text.archive or s_text.archived); -- add text table.insert (out, ' '); -- insert a space table.insert (out, ulx.url1.date); -- add date table.insert (out, ')'); -- close date text end else -- no title if index_date then -- when url date is 'index' table.insert (out, table.concat ({' ', s_text.Archive_index, ']'})); -- add the index link label table.insert (out, msg or ); -- add date mismatch message when url date is /*/ and |date= has valid date else table.insert (out, table.concat ({' ', s_text.Archived, '] '})); -- add link label for url has timestamp date (will include mismatch message if there is one) end if ulx.url1.date then if 'index' ~= ulx.url1.date then table.insert (out, ulx.url1.date); -- add date when data is not 'index' end table.insert (out, comma(ulx.url1.date)); -- add ',' if date format is mdy table.insert (out, ulx.url1.tail); -- add tail text else -- no date table.insert (out, ulx.url1.tail); -- add tail text end end if 0 < ulx.url1.extraurls then -- For multiple archive URLs local tot = ulx.url1.extraurls + 1 table.insert (out, '.') -- terminate first url table.insert (out, table.concat ({' ', s_text.addlarchives, ': '})); -- add header text for i=2, tot do -- loop through the additionals local index = table.concat ({'url', i}); -- make an index displayfield = ulx[index]['title'] and 'title' or 'date'; -- choose display text table.insert (out, '['); -- open extlink markup table.insert (out, ulx[index]['url']); -- add the url table.insert (out, ' '); -- the required space table.insert (out, ulx[index][displayfield]); -- add the label table.insert (out, ']'); -- close extlink markup table.insert (out, i==tot and '.' or ', '); -- add terminator end end return table.concat (out); -- make a big string and done else -- For Template:Cite archives if 'addlarchives' == ulx.url1.format then -- Multiple archive services table.insert (out, table.concat ({s_text.addlarchives, ': '})); -- add header text else -- Multiple pages from the same archive table.insert (out, table.concat ({s_text.addlpages, ' '})); -- add header text table.insert (out, ulx.url1.date); -- add date to header text table.insert (out, ': '); -- close header text end local tot = ulx.url1.extraurls + 1; for i=1, tot do -- loop through the additionals local index = table.concat ({'url', i}); -- make an index table.insert (out, '['); -- open extlink markup table.insert (out, ulx[index]['url']); -- add url table.insert (out, ' '); -- add required space displayfield = ulx[index]['title']; if 'addlarchives' == ulx.url1.format then if not displayfield then displayfield = ulx[index]['date'] end else -- must be addlpages if not displayfield then displayfield = table.concat ({s_text.Page, ' ', i}); end end table.insert (out, displayfield); -- add title, date, page label text table.insert (out, ']'); -- close extlink markup table.insert (out, (i==tot and '.' or ', ')); -- add terminator end return table.concat (out); -- make a big string and done end end --[[--------------------------< P A R A M E T E R _ N A M E _ X L A T E >-------------------------------------- for internaltionalization, translate local-language parameter names to their English equivalents TODO: return error message if multiple aliases of the same canonical parameter name are found? returns two tables: new_args - holds canonical form parameters and their values either from translation or because the parameter was already in canonical form origin - maps canonical-form parameter names to their untranslated (local language) form for error messaging in the local language unrecognized parameters are ignored ]] local function parameter_name_xlate (args, params, enum_params) local name; -- holds modifiable name of the parameter name during evaluation local enum; -- for enumerated parameters, holds the enumerator during evaluation local found = false; -- flag used to break out of nested for loops local new_args = {}; -- a table that holds canonical and translated parameter k/v pairs local origin = {}; -- a table that maps original (local language) parameter names to their canonical name for local language error messaging local unnamed_params; -- set true when unsupported positional parameters are detected for k, v in pairs (args) do -- loop through all of the arguments in the args table name = k; -- copy of original parameter name if 'string' == type (k) then if non_western_digits then -- true when non-western digits supported at this wiki name = mw.ustring.gsub (name, '%d', digits); -- convert this wiki's non-western digits to western digits end enum = name:match ('%d+$'); -- get parameter enumerator if it exists; nil else if not enum then -- no enumerator so looking for non-enumnerated parameters -- TODO: insert shortcut here? if params[name] then name holds the canonical parameter name; no need to search further for pname, aliases in pairs (params) do -- loop through each parameter the params table for _, alias in ipairs (aliases) do -- loop through each alias in the parameter's aliases table if name == alias then new_args[pname] = v; -- create a new entry in the new_args table origin [pname] = k; -- create an entry to make canonical parameter name to original local language parameter name found = true; -- flag so that we can break out of these nested for loops break; -- no need to search the rest of the aliases table for name so go on to the next k, v pair end end if found then -- true when we found an alias that matched name found = false; -- reset the flag break; -- go do next args k/v pair end end else -- enumerated parameters name = name:gsub ('%d$', '#'); -- replace enumeration digits with place holder for table search -- TODO: insert shortcut here? if num_params[name] then name holds the canonical parameter name; no need to search further for pname, aliases in pairs (enum_params) do -- loop through each parameter the num_params table for _, alias in ipairs (aliases) do -- loop through each alias in the parameter's aliases table if name == alias then pname = pname:gsub ('#$', enum); -- replace the '#' place holder with the actual enumerator new_args[pname] = v; -- create a new entry in the new_args table origin [pname] = k; -- create an entry to make canonical parameter name to original local language parameter name found = true; -- flag so that we can break out of these nested for loops break; -- no need to search the rest of the aliases table for name so go on to the next k, v pair end end if found then -- true when we found an alias that matched name found = false; -- reset the flag break; -- go do next args k/v pair end end end else unnamed_params = true; -- flag for unsupported positional parameters end end -- for k, v return new_args, origin, unnamed_params; end --[[--------------------------< W E B A R C H I V E >---------------------------------------------------------- template entry point ]] local function webarchive(frame) local args = getArgs (frame); local data = mw.loadData (table.concat ({ -- make a data module name; sandbox or live 'Module:Webarchive/data', frame:getTitle():find('sandbox', 1, true) and '/sandbox' or -- this instance is ./sandbox then append /sandbox })); categories = data.categories; -- fill in the forward declarations config = data.config; if data.digits.enable then digits = data.digits; -- for i18n; table of digits in the local wiki's language non_western_digits = true; -- use_non_western_digits end err_warn_msgs = data.err_warn_msgs; excepted_pages = data.excepted_pages; month_num = data.month_num; -- for i18n; table of month names in the local wiki's language prefixes = data.prefixes; services = data.services; s_text = data.s_text; uncategorized_namespaces = data.uncategorized_namespaces; uncategorized_subpages = data.uncategorized_subpages; local origin = {}; -- holds a map of English to local language parameter names used in the current template; not currently used local unnamed_params; -- boolean set to true when template call has unnamed parameters args, origin, unnamed_params = parameter_name_xlate (args, data.params, data.enum_params); -- translate parameter names in args to English local date, format, msg, udate, uri, url; local ldf = 'iso'; -- when there is no |date= parameter, render url dates in iso format if args.url and args.url1 then -- URL argument (first) return inlineError (data.crit_err_msgs.conflicting, {origin.url, origin.url1}); end url = args.url or args.url1; if not url then return inlineError (data.crit_err_msgs.empty); end -- these iabot bugs perportedly fixed; removing these causes lua script error --[[ -- at Template:Webarchive/testcases/Production; resolve that before deleting these tests if mw.ustring.find( url, "https://web.http", 1, true ) then -- track bug - TODO: IAbot bug; not known if the bug has been fixed; deferred track[categories.error] = 1; return inlineError (data.crit_err_msgs.iabot1); end if url == "https://web.archive.org/http:/" then -- track bug - TODO: IAbot bug; not known if the bug has been fixed; deferred track[categories.error] = 1; return inlineError (data.crit_err_msgs.iabot2); end ]] if not (url:lower():find ('^http') or url:find ('^//')) then return inlineError (data.crit_err_msgs.invalid_url ); end ulx.url1 = {} ulx.url1.url = url ulx.url1.extraurls = parseExtraArgs(args) local good = false; good, uri = pcall (mw.uri.new, ulx.url1.url); -- get a table of uri parts from this url; protected mode to prevent lua error when ulx.url1.url is malformed if not good or nil == uri.host then -- abandon when ulx.url1.url is malformed return inlineError (data.crit_err_msgs.invalid_url); end serviceName(uri.host, args.nolink) if args.date and args.date1 then -- Date argument return inlineError (data.crit_err_msgs.conflicting, {origin.date, origin.date1}); end date = args.date or args.date1; date = date and date:gsub (' +', ' '); -- replace multiple spaces with a single space if date and config.verifydates then if '*' == date then date = 'index'; ldf = 'iso'; -- set to default format elseif 'mdy' == date then date = nil; -- if date extracted from URL, ldf = 'mdy'; -- then |date=mdy overrides iso elseif 'dmy' == date then date = nil; -- if date extracted from URL, ldf = 'dmy'; -- then |date=dmy overrides iso elseif 'ymd' == date then date = nil; -- if date extracted from URL, ldf = 'ymd'; -- then |date=ymd overrides iso else date, ldf = decode_date (date); -- get an iso format date from date and get date's original format end end if 'wayback' == ulx.url1.service or 'locwebarchives' == ulx.url1.service or 'ukgwa' == ulx.url1.service then if date then if config.verifydates then if ldf then udate, msg = decodeWaybackDate (uri.path); -- get the url date in iso format and format of date in |date=; 'index' when wayback url date is * if not udate then -- this is the only 'fatal' error return return inlineError (data.crit_err_msgs[msg]); end if udate ~= date then -- date comparison using iso format dates date = udate; msg = table.concat ({ inlineRed (err_warn_msgs.mismatch, 'warning'), -- add warning message msg, -- add message if there is one }); end end end else -- no |date= udate, msg = decodeWaybackDate (uri.path); if not udate then -- this is the only 'fatal' error return return inlineError (data.crit_err_msgs[msg]); end if == udate then date = nil; -- unset else date = udate; end end elseif 'webcite' == ulx.url1.service then if date then if config.verifydates then if ldf then udate = decodeWebciteDate (uri.path); -- get the url date in iso format if 'query' ~= udate then -- skip if query if udate ~= date then -- date comparison using iso format dates date = udate; msg = table.concat ({ inlineRed (err_warn_msgs.mismatch, 'warning'), }); end end end end else date = decodeWebciteDate( uri.path, "iso" ) if date == "query" then date = nil; -- unset msg = inlineRed (err_warn_msgs.date_miss, 'warning'); elseif not date then -- invalid base62 string date = inlineRed (err_warn_msgs.date1, 'error'); end end elseif 'archiveis' == ulx.url1.service then if date then if config.verifydates then if ldf then udate, msg = decodeArchiveisDate (uri.path) -- get the url date in iso format if 'short link' ~= udate then -- skip if short link if udate ~= date then -- date comparison using iso format dates date = udate; msg = table.concat ({ inlineRed (err_warn_msgs.mismatch, 'warning'), -- add warning message msg, -- add message if there is one }); end end end end else -- no |date= udate, msg = decodeArchiveisDate( uri.path, "iso" ) if udate == "short link" then date = nil; -- unset msg = inlineRed (err_warn_msgs.date_miss, 'warning'); elseif == udate then date = nil; -- unset else date = udate; end end else -- some other service if not date then msg = inlineRed (err_warn_msgs.date_miss, 'warning'); end end if 'index' == date then ulx.url1.date = date .. (msg or ); -- create index + message (if there is one) elseif date then ulx.url1.date = makeDate (date, nil, nil, ldf) .. (msg or ); -- create a date in the wiki's local language + message (if there is one) else ulx.url1.date = msg; end format = args.format; -- Format argument if not format then format = "none" else for k, v in pairs (data.format_vals) do -- |format= accepts two specific values loop through a table of those values local found; -- declare a nil flag for _, p in ipairs (v) do -- loop through local language variants if format == p then -- when |format= value matches format = k; -- use name from table key found = true; -- declare found so that we can break out of outer for loop break; -- break out of inner for loop end end if found then break; end end if format == "addlpages" then if not ulx.url1.date then format = "none" end elseif format == "addlarchives" then format = "addlarchives" else format = "none" end end ulx.url1.format = format if args.title and args.title1 then -- Title argument return inlineError (data.crit_err_msgs.conflicting, {origin.title, origin.title1}); end ulx.url1.title = args.title or args.title1; local rend = createRendering() if not rend then return inlineError (data.crit_err_msgs.unknown); end return rend .. ((unnamed_params and inlineRed (err_warn_msgs.unnamed_params, 'warning')) or ) .. createTracking(); end --[[--------------------------< E X P O R T E D F U N C T I O N S >------------------------------------------ ]] return {webarchive = webarchive}; L.A. Daily News, published by AMC's Filmsite. Published December 3, 1997. Retrieved August 15, 2017.
  61. ↑ Writers Guild of America, West. The 101 Best Screenplays. Archived from the original on August 13, 2006. Retrieved on August 24, 2006.
  62. ↑ Top 100 Films. --[[ ---------------------------------- Lua module implementing the Template loop detected: Template:Webarchive template. A merger of the functionality of three templates: Template:Wayback, Template:Webcite and Template:Cite archives ]] --[[--------------------------< D E P E N D E N C I E S >------------------------------------------------------ ]] require('Module:No globals'); local getArgs = require ('Module:Arguments').getArgs; --[[--------------------------< F O R W A R D D E C L A R A T I O N S >-------------------------------------- ]] local categories = {}; -- category names local config = {}; -- global configuration settings local digits = {}; -- for i18n; table that translates local-wiki digits to western digits local err_warn_msgs = {}; -- error and warning messages local excepted_pages = {}; local month_num = {}; -- for i18n; table that translates local-wiki month names to western digits local prefixes = {}; -- service provider tail string prefixes local services = {}; -- archive service provider data from local s_text = {}; -- table of static text strings used to build final rendering local uncategorized_namespaces = {}; -- list of namespaces that we should not categorize local uncategorized_subpages = {}; -- list of subpages that should not be categorized --[[--------------------------< P A G E S C O P E I D E N T I F I E R S >---------------------------------- ]] local non_western_digits; -- boolean flag set true when data.digits.enable is true local this_page = mw.title.getCurrentTitle(); local track = {}; -- Associative array to hold tracking categories local ulx = {}; -- Associative array to hold template data --[[--------------------------< S U B S T I T U T E >---------------------------------------------------------- Populates numbered arguments in a message string using an argument table. ]] local function substitute (msg, args) return args and mw.message.newRawMessage (msg, args):plain() or msg; end --[[--------------------------< tableLength >----------------------- Given a 1-D table, return number of elements ]] local function tableLength(T) local count = 0 for _ in pairs(T) do count = count + 1 end return count end --[=[-------------------------< M A K E _ W I K I L I N K >---------------------------------------------------- Makes a wikilink; when both link and display text is provided, returns a wikilink in the form D; if only link is provided, returns a wikilink in the form L; if neither are provided or link is omitted, returns an empty string. ]=] local function make_wikilink (link, display, no_link) if nil == no_link then if link and ( ~= link) then if display and ( ~= display) then return table.concat ({'', display, ''}); else return table.concat ({'', link, ''}); end end return display or ; -- link not set so return the display text else -- no_link if display and ( ~= display) then -- if there is display text return display; -- return that else return link or ; -- return the target article name or empty string end end end --[[--------------------------< createTracking >----------------------- Return data in track[] ie. tracking categories ]] local function createTracking() if not excepted_pages[this_page.fullText] then -- namespace:title/fragment is allowed to be categorized (typically this module's / template's testcases page(s)) if uncategorized_namespaces[this_page.nsText] then return ; -- this page not to be categorized so return empty string end for _,v in ipairs (uncategorized_subpages) do -- cycle through page name patterns if this_page.text:match (v) then -- test page name against each pattern return ; -- this subpage type not to be categorized so return empty string end end end local out = {}; if tableLength(track) > 0 then for key, _ in pairs(track) do -- loop through table table.insert (out, make_wikilink (key)); -- and convert category names to links end end return table.concat (out); -- concat into one big string; empty string if table is empty end --[[--------------------------< inlineError >----------------------- Critical error. Render output completely in red. Add to tracking category. This function called as the last thing before abandoning this module ]] local function inlineError (msg, args) track[categories.error] = 1 return table.concat ({ 'Error in ', -- open the error message span config.tname, -- insert the local language template name ' template: ', substitute (msg, args), -- insert the formatted error message '.', -- close the span createTracking() -- add the category }) end --[[--------------------------< inlineRed >----------------------- Render a text fragment in red, such as a warning as part of the final output. Add tracking category. ]] local function inlineRed(msg, trackmsg) if trackmsg == "warning" then track[categories.warning] = 1; elseif trackmsg == "error" then track[categories.error] = 1; end return '' .. msg .. '' end --[[--------------------------< base62 >----------------------- Convert base-62 to base-10 Credit: https://de.wikipedia.org/wiki/Modul:Expr ]] local function base62( value ) local r = 1 -- default return value is input value is malformed if value:match ('%W') then -- value must only be in the set [0-9a-zA-Z] return; -- nil return when value contains extraneous characters end local n = #value -- number of characters in value local k = 1 local c r = 0 for i = n, 1, -1 do -- loop through all characters in value from ls digit to ms digit c = value:byte( i, i ) if c >= 48 and c <= 57 then -- character is digit 0-9 c = c - 48 elseif c >= 65 and c <= 90 then -- character is ascii a-z c = c - 55 else -- must be ascii A-Z c = c - 61 end r = r + c * k -- accumulate this base62 character's value k = k * 62 -- bump for next end -- for i return r end --[[--------------------------< D E C O D E _ D A T E >-------------------------------------------------------- Given a date string, return it in iso format along with an indicator of the date's format. Except that month names must be recognizable as legitimate month names with proper capitalization, and that the date string must match one of the recognized date formats, no error checking is done here; return nil else ]] local function decode_date (date_str) local patterns = { ['dmy'] = {'^(%d%d?) +([^%s%d]+) +(%d%d%d%d)$', 'd', 'm', 'y'}, -- %a does not recognize unicode combining characters used by some languages ['mdy'] = {'^([^%s%d]+) (%d%d?), +(%d%d%d%d)$', 'm', 'd', 'y'}, ['ymd'] = {'^(%d%d%d%d) +([^%s%d]+) (%d%d?)$', 'y', 'm', 'd'}, -- not mos compliant at en.wiki but may be acceptible at other wikis }; local t = {}; if non_western_digits then -- this wiki uses non-western digits? date_str = mw.ustring.gsub (date_str, '%d', digits); -- convert this wiki's non-western digits to western digits end if date_str:match ('^%d%d%d%d%-%d%d%-%d%d$') then -- already an iso format date, return western digits form return date_str, 'iso'; end for k, v in pairs (patterns) do local c1, c2, c3 = mw.ustring.match (date_str, patterns[k][1]); -- c1 .. c3 are captured but we don't know what they hold if c1 then -- set on match t = { -- translate unspecified captures to y, m, and d [patterns[k][2]] = c1, -- fill the table of captures with the captures [patterns[k][3]] = c2, -- take index names from src_pattern table and assign sequential captures [patterns[k][4]] = c3, }; if month_num[t.m] then -- when month not already a number t.m = month_num[t.m]; -- replace valid month name with a number else return nil, 'iso'; -- not a valid date form because month not valid end return mw.ustring.format ('%.4d-%.2d-%.2d', t.y, t.m, t.d), k; -- return date in iso format end end return nil, 'iso'; -- date could not be decoded; return nil and default iso date end --[[--------------------------< makeDate >----------------------- Given year, month, day numbers, (zero-padded or not) return a full date in df format where df may be one of: mdy, dmy, iso, ymd on entry, year, month, day are presumed to be correct for the date that they represent; all are required in this module, makeDate() is sometimes given an iso-format date in year: makeDate (2018-09-20, nil, nil, df) this works because table.concat() sees only one table member ]] local function makeDate (year, month, day, df) local format = { ['dmy'] = 'j F Y', ['mdy'] = 'F j, Y', ['ymd'] = 'Y F j', ['iso'] = 'Y-m-d', }; local date = table.concat ({year, month, day}, '-'); -- assemble year-initial numeric-format date (zero padding not required here) if non_western_digits then -- this wiki uses non-western digits? date = mw.ustring.gsub (date, '%d', digits); -- convert this wiki's non-western digits to western digits end return mw.getContentLanguage():formatDate (format[df], date); end --[[--------------------------< I S _ V A L I D _ D A T E >---------------------------------------------------- Returns true if date is after 31 December 1899 (why is 1900 the min year? shouldn't the internet's date-of-birth be min year?), not after today's date, and represents a valid date (29 February 2017 is not a valid date). Applies Gregorian leapyear rules. all arguments are required ]] local function is_valid_date (year, month, day) local days_in_month = {31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31}; local month_length; local y, m, d; local today = os.date ('*t'); -- fetch a table of current date parts if not year or == year or not month or == month or not day or == day then return false; -- something missing end y = tonumber (year); m = tonumber (month); d = tonumber (day); if 1900 > y or today.year < y or 1 > m or 12 < m then -- year and month are within bounds TODO: 1900? return false; end if (2==m) then -- if February month_length = 28; -- then 28 days unless if (0==(y%4) and (0~=(y%100) or 0==(y%400))) then -- is a leap year? month_length = 29; -- if leap year then 29 days in February end else month_length=days_in_month[m]; end if 1 > d or month_length < d then -- day is within bounds return false; end -- here when date parts represent a valid date return os.time({['year']=y, ['month']=m, ['day']=d, ['hour']=0}) <= os.time(); -- date at midnight must be less than or equal to current date/time end --[[--------------------------< decodeWebciteDate >----------------------- Given a URI-path to Webcite (eg. /67xHmVFWP) return the encoded date in df format returns date string in df format - webcite date is a unix timestamp encoded as bae62 or the string 'query' ]] local function decodeWebciteDate(path, df) local dt = {}; local decode; dt = mw.text.split(path, "/") -- valid URL formats that are not base62 -- http://www.webcitation.org/query?id=1138911916587475 -- http://www.webcitation.org/query?url=http..&date=2012-06-01+21:40:03 -- http://www.webcitation.org/1138911916587475 -- http://www.webcitation.org/cache/73e53dd1f16cf8c5da298418d2a6e452870cf50e -- http://www.webcitation.org/getfile.php?fileid=1c46e791d68e89e12d0c2532cc3cf629b8bc8c8e if dt[2]:find ('query', 1, true) or dt[2]:find ('cache', 1, true) or dt[2]:find ('getfile', 1, true) or tonumber(dt[2]) then return 'query'; end decode = base62(dt[2]); -- base62 string -> exponential number if not decode then return nil; -- nil return when dt[2] contains characters not in %w end dt = os.date('*t', string.format("%d", decode):sub(1,10)) -- exponential number -> text -> first 10 characters (a unix timestamp) -> a table of date parts decode = makeDate (dt.year, dt.month, dt.day, 'iso'); -- date comparisons are all done in iso format with western digits if non_western_digits then -- this wiki uses non-western digits? decode = mw.ustring.gsub (decode, '%d', digits); -- convert this wiki's non-western digits to western digits end return decode; end --[[--------------------------< decodeWaybackDate >----------------------- Given a URI-path to Wayback (eg. /web/20160901010101/http://example.com ) or Library of Congress Web Archives (eg. /all/20160901010101/http://example.com) or UK Government Web Archive (eg. /ukgwa/20160901010101/http://example.com or /tna/20160901010101/http://example.com) return the formatted date eg. "September 1, 2016" in df format Handle non-digits in snapshot ID such as "re_" and "-" and "*" returns two values: first value is one of these: valid date string in df format - wayback date is valid (including the text string 'index' when date is '/*/') empty string - wayback date is malformed (less than 8 digits, not a valid date) nil - wayback date is '/save/' or otherwise not a number second return value is an appropriate 'message' may or may not be formatted ]] local function decodeWaybackDate(path, df) local msg, snapdate; snapdate = path:gsub ('^/web/', ):gsub ('^/all/', ):gsub ('^/ukgwa/', ):gsub ('^/tna/', ):gsub ('^/', ); -- remove leading /web/, /all/, /ukgwa/, /tna/, or / snapdate = snapdate:match ('^[^/]+'); -- get timestamp if snapdate == "*" then -- eg. /web/*/http.., etc. return 'index'; -- return indicator that this url has an index date end snapdate = snapdate:gsub ('%a%a_%d?$', ):gsub ('%-', ); -- from date, remove any trailing "re_", dashes msg = ; if snapdate:match ('%*$') then -- a trailing '*' causes calendar display at archive .org snapdate = snapdate:gsub ('%*$', ); -- remove so not part of length calc later msg = inlineRed (err_warn_msgs.ts_cal, 'warning'); -- make a message end if not tonumber(snapdate) then return nil, 'ts_nan'; -- return nil (fatal error flag) and message selector end local dlen = snapdate:len(); if dlen < 8 then -- we need 8 digits TODO: but shouldn't this be testing for 14 digits? return , inlineRed (err_warn_msgs.ts_short, 'error'); -- return empty string and error message end local year, month, day = snapdate:match ('(%d%d%d%d)(%d%d)(%d%d)'); -- no need for snapdatelong here if not is_valid_date (year, month, day) then return , inlineRed (err_warn_msgs.ts_date, 'error'); -- return empty string and error message end snapdate = table.concat ({year, month, day}, '-'); -- date comparisons are all done in iso format if 14 == dlen then return snapdate, msg; -- return date with message if any else return snapdate, msg .. inlineRed (err_warn_msgs.ts_len, 'warning'); -- return date with warning message(s) end end --[[--------------------------< decodeArchiveisDate >----------------------- Given an Archive.is "long link" URI-path (e.g. /2016.08.28-144552/http://example.com) return the date in df format (e.g. if df = dmy, return 28 August 2016) Handles "." and "-" in snapshot date, so 2016.08.28-144552 is same as 20160828144552 returns two values: first value is one of these: valid date string in df format - archive.is date is valid (including the text string 'short link' when url is the short form) empty string - wayback date is malformed (not a number, less than 8 digits, not a valid date) nil - wayback date is '/save/' second return value is an appropriate 'message' may or may not be formatted ]] local function decodeArchiveisDate(path, df) local snapdate if path:match ('^/%w+$') then -- short form url path is '/' followed by some number of base 62 digits and nothing else return "short link" -- e.g. http://archive.is/hD1qz end snapdate = mw.text.split (path, '/')[2]:gsub('[%.%-]', ); -- get snapshot date, e.g. 2016.08.28-144552; remove periods and hyphens local dlen = string.len(snapdate) if dlen < 8 then -- we need 8 digits TODO: but shouldn't this be testing for 14 digits? return , inlineRed (err_warn_msgs.ts_short, 'error'); -- return empty string and error message end local year, month, day = snapdate:match ('(%d%d%d%d)(%d%d)(%d%d)'); -- no need for snapdatelong here if not is_valid_date (year, month, day) then return , inlineRed (err_warn_msgs.ts_date, 'error'); -- return empty string and error message end snapdate = table.concat ({year, month, day}, '-'); -- date comparisons are all done in iso format if 14 == dlen then return snapdate; -- return date else return snapdate, inlineRed (err_warn_msgs.ts_len, 'warning'); -- return date with warning message end end --[[--------------------------< serviceName >----------------------- Given a domain extracted by mw.uri.new() (eg. web.archive.org) set tail string and service ID ]] local function serviceName(host, no_link) local tracking; local index; host = host:lower():gsub ('^web%.(.+)', '%1'):gsub ('^www%.(.+)', '%1'); -- lowercase, remove web. and www. subdomains if services[host] then index = host; else for k, _ in pairs (services) do if host:find ('%f[%a]'..k:gsub ('([%.%-])', '%%%1')) then index = k; break; end end end if index then local out = {}; -- empty string in [1] so that concatenated result has leading single space ulx.url1.service = services[index][4] or 'other'; tracking = services[index][5] or categories.other; -- build tail string if false == services[index][1] then -- select prefix table.insert (out, prefixes.at); elseif true == services[index][1] then table.insert (out, prefixes.atthe); else table.insert (out, services[index][1]); end table.insert (out, make_wikilink (services[index][2], services[index][3], no_link)); -- add article wikilink if services[index][6] then -- add tail postfix if it exists table.insert (out, services[index][6]); end ulx.url1.tail = table.concat (out, ' '); -- put it all together; result has leading space character else -- here when unknown archive ulx.url1.service = 'other'; tracking = categories.unknown; ulx.url1.tail = table.concat ({, prefixes.at, host, inlineRed (err_warn_msgs.unknown_url, error)}, ' '); end track[tracking] = 1 end --[[--------------------------< parseExtraArgs >----------------------- Parse numbered arguments starting at 2, such as url2..url10, date2..date10, title2..title10 For example: Template loop detected: Template:Webarchive Three url arguments not in numeric sequence (1..4..7). Function only processes arguments numbered 2 or greater (in this case 4 and 7) It creates numeric sequenced table entries like: urlx.url2.url = <argument value for url4> urlx.url3.url = <argument value for url7> Returns the number of URL arguments found numbered 2 or greater (in this case returns "2") ]] local function parseExtraArgs(args) local i, j, argurl, argurl2, argdate, argtitle j = 2 for i = 2, config.maxurls do argurl = "url" .. i if args[argurl] then argurl2 = "url" .. j ulx[argurl2] = {} ulx[argurl2]["url"] = args[argurl] argdate = "date" .. i if args[argdate] then ulx[argurl2]["date"] = args[argdate] else ulx[argurl2]["date"] = inlineRed (err_warn_msgs.date_miss, 'warning'); end argtitle = "title" .. i if args[argtitle] then ulx[argurl2]["title"] = args[argtitle] else ulx[argurl2]["title"] = nil end j = j + 1 end end if j == 2 then return 0 else return j - 2 end end --[[--------------------------< comma >----------------------- Given a date string, return "," if it's MDY ]] local function comma(date) return (date and date:match ('%a+ +%d%d?(,) +%d%d%d%d')) or ; end --[[--------------------------< createRendering >----------------------- Return a rendering of the data in ulx[][] ]] local function createRendering() local displayfield local out = {}; local index_date, msg = ulx.url1.date:match ('(index)(.*)'); -- when ulx.url1.date extract 'index' text and message text (if there is a message) ulx.url1.date = ulx.url1.date:gsub ('index.*', 'index'); -- remove message if 'none' == ulx.url1.format then -- For Template:Wayback, Template:Webcite table.insert (out, '['); -- open extlink markup table.insert (out, ulx.url1.url); -- add url if ulx.url1.title then table.insert (out, ' ') -- the required space table.insert (out, ulx.url1.title) -- the title table.insert (out, ']'); -- close extlink markup table.insert (out, ulx.url1.tail); -- tail text if ulx.url1.date then table.insert (out, ' ('); -- open date text; TODO: why the html entity? replace with regular space? table.insert (out, 'index' == ulx.url1.date and s_text.archive or s_text.archived); -- add text table.insert (out, ' '); -- insert a space table.insert (out, ulx.url1.date); -- add date table.insert (out, ')'); -- close date text end else -- no title if index_date then -- when url date is 'index' table.insert (out, table.concat ({' ', s_text.Archive_index, ']'})); -- add the index link label table.insert (out, msg or ); -- add date mismatch message when url date is /*/ and |date= has valid date else table.insert (out, table.concat ({' ', s_text.Archived, '] '})); -- add link label for url has timestamp date (will include mismatch message if there is one) end if ulx.url1.date then if 'index' ~= ulx.url1.date then table.insert (out, ulx.url1.date); -- add date when data is not 'index' end table.insert (out, comma(ulx.url1.date)); -- add ',' if date format is mdy table.insert (out, ulx.url1.tail); -- add tail text else -- no date table.insert (out, ulx.url1.tail); -- add tail text end end if 0 < ulx.url1.extraurls then -- For multiple archive URLs local tot = ulx.url1.extraurls + 1 table.insert (out, '.') -- terminate first url table.insert (out, table.concat ({' ', s_text.addlarchives, ': '})); -- add header text for i=2, tot do -- loop through the additionals local index = table.concat ({'url', i}); -- make an index displayfield = ulx[index]['title'] and 'title' or 'date'; -- choose display text table.insert (out, '['); -- open extlink markup table.insert (out, ulx[index]['url']); -- add the url table.insert (out, ' '); -- the required space table.insert (out, ulx[index][displayfield]); -- add the label table.insert (out, ']'); -- close extlink markup table.insert (out, i==tot and '.' or ', '); -- add terminator end end return table.concat (out); -- make a big string and done else -- For Template:Cite archives if 'addlarchives' == ulx.url1.format then -- Multiple archive services table.insert (out, table.concat ({s_text.addlarchives, ': '})); -- add header text else -- Multiple pages from the same archive table.insert (out, table.concat ({s_text.addlpages, ' '})); -- add header text table.insert (out, ulx.url1.date); -- add date to header text table.insert (out, ': '); -- close header text end local tot = ulx.url1.extraurls + 1; for i=1, tot do -- loop through the additionals local index = table.concat ({'url', i}); -- make an index table.insert (out, '['); -- open extlink markup table.insert (out, ulx[index]['url']); -- add url table.insert (out, ' '); -- add required space displayfield = ulx[index]['title']; if 'addlarchives' == ulx.url1.format then if not displayfield then displayfield = ulx[index]['date'] end else -- must be addlpages if not displayfield then displayfield = table.concat ({s_text.Page, ' ', i}); end end table.insert (out, displayfield); -- add title, date, page label text table.insert (out, ']'); -- close extlink markup table.insert (out, (i==tot and '.' or ', ')); -- add terminator end return table.concat (out); -- make a big string and done end end --[[--------------------------< P A R A M E T E R _ N A M E _ X L A T E >-------------------------------------- for internaltionalization, translate local-language parameter names to their English equivalents TODO: return error message if multiple aliases of the same canonical parameter name are found? returns two tables: new_args - holds canonical form parameters and their values either from translation or because the parameter was already in canonical form origin - maps canonical-form parameter names to their untranslated (local language) form for error messaging in the local language unrecognized parameters are ignored ]] local function parameter_name_xlate (args, params, enum_params) local name; -- holds modifiable name of the parameter name during evaluation local enum; -- for enumerated parameters, holds the enumerator during evaluation local found = false; -- flag used to break out of nested for loops local new_args = {}; -- a table that holds canonical and translated parameter k/v pairs local origin = {}; -- a table that maps original (local language) parameter names to their canonical name for local language error messaging local unnamed_params; -- set true when unsupported positional parameters are detected for k, v in pairs (args) do -- loop through all of the arguments in the args table name = k; -- copy of original parameter name if 'string' == type (k) then if non_western_digits then -- true when non-western digits supported at this wiki name = mw.ustring.gsub (name, '%d', digits); -- convert this wiki's non-western digits to western digits end enum = name:match ('%d+$'); -- get parameter enumerator if it exists; nil else if not enum then -- no enumerator so looking for non-enumnerated parameters -- TODO: insert shortcut here? if params[name] then name holds the canonical parameter name; no need to search further for pname, aliases in pairs (params) do -- loop through each parameter the params table for _, alias in ipairs (aliases) do -- loop through each alias in the parameter's aliases table if name == alias then new_args[pname] = v; -- create a new entry in the new_args table origin [pname] = k; -- create an entry to make canonical parameter name to original local language parameter name found = true; -- flag so that we can break out of these nested for loops break; -- no need to search the rest of the aliases table for name so go on to the next k, v pair end end if found then -- true when we found an alias that matched name found = false; -- reset the flag break; -- go do next args k/v pair end end else -- enumerated parameters name = name:gsub ('%d$', '#'); -- replace enumeration digits with place holder for table search -- TODO: insert shortcut here? if num_params[name] then name holds the canonical parameter name; no need to search further for pname, aliases in pairs (enum_params) do -- loop through each parameter the num_params table for _, alias in ipairs (aliases) do -- loop through each alias in the parameter's aliases table if name == alias then pname = pname:gsub ('#$', enum); -- replace the '#' place holder with the actual enumerator new_args[pname] = v; -- create a new entry in the new_args table origin [pname] = k; -- create an entry to make canonical parameter name to original local language parameter name found = true; -- flag so that we can break out of these nested for loops break; -- no need to search the rest of the aliases table for name so go on to the next k, v pair end end if found then -- true when we found an alias that matched name found = false; -- reset the flag break; -- go do next args k/v pair end end end else unnamed_params = true; -- flag for unsupported positional parameters end end -- for k, v return new_args, origin, unnamed_params; end --[[--------------------------< W E B A R C H I V E >---------------------------------------------------------- template entry point ]] local function webarchive(frame) local args = getArgs (frame); local data = mw.loadData (table.concat ({ -- make a data module name; sandbox or live 'Module:Webarchive/data', frame:getTitle():find('sandbox', 1, true) and '/sandbox' or -- this instance is ./sandbox then append /sandbox })); categories = data.categories; -- fill in the forward declarations config = data.config; if data.digits.enable then digits = data.digits; -- for i18n; table of digits in the local wiki's language non_western_digits = true; -- use_non_western_digits end err_warn_msgs = data.err_warn_msgs; excepted_pages = data.excepted_pages; month_num = data.month_num; -- for i18n; table of month names in the local wiki's language prefixes = data.prefixes; services = data.services; s_text = data.s_text; uncategorized_namespaces = data.uncategorized_namespaces; uncategorized_subpages = data.uncategorized_subpages; local origin = {}; -- holds a map of English to local language parameter names used in the current template; not currently used local unnamed_params; -- boolean set to true when template call has unnamed parameters args, origin, unnamed_params = parameter_name_xlate (args, data.params, data.enum_params); -- translate parameter names in args to English local date, format, msg, udate, uri, url; local ldf = 'iso'; -- when there is no |date= parameter, render url dates in iso format if args.url and args.url1 then -- URL argument (first) return inlineError (data.crit_err_msgs.conflicting, {origin.url, origin.url1}); end url = args.url or args.url1; if not url then return inlineError (data.crit_err_msgs.empty); end -- these iabot bugs perportedly fixed; removing these causes lua script error --[[ -- at Template:Webarchive/testcases/Production; resolve that before deleting these tests if mw.ustring.find( url, "https://web.http", 1, true ) then -- track bug - TODO: IAbot bug; not known if the bug has been fixed; deferred track[categories.error] = 1; return inlineError (data.crit_err_msgs.iabot1); end if url == "https://web.archive.org/http:/" then -- track bug - TODO: IAbot bug; not known if the bug has been fixed; deferred track[categories.error] = 1; return inlineError (data.crit_err_msgs.iabot2); end ]] if not (url:lower():find ('^http') or url:find ('^//')) then return inlineError (data.crit_err_msgs.invalid_url ); end ulx.url1 = {} ulx.url1.url = url ulx.url1.extraurls = parseExtraArgs(args) local good = false; good, uri = pcall (mw.uri.new, ulx.url1.url); -- get a table of uri parts from this url; protected mode to prevent lua error when ulx.url1.url is malformed if not good or nil == uri.host then -- abandon when ulx.url1.url is malformed return inlineError (data.crit_err_msgs.invalid_url); end serviceName(uri.host, args.nolink) if args.date and args.date1 then -- Date argument return inlineError (data.crit_err_msgs.conflicting, {origin.date, origin.date1}); end date = args.date or args.date1; date = date and date:gsub (' +', ' '); -- replace multiple spaces with a single space if date and config.verifydates then if '*' == date then date = 'index'; ldf = 'iso'; -- set to default format elseif 'mdy' == date then date = nil; -- if date extracted from URL, ldf = 'mdy'; -- then |date=mdy overrides iso elseif 'dmy' == date then date = nil; -- if date extracted from URL, ldf = 'dmy'; -- then |date=dmy overrides iso elseif 'ymd' == date then date = nil; -- if date extracted from URL, ldf = 'ymd'; -- then |date=ymd overrides iso else date, ldf = decode_date (date); -- get an iso format date from date and get date's original format end end if 'wayback' == ulx.url1.service or 'locwebarchives' == ulx.url1.service or 'ukgwa' == ulx.url1.service then if date then if config.verifydates then if ldf then udate, msg = decodeWaybackDate (uri.path); -- get the url date in iso format and format of date in |date=; 'index' when wayback url date is * if not udate then -- this is the only 'fatal' error return return inlineError (data.crit_err_msgs[msg]); end if udate ~= date then -- date comparison using iso format dates date = udate; msg = table.concat ({ inlineRed (err_warn_msgs.mismatch, 'warning'), -- add warning message msg, -- add message if there is one }); end end end else -- no |date= udate, msg = decodeWaybackDate (uri.path); if not udate then -- this is the only 'fatal' error return return inlineError (data.crit_err_msgs[msg]); end if == udate then date = nil; -- unset else date = udate; end end elseif 'webcite' == ulx.url1.service then if date then if config.verifydates then if ldf then udate = decodeWebciteDate (uri.path); -- get the url date in iso format if 'query' ~= udate then -- skip if query if udate ~= date then -- date comparison using iso format dates date = udate; msg = table.concat ({ inlineRed (err_warn_msgs.mismatch, 'warning'), }); end end end end else date = decodeWebciteDate( uri.path, "iso" ) if date == "query" then date = nil; -- unset msg = inlineRed (err_warn_msgs.date_miss, 'warning'); elseif not date then -- invalid base62 string date = inlineRed (err_warn_msgs.date1, 'error'); end end elseif 'archiveis' == ulx.url1.service then if date then if config.verifydates then if ldf then udate, msg = decodeArchiveisDate (uri.path) -- get the url date in iso format if 'short link' ~= udate then -- skip if short link if udate ~= date then -- date comparison using iso format dates date = udate; msg = table.concat ({ inlineRed (err_warn_msgs.mismatch, 'warning'), -- add warning message msg, -- add message if there is one }); end end end end else -- no |date= udate, msg = decodeArchiveisDate( uri.path, "iso" ) if udate == "short link" then date = nil; -- unset msg = inlineRed (err_warn_msgs.date_miss, 'warning'); elseif == udate then date = nil; -- unset else date = udate; end end else -- some other service if not date then msg = inlineRed (err_warn_msgs.date_miss, 'warning'); end end if 'index' == date then ulx.url1.date = date .. (msg or ); -- create index + message (if there is one) elseif date then ulx.url1.date = makeDate (date, nil, nil, ldf) .. (msg or ); -- create a date in the wiki's local language + message (if there is one) else ulx.url1.date = msg; end format = args.format; -- Format argument if not format then format = "none" else for k, v in pairs (data.format_vals) do -- |format= accepts two specific values loop through a table of those values local found; -- declare a nil flag for _, p in ipairs (v) do -- loop through local language variants if format == p then -- when |format= value matches format = k; -- use name from table key found = true; -- declare found so that we can break out of outer for loop break; -- break out of inner for loop end end if found then break; end end if format == "addlpages" then if not ulx.url1.date then format = "none" end elseif format == "addlarchives" then format = "addlarchives" else format = "none" end end ulx.url1.format = format if args.title and args.title1 then -- Title argument return inlineError (data.crit_err_msgs.conflicting, {origin.title, origin.title1}); end ulx.url1.title = args.title or args.title1; local rend = createRendering() if not rend then return inlineError (data.crit_err_msgs.unknown); end return rend .. ((unnamed_params and inlineRed (err_warn_msgs.unnamed_params, 'warning')) or ) .. createTracking(); end --[[--------------------------< E X P O R T E D F U N C T I O N S >------------------------------------------ ]] return {webarchive = webarchive}; Time Out, published by AMC's Filmsite. Retrieved August 15, 2017.
  63. ↑ Top 100 Films. --[[ ---------------------------------- Lua module implementing the Template loop detected: Template:Webarchive template. A merger of the functionality of three templates: Template:Wayback, Template:Webcite and Template:Cite archives ]] --[[--------------------------< D E P E N D E N C I E S >------------------------------------------------------ ]] require('Module:No globals'); local getArgs = require ('Module:Arguments').getArgs; --[[--------------------------< F O R W A R D D E C L A R A T I O N S >-------------------------------------- ]] local categories = {}; -- category names local config = {}; -- global configuration settings local digits = {}; -- for i18n; table that translates local-wiki digits to western digits local err_warn_msgs = {}; -- error and warning messages local excepted_pages = {}; local month_num = {}; -- for i18n; table that translates local-wiki month names to western digits local prefixes = {}; -- service provider tail string prefixes local services = {}; -- archive service provider data from local s_text = {}; -- table of static text strings used to build final rendering local uncategorized_namespaces = {}; -- list of namespaces that we should not categorize local uncategorized_subpages = {}; -- list of subpages that should not be categorized --[[--------------------------< P A G E S C O P E I D E N T I F I E R S >---------------------------------- ]] local non_western_digits; -- boolean flag set true when data.digits.enable is true local this_page = mw.title.getCurrentTitle(); local track = {}; -- Associative array to hold tracking categories local ulx = {}; -- Associative array to hold template data --[[--------------------------< S U B S T I T U T E >---------------------------------------------------------- Populates numbered arguments in a message string using an argument table. ]] local function substitute (msg, args) return args and mw.message.newRawMessage (msg, args):plain() or msg; end --[[--------------------------< tableLength >----------------------- Given a 1-D table, return number of elements ]] local function tableLength(T) local count = 0 for _ in pairs(T) do count = count + 1 end return count end --[=[-------------------------< M A K E _ W I K I L I N K >---------------------------------------------------- Makes a wikilink; when both link and display text is provided, returns a wikilink in the form D; if only link is provided, returns a wikilink in the form L; if neither are provided or link is omitted, returns an empty string. ]=] local function make_wikilink (link, display, no_link) if nil == no_link then if link and ( ~= link) then if display and ( ~= display) then return table.concat ({'', display, ''}); else return table.concat ({'', link, ''}); end end return display or ; -- link not set so return the display text else -- no_link if display and ( ~= display) then -- if there is display text return display; -- return that else return link or ; -- return the target article name or empty string end end end --[[--------------------------< createTracking >----------------------- Return data in track[] ie. tracking categories ]] local function createTracking() if not excepted_pages[this_page.fullText] then -- namespace:title/fragment is allowed to be categorized (typically this module's / template's testcases page(s)) if uncategorized_namespaces[this_page.nsText] then return ; -- this page not to be categorized so return empty string end for _,v in ipairs (uncategorized_subpages) do -- cycle through page name patterns if this_page.text:match (v) then -- test page name against each pattern return ; -- this subpage type not to be categorized so return empty string end end end local out = {}; if tableLength(track) > 0 then for key, _ in pairs(track) do -- loop through table table.insert (out, make_wikilink (key)); -- and convert category names to links end end return table.concat (out); -- concat into one big string; empty string if table is empty end --[[--------------------------< inlineError >----------------------- Critical error. Render output completely in red. Add to tracking category. This function called as the last thing before abandoning this module ]] local function inlineError (msg, args) track[categories.error] = 1 return table.concat ({ 'Error in ', -- open the error message span config.tname, -- insert the local language template name ' template: ', substitute (msg, args), -- insert the formatted error message '.', -- close the span createTracking() -- add the category }) end --[[--------------------------< inlineRed >----------------------- Render a text fragment in red, such as a warning as part of the final output. Add tracking category. ]] local function inlineRed(msg, trackmsg) if trackmsg == "warning" then track[categories.warning] = 1; elseif trackmsg == "error" then track[categories.error] = 1; end return '' .. msg .. '' end --[[--------------------------< base62 >----------------------- Convert base-62 to base-10 Credit: https://de.wikipedia.org/wiki/Modul:Expr ]] local function base62( value ) local r = 1 -- default return value is input value is malformed if value:match ('%W') then -- value must only be in the set [0-9a-zA-Z] return; -- nil return when value contains extraneous characters end local n = #value -- number of characters in value local k = 1 local c r = 0 for i = n, 1, -1 do -- loop through all characters in value from ls digit to ms digit c = value:byte( i, i ) if c >= 48 and c <= 57 then -- character is digit 0-9 c = c - 48 elseif c >= 65 and c <= 90 then -- character is ascii a-z c = c - 55 else -- must be ascii A-Z c = c - 61 end r = r + c * k -- accumulate this base62 character's value k = k * 62 -- bump for next end -- for i return r end --[[--------------------------< D E C O D E _ D A T E >-------------------------------------------------------- Given a date string, return it in iso format along with an indicator of the date's format. Except that month names must be recognizable as legitimate month names with proper capitalization, and that the date string must match one of the recognized date formats, no error checking is done here; return nil else ]] local function decode_date (date_str) local patterns = { ['dmy'] = {'^(%d%d?) +([^%s%d]+) +(%d%d%d%d)$', 'd', 'm', 'y'}, -- %a does not recognize unicode combining characters used by some languages ['mdy'] = {'^([^%s%d]+) (%d%d?), +(%d%d%d%d)$', 'm', 'd', 'y'}, ['ymd'] = {'^(%d%d%d%d) +([^%s%d]+) (%d%d?)$', 'y', 'm', 'd'}, -- not mos compliant at en.wiki but may be acceptible at other wikis }; local t = {}; if non_western_digits then -- this wiki uses non-western digits? date_str = mw.ustring.gsub (date_str, '%d', digits); -- convert this wiki's non-western digits to western digits end if date_str:match ('^%d%d%d%d%-%d%d%-%d%d$') then -- already an iso format date, return western digits form return date_str, 'iso'; end for k, v in pairs (patterns) do local c1, c2, c3 = mw.ustring.match (date_str, patterns[k][1]); -- c1 .. c3 are captured but we don't know what they hold if c1 then -- set on match t = { -- translate unspecified captures to y, m, and d [patterns[k][2]] = c1, -- fill the table of captures with the captures [patterns[k][3]] = c2, -- take index names from src_pattern table and assign sequential captures [patterns[k][4]] = c3, }; if month_num[t.m] then -- when month not already a number t.m = month_num[t.m]; -- replace valid month name with a number else return nil, 'iso'; -- not a valid date form because month not valid end return mw.ustring.format ('%.4d-%.2d-%.2d', t.y, t.m, t.d), k; -- return date in iso format end end return nil, 'iso'; -- date could not be decoded; return nil and default iso date end --[[--------------------------< makeDate >----------------------- Given year, month, day numbers, (zero-padded or not) return a full date in df format where df may be one of: mdy, dmy, iso, ymd on entry, year, month, day are presumed to be correct for the date that they represent; all are required in this module, makeDate() is sometimes given an iso-format date in year: makeDate (2018-09-20, nil, nil, df) this works because table.concat() sees only one table member ]] local function makeDate (year, month, day, df) local format = { ['dmy'] = 'j F Y', ['mdy'] = 'F j, Y', ['ymd'] = 'Y F j', ['iso'] = 'Y-m-d', }; local date = table.concat ({year, month, day}, '-'); -- assemble year-initial numeric-format date (zero padding not required here) if non_western_digits then -- this wiki uses non-western digits? date = mw.ustring.gsub (date, '%d', digits); -- convert this wiki's non-western digits to western digits end return mw.getContentLanguage():formatDate (format[df], date); end --[[--------------------------< I S _ V A L I D _ D A T E >---------------------------------------------------- Returns true if date is after 31 December 1899 (why is 1900 the min year? shouldn't the internet's date-of-birth be min year?), not after today's date, and represents a valid date (29 February 2017 is not a valid date). Applies Gregorian leapyear rules. all arguments are required ]] local function is_valid_date (year, month, day) local days_in_month = {31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31}; local month_length; local y, m, d; local today = os.date ('*t'); -- fetch a table of current date parts if not year or == year or not month or == month or not day or == day then return false; -- something missing end y = tonumber (year); m = tonumber (month); d = tonumber (day); if 1900 > y or today.year < y or 1 > m or 12 < m then -- year and month are within bounds TODO: 1900? return false; end if (2==m) then -- if February month_length = 28; -- then 28 days unless if (0==(y%4) and (0~=(y%100) or 0==(y%400))) then -- is a leap year? month_length = 29; -- if leap year then 29 days in February end else month_length=days_in_month[m]; end if 1 > d or month_length < d then -- day is within bounds return false; end -- here when date parts represent a valid date return os.time({['year']=y, ['month']=m, ['day']=d, ['hour']=0}) <= os.time(); -- date at midnight must be less than or equal to current date/time end --[[--------------------------< decodeWebciteDate >----------------------- Given a URI-path to Webcite (eg. /67xHmVFWP) return the encoded date in df format returns date string in df format - webcite date is a unix timestamp encoded as bae62 or the string 'query' ]] local function decodeWebciteDate(path, df) local dt = {}; local decode; dt = mw.text.split(path, "/") -- valid URL formats that are not base62 -- http://www.webcitation.org/query?id=1138911916587475 -- http://www.webcitation.org/query?url=http..&date=2012-06-01+21:40:03 -- http://www.webcitation.org/1138911916587475 -- http://www.webcitation.org/cache/73e53dd1f16cf8c5da298418d2a6e452870cf50e -- http://www.webcitation.org/getfile.php?fileid=1c46e791d68e89e12d0c2532cc3cf629b8bc8c8e if dt[2]:find ('query', 1, true) or dt[2]:find ('cache', 1, true) or dt[2]:find ('getfile', 1, true) or tonumber(dt[2]) then return 'query'; end decode = base62(dt[2]); -- base62 string -> exponential number if not decode then return nil; -- nil return when dt[2] contains characters not in %w end dt = os.date('*t', string.format("%d", decode):sub(1,10)) -- exponential number -> text -> first 10 characters (a unix timestamp) -> a table of date parts decode = makeDate (dt.year, dt.month, dt.day, 'iso'); -- date comparisons are all done in iso format with western digits if non_western_digits then -- this wiki uses non-western digits? decode = mw.ustring.gsub (decode, '%d', digits); -- convert this wiki's non-western digits to western digits end return decode; end --[[--------------------------< decodeWaybackDate >----------------------- Given a URI-path to Wayback (eg. /web/20160901010101/http://example.com ) or Library of Congress Web Archives (eg. /all/20160901010101/http://example.com) or UK Government Web Archive (eg. /ukgwa/20160901010101/http://example.com or /tna/20160901010101/http://example.com) return the formatted date eg. "September 1, 2016" in df format Handle non-digits in snapshot ID such as "re_" and "-" and "*" returns two values: first value is one of these: valid date string in df format - wayback date is valid (including the text string 'index' when date is '/*/') empty string - wayback date is malformed (less than 8 digits, not a valid date) nil - wayback date is '/save/' or otherwise not a number second return value is an appropriate 'message' may or may not be formatted ]] local function decodeWaybackDate(path, df) local msg, snapdate; snapdate = path:gsub ('^/web/', ):gsub ('^/all/', ):gsub ('^/ukgwa/', ):gsub ('^/tna/', ):gsub ('^/', ); -- remove leading /web/, /all/, /ukgwa/, /tna/, or / snapdate = snapdate:match ('^[^/]+'); -- get timestamp if snapdate == "*" then -- eg. /web/*/http.., etc. return 'index'; -- return indicator that this url has an index date end snapdate = snapdate:gsub ('%a%a_%d?$', ):gsub ('%-', ); -- from date, remove any trailing "re_", dashes msg = ; if snapdate:match ('%*$') then -- a trailing '*' causes calendar display at archive .org snapdate = snapdate:gsub ('%*$', ); -- remove so not part of length calc later msg = inlineRed (err_warn_msgs.ts_cal, 'warning'); -- make a message end if not tonumber(snapdate) then return nil, 'ts_nan'; -- return nil (fatal error flag) and message selector end local dlen = snapdate:len(); if dlen < 8 then -- we need 8 digits TODO: but shouldn't this be testing for 14 digits? return , inlineRed (err_warn_msgs.ts_short, 'error'); -- return empty string and error message end local year, month, day = snapdate:match ('(%d%d%d%d)(%d%d)(%d%d)'); -- no need for snapdatelong here if not is_valid_date (year, month, day) then return , inlineRed (err_warn_msgs.ts_date, 'error'); -- return empty string and error message end snapdate = table.concat ({year, month, day}, '-'); -- date comparisons are all done in iso format if 14 == dlen then return snapdate, msg; -- return date with message if any else return snapdate, msg .. inlineRed (err_warn_msgs.ts_len, 'warning'); -- return date with warning message(s) end end --[[--------------------------< decodeArchiveisDate >----------------------- Given an Archive.is "long link" URI-path (e.g. /2016.08.28-144552/http://example.com) return the date in df format (e.g. if df = dmy, return 28 August 2016) Handles "." and "-" in snapshot date, so 2016.08.28-144552 is same as 20160828144552 returns two values: first value is one of these: valid date string in df format - archive.is date is valid (including the text string 'short link' when url is the short form) empty string - wayback date is malformed (not a number, less than 8 digits, not a valid date) nil - wayback date is '/save/' second return value is an appropriate 'message' may or may not be formatted ]] local function decodeArchiveisDate(path, df) local snapdate if path:match ('^/%w+$') then -- short form url path is '/' followed by some number of base 62 digits and nothing else return "short link" -- e.g. http://archive.is/hD1qz end snapdate = mw.text.split (path, '/')[2]:gsub('[%.%-]', ); -- get snapshot date, e.g. 2016.08.28-144552; remove periods and hyphens local dlen = string.len(snapdate) if dlen < 8 then -- we need 8 digits TODO: but shouldn't this be testing for 14 digits? return , inlineRed (err_warn_msgs.ts_short, 'error'); -- return empty string and error message end local year, month, day = snapdate:match ('(%d%d%d%d)(%d%d)(%d%d)'); -- no need for snapdatelong here if not is_valid_date (year, month, day) then return , inlineRed (err_warn_msgs.ts_date, 'error'); -- return empty string and error message end snapdate = table.concat ({year, month, day}, '-'); -- date comparisons are all done in iso format if 14 == dlen then return snapdate; -- return date else return snapdate, inlineRed (err_warn_msgs.ts_len, 'warning'); -- return date with warning message end end --[[--------------------------< serviceName >----------------------- Given a domain extracted by mw.uri.new() (eg. web.archive.org) set tail string and service ID ]] local function serviceName(host, no_link) local tracking; local index; host = host:lower():gsub ('^web%.(.+)', '%1'):gsub ('^www%.(.+)', '%1'); -- lowercase, remove web. and www. subdomains if services[host] then index = host; else for k, _ in pairs (services) do if host:find ('%f[%a]'..k:gsub ('([%.%-])', '%%%1')) then index = k; break; end end end if index then local out = {}; -- empty string in [1] so that concatenated result has leading single space ulx.url1.service = services[index][4] or 'other'; tracking = services[index][5] or categories.other; -- build tail string if false == services[index][1] then -- select prefix table.insert (out, prefixes.at); elseif true == services[index][1] then table.insert (out, prefixes.atthe); else table.insert (out, services[index][1]); end table.insert (out, make_wikilink (services[index][2], services[index][3], no_link)); -- add article wikilink if services[index][6] then -- add tail postfix if it exists table.insert (out, services[index][6]); end ulx.url1.tail = table.concat (out, ' '); -- put it all together; result has leading space character else -- here when unknown archive ulx.url1.service = 'other'; tracking = categories.unknown; ulx.url1.tail = table.concat ({, prefixes.at, host, inlineRed (err_warn_msgs.unknown_url, error)}, ' '); end track[tracking] = 1 end --[[--------------------------< parseExtraArgs >----------------------- Parse numbered arguments starting at 2, such as url2..url10, date2..date10, title2..title10 For example: Template loop detected: Template:Webarchive Three url arguments not in numeric sequence (1..4..7). Function only processes arguments numbered 2 or greater (in this case 4 and 7) It creates numeric sequenced table entries like: urlx.url2.url = <argument value for url4> urlx.url3.url = <argument value for url7> Returns the number of URL arguments found numbered 2 or greater (in this case returns "2") ]] local function parseExtraArgs(args) local i, j, argurl, argurl2, argdate, argtitle j = 2 for i = 2, config.maxurls do argurl = "url" .. i if args[argurl] then argurl2 = "url" .. j ulx[argurl2] = {} ulx[argurl2]["url"] = args[argurl] argdate = "date" .. i if args[argdate] then ulx[argurl2]["date"] = args[argdate] else ulx[argurl2]["date"] = inlineRed (err_warn_msgs.date_miss, 'warning'); end argtitle = "title" .. i if args[argtitle] then ulx[argurl2]["title"] = args[argtitle] else ulx[argurl2]["title"] = nil end j = j + 1 end end if j == 2 then return 0 else return j - 2 end end --[[--------------------------< comma >----------------------- Given a date string, return "," if it's MDY ]] local function comma(date) return (date and date:match ('%a+ +%d%d?(,) +%d%d%d%d')) or ; end --[[--------------------------< createRendering >----------------------- Return a rendering of the data in ulx[][] ]] local function createRendering() local displayfield local out = {}; local index_date, msg = ulx.url1.date:match ('(index)(.*)'); -- when ulx.url1.date extract 'index' text and message text (if there is a message) ulx.url1.date = ulx.url1.date:gsub ('index.*', 'index'); -- remove message if 'none' == ulx.url1.format then -- For Template:Wayback, Template:Webcite table.insert (out, '['); -- open extlink markup table.insert (out, ulx.url1.url); -- add url if ulx.url1.title then table.insert (out, ' ') -- the required space table.insert (out, ulx.url1.title) -- the title table.insert (out, ']'); -- close extlink markup table.insert (out, ulx.url1.tail); -- tail text if ulx.url1.date then table.insert (out, ' ('); -- open date text; TODO: why the html entity? replace with regular space? table.insert (out, 'index' == ulx.url1.date and s_text.archive or s_text.archived); -- add text table.insert (out, ' '); -- insert a space table.insert (out, ulx.url1.date); -- add date table.insert (out, ')'); -- close date text end else -- no title if index_date then -- when url date is 'index' table.insert (out, table.concat ({' ', s_text.Archive_index, ']'})); -- add the index link label table.insert (out, msg or ); -- add date mismatch message when url date is /*/ and |date= has valid date else table.insert (out, table.concat ({' ', s_text.Archived, '] '})); -- add link label for url has timestamp date (will include mismatch message if there is one) end if ulx.url1.date then if 'index' ~= ulx.url1.date then table.insert (out, ulx.url1.date); -- add date when data is not 'index' end table.insert (out, comma(ulx.url1.date)); -- add ',' if date format is mdy table.insert (out, ulx.url1.tail); -- add tail text else -- no date table.insert (out, ulx.url1.tail); -- add tail text end end if 0 < ulx.url1.extraurls then -- For multiple archive URLs local tot = ulx.url1.extraurls + 1 table.insert (out, '.') -- terminate first url table.insert (out, table.concat ({' ', s_text.addlarchives, ': '})); -- add header text for i=2, tot do -- loop through the additionals local index = table.concat ({'url', i}); -- make an index displayfield = ulx[index]['title'] and 'title' or 'date'; -- choose display text table.insert (out, '['); -- open extlink markup table.insert (out, ulx[index]['url']); -- add the url table.insert (out, ' '); -- the required space table.insert (out, ulx[index][displayfield]); -- add the label table.insert (out, ']'); -- close extlink markup table.insert (out, i==tot and '.' or ', '); -- add terminator end end return table.concat (out); -- make a big string and done else -- For Template:Cite archives if 'addlarchives' == ulx.url1.format then -- Multiple archive services table.insert (out, table.concat ({s_text.addlarchives, ': '})); -- add header text else -- Multiple pages from the same archive table.insert (out, table.concat ({s_text.addlpages, ' '})); -- add header text table.insert (out, ulx.url1.date); -- add date to header text table.insert (out, ': '); -- close header text end local tot = ulx.url1.extraurls + 1; for i=1, tot do -- loop through the additionals local index = table.concat ({'url', i}); -- make an index table.insert (out, '['); -- open extlink markup table.insert (out, ulx[index]['url']); -- add url table.insert (out, ' '); -- add required space displayfield = ulx[index]['title']; if 'addlarchives' == ulx.url1.format then if not displayfield then displayfield = ulx[index]['date'] end else -- must be addlpages if not displayfield then displayfield = table.concat ({s_text.Page, ' ', i}); end end table.insert (out, displayfield); -- add title, date, page label text table.insert (out, ']'); -- close extlink markup table.insert (out, (i==tot and '.' or ', ')); -- add terminator end return table.concat (out); -- make a big string and done end end --[[--------------------------< P A R A M E T E R _ N A M E _ X L A T E >-------------------------------------- for internaltionalization, translate local-language parameter names to their English equivalents TODO: return error message if multiple aliases of the same canonical parameter name are found? returns two tables: new_args - holds canonical form parameters and their values either from translation or because the parameter was already in canonical form origin - maps canonical-form parameter names to their untranslated (local language) form for error messaging in the local language unrecognized parameters are ignored ]] local function parameter_name_xlate (args, params, enum_params) local name; -- holds modifiable name of the parameter name during evaluation local enum; -- for enumerated parameters, holds the enumerator during evaluation local found = false; -- flag used to break out of nested for loops local new_args = {}; -- a table that holds canonical and translated parameter k/v pairs local origin = {}; -- a table that maps original (local language) parameter names to their canonical name for local language error messaging local unnamed_params; -- set true when unsupported positional parameters are detected for k, v in pairs (args) do -- loop through all of the arguments in the args table name = k; -- copy of original parameter name if 'string' == type (k) then if non_western_digits then -- true when non-western digits supported at this wiki name = mw.ustring.gsub (name, '%d', digits); -- convert this wiki's non-western digits to western digits end enum = name:match ('%d+$'); -- get parameter enumerator if it exists; nil else if not enum then -- no enumerator so looking for non-enumnerated parameters -- TODO: insert shortcut here? if params[name] then name holds the canonical parameter name; no need to search further for pname, aliases in pairs (params) do -- loop through each parameter the params table for _, alias in ipairs (aliases) do -- loop through each alias in the parameter's aliases table if name == alias then new_args[pname] = v; -- create a new entry in the new_args table origin [pname] = k; -- create an entry to make canonical parameter name to original local language parameter name found = true; -- flag so that we can break out of these nested for loops break; -- no need to search the rest of the aliases table for name so go on to the next k, v pair end end if found then -- true when we found an alias that matched name found = false; -- reset the flag break; -- go do next args k/v pair end end else -- enumerated parameters name = name:gsub ('%d$', '#'); -- replace enumeration digits with place holder for table search -- TODO: insert shortcut here? if num_params[name] then name holds the canonical parameter name; no need to search further for pname, aliases in pairs (enum_params) do -- loop through each parameter the num_params table for _, alias in ipairs (aliases) do -- loop through each alias in the parameter's aliases table if name == alias then pname = pname:gsub ('#$', enum); -- replace the '#' place holder with the actual enumerator new_args[pname] = v; -- create a new entry in the new_args table origin [pname] = k; -- create an entry to make canonical parameter name to original local language parameter name found = true; -- flag so that we can break out of these nested for loops break; -- no need to search the rest of the aliases table for name so go on to the next k, v pair end end if found then -- true when we found an alias that matched name found = false; -- reset the flag break; -- go do next args k/v pair end end end else unnamed_params = true; -- flag for unsupported positional parameters end end -- for k, v return new_args, origin, unnamed_params; end --[[--------------------------< W E B A R C H I V E >---------------------------------------------------------- template entry point ]] local function webarchive(frame) local args = getArgs (frame); local data = mw.loadData (table.concat ({ -- make a data module name; sandbox or live 'Module:Webarchive/data', frame:getTitle():find('sandbox', 1, true) and '/sandbox' or -- this instance is ./sandbox then append /sandbox })); categories = data.categories; -- fill in the forward declarations config = data.config; if data.digits.enable then digits = data.digits; -- for i18n; table of digits in the local wiki's language non_western_digits = true; -- use_non_western_digits end err_warn_msgs = data.err_warn_msgs; excepted_pages = data.excepted_pages; month_num = data.month_num; -- for i18n; table of month names in the local wiki's language prefixes = data.prefixes; services = data.services; s_text = data.s_text; uncategorized_namespaces = data.uncategorized_namespaces; uncategorized_subpages = data.uncategorized_subpages; local origin = {}; -- holds a map of English to local language parameter names used in the current template; not currently used local unnamed_params; -- boolean set to true when template call has unnamed parameters args, origin, unnamed_params = parameter_name_xlate (args, data.params, data.enum_params); -- translate parameter names in args to English local date, format, msg, udate, uri, url; local ldf = 'iso'; -- when there is no |date= parameter, render url dates in iso format if args.url and args.url1 then -- URL argument (first) return inlineError (data.crit_err_msgs.conflicting, {origin.url, origin.url1}); end url = args.url or args.url1; if not url then return inlineError (data.crit_err_msgs.empty); end -- these iabot bugs perportedly fixed; removing these causes lua script error --[[ -- at Template:Webarchive/testcases/Production; resolve that before deleting these tests if mw.ustring.find( url, "https://web.http", 1, true ) then -- track bug - TODO: IAbot bug; not known if the bug has been fixed; deferred track[categories.error] = 1; return inlineError (data.crit_err_msgs.iabot1); end if url == "https://web.archive.org/http:/" then -- track bug - TODO: IAbot bug; not known if the bug has been fixed; deferred track[categories.error] = 1; return inlineError (data.crit_err_msgs.iabot2); end ]] if not (url:lower():find ('^http') or url:find ('^//')) then return inlineError (data.crit_err_msgs.invalid_url ); end ulx.url1 = {} ulx.url1.url = url ulx.url1.extraurls = parseExtraArgs(args) local good = false; good, uri = pcall (mw.uri.new, ulx.url1.url); -- get a table of uri parts from this url; protected mode to prevent lua error when ulx.url1.url is malformed if not good or nil == uri.host then -- abandon when ulx.url1.url is malformed return inlineError (data.crit_err_msgs.invalid_url); end serviceName(uri.host, args.nolink) if args.date and args.date1 then -- Date argument return inlineError (data.crit_err_msgs.conflicting, {origin.date, origin.date1}); end date = args.date or args.date1; date = date and date:gsub (' +', ' '); -- replace multiple spaces with a single space if date and config.verifydates then if '*' == date then date = 'index'; ldf = 'iso'; -- set to default format elseif 'mdy' == date then date = nil; -- if date extracted from URL, ldf = 'mdy'; -- then |date=mdy overrides iso elseif 'dmy' == date then date = nil; -- if date extracted from URL, ldf = 'dmy'; -- then |date=dmy overrides iso elseif 'ymd' == date then date = nil; -- if date extracted from URL, ldf = 'ymd'; -- then |date=ymd overrides iso else date, ldf = decode_date (date); -- get an iso format date from date and get date's original format end end if 'wayback' == ulx.url1.service or 'locwebarchives' == ulx.url1.service or 'ukgwa' == ulx.url1.service then if date then if config.verifydates then if ldf then udate, msg = decodeWaybackDate (uri.path); -- get the url date in iso format and format of date in |date=; 'index' when wayback url date is * if not udate then -- this is the only 'fatal' error return return inlineError (data.crit_err_msgs[msg]); end if udate ~= date then -- date comparison using iso format dates date = udate; msg = table.concat ({ inlineRed (err_warn_msgs.mismatch, 'warning'), -- add warning message msg, -- add message if there is one }); end end end else -- no |date= udate, msg = decodeWaybackDate (uri.path); if not udate then -- this is the only 'fatal' error return return inlineError (data.crit_err_msgs[msg]); end if == udate then date = nil; -- unset else date = udate; end end elseif 'webcite' == ulx.url1.service then if date then if config.verifydates then if ldf then udate = decodeWebciteDate (uri.path); -- get the url date in iso format if 'query' ~= udate then -- skip if query if udate ~= date then -- date comparison using iso format dates date = udate; msg = table.concat ({ inlineRed (err_warn_msgs.mismatch, 'warning'), }); end end end end else date = decodeWebciteDate( uri.path, "iso" ) if date == "query" then date = nil; -- unset msg = inlineRed (err_warn_msgs.date_miss, 'warning'); elseif not date then -- invalid base62 string date = inlineRed (err_warn_msgs.date1, 'error'); end end elseif 'archiveis' == ulx.url1.service then if date then if config.verifydates then if ldf then udate, msg = decodeArchiveisDate (uri.path) -- get the url date in iso format if 'short link' ~= udate then -- skip if short link if udate ~= date then -- date comparison using iso format dates date = udate; msg = table.concat ({ inlineRed (err_warn_msgs.mismatch, 'warning'), -- add warning message msg, -- add message if there is one }); end end end end else -- no |date= udate, msg = decodeArchiveisDate( uri.path, "iso" ) if udate == "short link" then date = nil; -- unset msg = inlineRed (err_warn_msgs.date_miss, 'warning'); elseif == udate then date = nil; -- unset else date = udate; end end else -- some other service if not date then msg = inlineRed (err_warn_msgs.date_miss, 'warning'); end end if 'index' == date then ulx.url1.date = date .. (msg or ); -- create index + message (if there is one) elseif date then ulx.url1.date = makeDate (date, nil, nil, ldf) .. (msg or ); -- create a date in the wiki's local language + message (if there is one) else ulx.url1.date = msg; end format = args.format; -- Format argument if not format then format = "none" else for k, v in pairs (data.format_vals) do -- |format= accepts two specific values loop through a table of those values local found; -- declare a nil flag for _, p in ipairs (v) do -- loop through local language variants if format == p then -- when |format= value matches format = k; -- use name from table key found = true; -- declare found so that we can break out of outer for loop break; -- break out of inner for loop end end if found then break; end end if format == "addlpages" then if not ulx.url1.date then format = "none" end elseif format == "addlarchives" then format = "addlarchives" else format = "none" end end ulx.url1.format = format if args.title and args.title1 then -- Title argument return inlineError (data.crit_err_msgs.conflicting, {origin.title, origin.title1}); end ulx.url1.title = args.title or args.title1; local rend = createRendering() if not rend then return inlineError (data.crit_err_msgs.unknown); end return rend .. ((unnamed_params and inlineRed (err_warn_msgs.unnamed_params, 'warning')) or ) .. createTracking(); end --[[--------------------------< E X P O R T E D F U N C T I O N S >------------------------------------------ ]] return {webarchive = webarchive}; Time Out (magazine), published by AMC's Filmsite. Retrieved August 15, 2017.
  64. ↑ Empire's The 500 Greatest Movies of All Time. Empire magazine. Retrieved on October 23, 2010.
  65. ↑ The Best 1,000 Movies Ever Made. The New York Times via Internet Archive. Retrieved August 15, 2017.
  66. ↑ The 100 Greatest Movies of All Time. Total Film via Internet Archive. Retrieved August 15, 2017.
  67. ↑ Filmsite's 100 Greatest Films. AMC FilmSite.org. Retrieved on October 23, 2010.
  68. ↑ The Best Movies of All Time (10,790 Most Notable). Films101.com. Retrieved on October 23, 2010.
  69. ↑ CineMontage's The 75 Best Edited Films. Editors Guild Magazine volume 1, issue 3 via Internet Archive. Published May 2012. Retrieved August 15, 2017.
  70. ↑ The 75 Best Edited Films Of All Time --[[ ---------------------------------- Lua module implementing the Template loop detected: Template:Webarchive template. A merger of the functionality of three templates: Template:Wayback, Template:Webcite and Template:Cite archives ]] --[[--------------------------< D E P E N D E N C I E S >------------------------------------------------------ ]] require('Module:No globals'); local getArgs = require ('Module:Arguments').getArgs; --[[--------------------------< F O R W A R D D E C L A R A T I O N S >-------------------------------------- ]] local categories = {}; -- category names local config = {}; -- global configuration settings local digits = {}; -- for i18n; table that translates local-wiki digits to western digits local err_warn_msgs = {}; -- error and warning messages local excepted_pages = {}; local month_num = {}; -- for i18n; table that translates local-wiki month names to western digits local prefixes = {}; -- service provider tail string prefixes local services = {}; -- archive service provider data from local s_text = {}; -- table of static text strings used to build final rendering local uncategorized_namespaces = {}; -- list of namespaces that we should not categorize local uncategorized_subpages = {}; -- list of subpages that should not be categorized --[[--------------------------< P A G E S C O P E I D E N T I F I E R S >---------------------------------- ]] local non_western_digits; -- boolean flag set true when data.digits.enable is true local this_page = mw.title.getCurrentTitle(); local track = {}; -- Associative array to hold tracking categories local ulx = {}; -- Associative array to hold template data --[[--------------------------< S U B S T I T U T E >---------------------------------------------------------- Populates numbered arguments in a message string using an argument table. ]] local function substitute (msg, args) return args and mw.message.newRawMessage (msg, args):plain() or msg; end --[[--------------------------< tableLength >----------------------- Given a 1-D table, return number of elements ]] local function tableLength(T) local count = 0 for _ in pairs(T) do count = count + 1 end return count end --[=[-------------------------< M A K E _ W I K I L I N K >---------------------------------------------------- Makes a wikilink; when both link and display text is provided, returns a wikilink in the form D; if only link is provided, returns a wikilink in the form L; if neither are provided or link is omitted, returns an empty string. ]=] local function make_wikilink (link, display, no_link) if nil == no_link then if link and ( ~= link) then if display and ( ~= display) then return table.concat ({'', display, ''}); else return table.concat ({'', link, ''}); end end return display or ; -- link not set so return the display text else -- no_link if display and ( ~= display) then -- if there is display text return display; -- return that else return link or ; -- return the target article name or empty string end end end --[[--------------------------< createTracking >----------------------- Return data in track[] ie. tracking categories ]] local function createTracking() if not excepted_pages[this_page.fullText] then -- namespace:title/fragment is allowed to be categorized (typically this module's / template's testcases page(s)) if uncategorized_namespaces[this_page.nsText] then return ; -- this page not to be categorized so return empty string end for _,v in ipairs (uncategorized_subpages) do -- cycle through page name patterns if this_page.text:match (v) then -- test page name against each pattern return ; -- this subpage type not to be categorized so return empty string end end end local out = {}; if tableLength(track) > 0 then for key, _ in pairs(track) do -- loop through table table.insert (out, make_wikilink (key)); -- and convert category names to links end end return table.concat (out); -- concat into one big string; empty string if table is empty end --[[--------------------------< inlineError >----------------------- Critical error. Render output completely in red. Add to tracking category. This function called as the last thing before abandoning this module ]] local function inlineError (msg, args) track[categories.error] = 1 return table.concat ({ 'Error in ', -- open the error message span config.tname, -- insert the local language template name ' template: ', substitute (msg, args), -- insert the formatted error message '.', -- close the span createTracking() -- add the category }) end --[[--------------------------< inlineRed >----------------------- Render a text fragment in red, such as a warning as part of the final output. Add tracking category. ]] local function inlineRed(msg, trackmsg) if trackmsg == "warning" then track[categories.warning] = 1; elseif trackmsg == "error" then track[categories.error] = 1; end return '' .. msg .. '' end --[[--------------------------< base62 >----------------------- Convert base-62 to base-10 Credit: https://de.wikipedia.org/wiki/Modul:Expr ]] local function base62( value ) local r = 1 -- default return value is input value is malformed if value:match ('%W') then -- value must only be in the set [0-9a-zA-Z] return; -- nil return when value contains extraneous characters end local n = #value -- number of characters in value local k = 1 local c r = 0 for i = n, 1, -1 do -- loop through all characters in value from ls digit to ms digit c = value:byte( i, i ) if c >= 48 and c <= 57 then -- character is digit 0-9 c = c - 48 elseif c >= 65 and c <= 90 then -- character is ascii a-z c = c - 55 else -- must be ascii A-Z c = c - 61 end r = r + c * k -- accumulate this base62 character's value k = k * 62 -- bump for next end -- for i return r end --[[--------------------------< D E C O D E _ D A T E >-------------------------------------------------------- Given a date string, return it in iso format along with an indicator of the date's format. Except that month names must be recognizable as legitimate month names with proper capitalization, and that the date string must match one of the recognized date formats, no error checking is done here; return nil else ]] local function decode_date (date_str) local patterns = { ['dmy'] = {'^(%d%d?) +([^%s%d]+) +(%d%d%d%d)$', 'd', 'm', 'y'}, -- %a does not recognize unicode combining characters used by some languages ['mdy'] = {'^([^%s%d]+) (%d%d?), +(%d%d%d%d)$', 'm', 'd', 'y'}, ['ymd'] = {'^(%d%d%d%d) +([^%s%d]+) (%d%d?)$', 'y', 'm', 'd'}, -- not mos compliant at en.wiki but may be acceptible at other wikis }; local t = {}; if non_western_digits then -- this wiki uses non-western digits? date_str = mw.ustring.gsub (date_str, '%d', digits); -- convert this wiki's non-western digits to western digits end if date_str:match ('^%d%d%d%d%-%d%d%-%d%d$') then -- already an iso format date, return western digits form return date_str, 'iso'; end for k, v in pairs (patterns) do local c1, c2, c3 = mw.ustring.match (date_str, patterns[k][1]); -- c1 .. c3 are captured but we don't know what they hold if c1 then -- set on match t = { -- translate unspecified captures to y, m, and d [patterns[k][2]] = c1, -- fill the table of captures with the captures [patterns[k][3]] = c2, -- take index names from src_pattern table and assign sequential captures [patterns[k][4]] = c3, }; if month_num[t.m] then -- when month not already a number t.m = month_num[t.m]; -- replace valid month name with a number else return nil, 'iso'; -- not a valid date form because month not valid end return mw.ustring.format ('%.4d-%.2d-%.2d', t.y, t.m, t.d), k; -- return date in iso format end end return nil, 'iso'; -- date could not be decoded; return nil and default iso date end --[[--------------------------< makeDate >----------------------- Given year, month, day numbers, (zero-padded or not) return a full date in df format where df may be one of: mdy, dmy, iso, ymd on entry, year, month, day are presumed to be correct for the date that they represent; all are required in this module, makeDate() is sometimes given an iso-format date in year: makeDate (2018-09-20, nil, nil, df) this works because table.concat() sees only one table member ]] local function makeDate (year, month, day, df) local format = { ['dmy'] = 'j F Y', ['mdy'] = 'F j, Y', ['ymd'] = 'Y F j', ['iso'] = 'Y-m-d', }; local date = table.concat ({year, month, day}, '-'); -- assemble year-initial numeric-format date (zero padding not required here) if non_western_digits then -- this wiki uses non-western digits? date = mw.ustring.gsub (date, '%d', digits); -- convert this wiki's non-western digits to western digits end return mw.getContentLanguage():formatDate (format[df], date); end --[[--------------------------< I S _ V A L I D _ D A T E >---------------------------------------------------- Returns true if date is after 31 December 1899 (why is 1900 the min year? shouldn't the internet's date-of-birth be min year?), not after today's date, and represents a valid date (29 February 2017 is not a valid date). Applies Gregorian leapyear rules. all arguments are required ]] local function is_valid_date (year, month, day) local days_in_month = {31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31}; local month_length; local y, m, d; local today = os.date ('*t'); -- fetch a table of current date parts if not year or == year or not month or == month or not day or == day then return false; -- something missing end y = tonumber (year); m = tonumber (month); d = tonumber (day); if 1900 > y or today.year < y or 1 > m or 12 < m then -- year and month are within bounds TODO: 1900? return false; end if (2==m) then -- if February month_length = 28; -- then 28 days unless if (0==(y%4) and (0~=(y%100) or 0==(y%400))) then -- is a leap year? month_length = 29; -- if leap year then 29 days in February end else month_length=days_in_month[m]; end if 1 > d or month_length < d then -- day is within bounds return false; end -- here when date parts represent a valid date return os.time({['year']=y, ['month']=m, ['day']=d, ['hour']=0}) <= os.time(); -- date at midnight must be less than or equal to current date/time end --[[--------------------------< decodeWebciteDate >----------------------- Given a URI-path to Webcite (eg. /67xHmVFWP) return the encoded date in df format returns date string in df format - webcite date is a unix timestamp encoded as bae62 or the string 'query' ]] local function decodeWebciteDate(path, df) local dt = {}; local decode; dt = mw.text.split(path, "/") -- valid URL formats that are not base62 -- http://www.webcitation.org/query?id=1138911916587475 -- http://www.webcitation.org/query?url=http..&date=2012-06-01+21:40:03 -- http://www.webcitation.org/1138911916587475 -- http://www.webcitation.org/cache/73e53dd1f16cf8c5da298418d2a6e452870cf50e -- http://www.webcitation.org/getfile.php?fileid=1c46e791d68e89e12d0c2532cc3cf629b8bc8c8e if dt[2]:find ('query', 1, true) or dt[2]:find ('cache', 1, true) or dt[2]:find ('getfile', 1, true) or tonumber(dt[2]) then return 'query'; end decode = base62(dt[2]); -- base62 string -> exponential number if not decode then return nil; -- nil return when dt[2] contains characters not in %w end dt = os.date('*t', string.format("%d", decode):sub(1,10)) -- exponential number -> text -> first 10 characters (a unix timestamp) -> a table of date parts decode = makeDate (dt.year, dt.month, dt.day, 'iso'); -- date comparisons are all done in iso format with western digits if non_western_digits then -- this wiki uses non-western digits? decode = mw.ustring.gsub (decode, '%d', digits); -- convert this wiki's non-western digits to western digits end return decode; end --[[--------------------------< decodeWaybackDate >----------------------- Given a URI-path to Wayback (eg. /web/20160901010101/http://example.com ) or Library of Congress Web Archives (eg. /all/20160901010101/http://example.com) or UK Government Web Archive (eg. /ukgwa/20160901010101/http://example.com or /tna/20160901010101/http://example.com) return the formatted date eg. "September 1, 2016" in df format Handle non-digits in snapshot ID such as "re_" and "-" and "*" returns two values: first value is one of these: valid date string in df format - wayback date is valid (including the text string 'index' when date is '/*/') empty string - wayback date is malformed (less than 8 digits, not a valid date) nil - wayback date is '/save/' or otherwise not a number second return value is an appropriate 'message' may or may not be formatted ]] local function decodeWaybackDate(path, df) local msg, snapdate; snapdate = path:gsub ('^/web/', ):gsub ('^/all/', ):gsub ('^/ukgwa/', ):gsub ('^/tna/', ):gsub ('^/', ); -- remove leading /web/, /all/, /ukgwa/, /tna/, or / snapdate = snapdate:match ('^[^/]+'); -- get timestamp if snapdate == "*" then -- eg. /web/*/http.., etc. return 'index'; -- return indicator that this url has an index date end snapdate = snapdate:gsub ('%a%a_%d?$', ):gsub ('%-', ); -- from date, remove any trailing "re_", dashes msg = ; if snapdate:match ('%*$') then -- a trailing '*' causes calendar display at archive .org snapdate = snapdate:gsub ('%*$', ); -- remove so not part of length calc later msg = inlineRed (err_warn_msgs.ts_cal, 'warning'); -- make a message end if not tonumber(snapdate) then return nil, 'ts_nan'; -- return nil (fatal error flag) and message selector end local dlen = snapdate:len(); if dlen < 8 then -- we need 8 digits TODO: but shouldn't this be testing for 14 digits? return , inlineRed (err_warn_msgs.ts_short, 'error'); -- return empty string and error message end local year, month, day = snapdate:match ('(%d%d%d%d)(%d%d)(%d%d)'); -- no need for snapdatelong here if not is_valid_date (year, month, day) then return , inlineRed (err_warn_msgs.ts_date, 'error'); -- return empty string and error message end snapdate = table.concat ({year, month, day}, '-'); -- date comparisons are all done in iso format if 14 == dlen then return snapdate, msg; -- return date with message if any else return snapdate, msg .. inlineRed (err_warn_msgs.ts_len, 'warning'); -- return date with warning message(s) end end --[[--------------------------< decodeArchiveisDate >----------------------- Given an Archive.is "long link" URI-path (e.g. /2016.08.28-144552/http://example.com) return the date in df format (e.g. if df = dmy, return 28 August 2016) Handles "." and "-" in snapshot date, so 2016.08.28-144552 is same as 20160828144552 returns two values: first value is one of these: valid date string in df format - archive.is date is valid (including the text string 'short link' when url is the short form) empty string - wayback date is malformed (not a number, less than 8 digits, not a valid date) nil - wayback date is '/save/' second return value is an appropriate 'message' may or may not be formatted ]] local function decodeArchiveisDate(path, df) local snapdate if path:match ('^/%w+$') then -- short form url path is '/' followed by some number of base 62 digits and nothing else return "short link" -- e.g. http://archive.is/hD1qz end snapdate = mw.text.split (path, '/')[2]:gsub('[%.%-]', ); -- get snapshot date, e.g. 2016.08.28-144552; remove periods and hyphens local dlen = string.len(snapdate) if dlen < 8 then -- we need 8 digits TODO: but shouldn't this be testing for 14 digits? return , inlineRed (err_warn_msgs.ts_short, 'error'); -- return empty string and error message end local year, month, day = snapdate:match ('(%d%d%d%d)(%d%d)(%d%d)'); -- no need for snapdatelong here if not is_valid_date (year, month, day) then return , inlineRed (err_warn_msgs.ts_date, 'error'); -- return empty string and error message end snapdate = table.concat ({year, month, day}, '-'); -- date comparisons are all done in iso format if 14 == dlen then return snapdate; -- return date else return snapdate, inlineRed (err_warn_msgs.ts_len, 'warning'); -- return date with warning message end end --[[--------------------------< serviceName >----------------------- Given a domain extracted by mw.uri.new() (eg. web.archive.org) set tail string and service ID ]] local function serviceName(host, no_link) local tracking; local index; host = host:lower():gsub ('^web%.(.+)', '%1'):gsub ('^www%.(.+)', '%1'); -- lowercase, remove web. and www. subdomains if services[host] then index = host; else for k, _ in pairs (services) do if host:find ('%f[%a]'..k:gsub ('([%.%-])', '%%%1')) then index = k; break; end end end if index then local out = {}; -- empty string in [1] so that concatenated result has leading single space ulx.url1.service = services[index][4] or 'other'; tracking = services[index][5] or categories.other; -- build tail string if false == services[index][1] then -- select prefix table.insert (out, prefixes.at); elseif true == services[index][1] then table.insert (out, prefixes.atthe); else table.insert (out, services[index][1]); end table.insert (out, make_wikilink (services[index][2], services[index][3], no_link)); -- add article wikilink if services[index][6] then -- add tail postfix if it exists table.insert (out, services[index][6]); end ulx.url1.tail = table.concat (out, ' '); -- put it all together; result has leading space character else -- here when unknown archive ulx.url1.service = 'other'; tracking = categories.unknown; ulx.url1.tail = table.concat ({, prefixes.at, host, inlineRed (err_warn_msgs.unknown_url, error)}, ' '); end track[tracking] = 1 end --[[--------------------------< parseExtraArgs >----------------------- Parse numbered arguments starting at 2, such as url2..url10, date2..date10, title2..title10 For example: Template loop detected: Template:Webarchive Three url arguments not in numeric sequence (1..4..7). Function only processes arguments numbered 2 or greater (in this case 4 and 7) It creates numeric sequenced table entries like: urlx.url2.url = <argument value for url4> urlx.url3.url = <argument value for url7> Returns the number of URL arguments found numbered 2 or greater (in this case returns "2") ]] local function parseExtraArgs(args) local i, j, argurl, argurl2, argdate, argtitle j = 2 for i = 2, config.maxurls do argurl = "url" .. i if args[argurl] then argurl2 = "url" .. j ulx[argurl2] = {} ulx[argurl2]["url"] = args[argurl] argdate = "date" .. i if args[argdate] then ulx[argurl2]["date"] = args[argdate] else ulx[argurl2]["date"] = inlineRed (err_warn_msgs.date_miss, 'warning'); end argtitle = "title" .. i if args[argtitle] then ulx[argurl2]["title"] = args[argtitle] else ulx[argurl2]["title"] = nil end j = j + 1 end end if j == 2 then return 0 else return j - 2 end end --[[--------------------------< comma >----------------------- Given a date string, return "," if it's MDY ]] local function comma(date) return (date and date:match ('%a+ +%d%d?(,) +%d%d%d%d')) or ; end --[[--------------------------< createRendering >----------------------- Return a rendering of the data in ulx[][] ]] local function createRendering() local displayfield local out = {}; local index_date, msg = ulx.url1.date:match ('(index)(.*)'); -- when ulx.url1.date extract 'index' text and message text (if there is a message) ulx.url1.date = ulx.url1.date:gsub ('index.*', 'index'); -- remove message if 'none' == ulx.url1.format then -- For Template:Wayback, Template:Webcite table.insert (out, '['); -- open extlink markup table.insert (out, ulx.url1.url); -- add url if ulx.url1.title then table.insert (out, ' ') -- the required space table.insert (out, ulx.url1.title) -- the title table.insert (out, ']'); -- close extlink markup table.insert (out, ulx.url1.tail); -- tail text if ulx.url1.date then table.insert (out, ' ('); -- open date text; TODO: why the html entity? replace with regular space? table.insert (out, 'index' == ulx.url1.date and s_text.archive or s_text.archived); -- add text table.insert (out, ' '); -- insert a space table.insert (out, ulx.url1.date); -- add date table.insert (out, ')'); -- close date text end else -- no title if index_date then -- when url date is 'index' table.insert (out, table.concat ({' ', s_text.Archive_index, ']'})); -- add the index link label table.insert (out, msg or ); -- add date mismatch message when url date is /*/ and |date= has valid date else table.insert (out, table.concat ({' ', s_text.Archived, '] '})); -- add link label for url has timestamp date (will include mismatch message if there is one) end if ulx.url1.date then if 'index' ~= ulx.url1.date then table.insert (out, ulx.url1.date); -- add date when data is not 'index' end table.insert (out, comma(ulx.url1.date)); -- add ',' if date format is mdy table.insert (out, ulx.url1.tail); -- add tail text else -- no date table.insert (out, ulx.url1.tail); -- add tail text end end if 0 < ulx.url1.extraurls then -- For multiple archive URLs local tot = ulx.url1.extraurls + 1 table.insert (out, '.') -- terminate first url table.insert (out, table.concat ({' ', s_text.addlarchives, ': '})); -- add header text for i=2, tot do -- loop through the additionals local index = table.concat ({'url', i}); -- make an index displayfield = ulx[index]['title'] and 'title' or 'date'; -- choose display text table.insert (out, '['); -- open extlink markup table.insert (out, ulx[index]['url']); -- add the url table.insert (out, ' '); -- the required space table.insert (out, ulx[index][displayfield]); -- add the label table.insert (out, ']'); -- close extlink markup table.insert (out, i==tot and '.' or ', '); -- add terminator end end return table.concat (out); -- make a big string and done else -- For Template:Cite archives if 'addlarchives' == ulx.url1.format then -- Multiple archive services table.insert (out, table.concat ({s_text.addlarchives, ': '})); -- add header text else -- Multiple pages from the same archive table.insert (out, table.concat ({s_text.addlpages, ' '})); -- add header text table.insert (out, ulx.url1.date); -- add date to header text table.insert (out, ': '); -- close header text end local tot = ulx.url1.extraurls + 1; for i=1, tot do -- loop through the additionals local index = table.concat ({'url', i}); -- make an index table.insert (out, '['); -- open extlink markup table.insert (out, ulx[index]['url']); -- add url table.insert (out, ' '); -- add required space displayfield = ulx[index]['title']; if 'addlarchives' == ulx.url1.format then if not displayfield then displayfield = ulx[index]['date'] end else -- must be addlpages if not displayfield then displayfield = table.concat ({s_text.Page, ' ', i}); end end table.insert (out, displayfield); -- add title, date, page label text table.insert (out, ']'); -- close extlink markup table.insert (out, (i==tot and '.' or ', ')); -- add terminator end return table.concat (out); -- make a big string and done end end --[[--------------------------< P A R A M E T E R _ N A M E _ X L A T E >-------------------------------------- for internaltionalization, translate local-language parameter names to their English equivalents TODO: return error message if multiple aliases of the same canonical parameter name are found? returns two tables: new_args - holds canonical form parameters and their values either from translation or because the parameter was already in canonical form origin - maps canonical-form parameter names to their untranslated (local language) form for error messaging in the local language unrecognized parameters are ignored ]] local function parameter_name_xlate (args, params, enum_params) local name; -- holds modifiable name of the parameter name during evaluation local enum; -- for enumerated parameters, holds the enumerator during evaluation local found = false; -- flag used to break out of nested for loops local new_args = {}; -- a table that holds canonical and translated parameter k/v pairs local origin = {}; -- a table that maps original (local language) parameter names to their canonical name for local language error messaging local unnamed_params; -- set true when unsupported positional parameters are detected for k, v in pairs (args) do -- loop through all of the arguments in the args table name = k; -- copy of original parameter name if 'string' == type (k) then if non_western_digits then -- true when non-western digits supported at this wiki name = mw.ustring.gsub (name, '%d', digits); -- convert this wiki's non-western digits to western digits end enum = name:match ('%d+$'); -- get parameter enumerator if it exists; nil else if not enum then -- no enumerator so looking for non-enumnerated parameters -- TODO: insert shortcut here? if params[name] then name holds the canonical parameter name; no need to search further for pname, aliases in pairs (params) do -- loop through each parameter the params table for _, alias in ipairs (aliases) do -- loop through each alias in the parameter's aliases table if name == alias then new_args[pname] = v; -- create a new entry in the new_args table origin [pname] = k; -- create an entry to make canonical parameter name to original local language parameter name found = true; -- flag so that we can break out of these nested for loops break; -- no need to search the rest of the aliases table for name so go on to the next k, v pair end end if found then -- true when we found an alias that matched name found = false; -- reset the flag break; -- go do next args k/v pair end end else -- enumerated parameters name = name:gsub ('%d$', '#'); -- replace enumeration digits with place holder for table search -- TODO: insert shortcut here? if num_params[name] then name holds the canonical parameter name; no need to search further for pname, aliases in pairs (enum_params) do -- loop through each parameter the num_params table for _, alias in ipairs (aliases) do -- loop through each alias in the parameter's aliases table if name == alias then pname = pname:gsub ('#$', enum); -- replace the '#' place holder with the actual enumerator new_args[pname] = v; -- create a new entry in the new_args table origin [pname] = k; -- create an entry to make canonical parameter name to original local language parameter name found = true; -- flag so that we can break out of these nested for loops break; -- no need to search the rest of the aliases table for name so go on to the next k, v pair end end if found then -- true when we found an alias that matched name found = false; -- reset the flag break; -- go do next args k/v pair end end end else unnamed_params = true; -- flag for unsupported positional parameters end end -- for k, v return new_args, origin, unnamed_params; end --[[--------------------------< W E B A R C H I V E >---------------------------------------------------------- template entry point ]] local function webarchive(frame) local args = getArgs (frame); local data = mw.loadData (table.concat ({ -- make a data module name; sandbox or live 'Module:Webarchive/data', frame:getTitle():find('sandbox', 1, true) and '/sandbox' or -- this instance is ./sandbox then append /sandbox })); categories = data.categories; -- fill in the forward declarations config = data.config; if data.digits.enable then digits = data.digits; -- for i18n; table of digits in the local wiki's language non_western_digits = true; -- use_non_western_digits end err_warn_msgs = data.err_warn_msgs; excepted_pages = data.excepted_pages; month_num = data.month_num; -- for i18n; table of month names in the local wiki's language prefixes = data.prefixes; services = data.services; s_text = data.s_text; uncategorized_namespaces = data.uncategorized_namespaces; uncategorized_subpages = data.uncategorized_subpages; local origin = {}; -- holds a map of English to local language parameter names used in the current template; not currently used local unnamed_params; -- boolean set to true when template call has unnamed parameters args, origin, unnamed_params = parameter_name_xlate (args, data.params, data.enum_params); -- translate parameter names in args to English local date, format, msg, udate, uri, url; local ldf = 'iso'; -- when there is no |date= parameter, render url dates in iso format if args.url and args.url1 then -- URL argument (first) return inlineError (data.crit_err_msgs.conflicting, {origin.url, origin.url1}); end url = args.url or args.url1; if not url then return inlineError (data.crit_err_msgs.empty); end -- these iabot bugs perportedly fixed; removing these causes lua script error --[[ -- at Template:Webarchive/testcases/Production; resolve that before deleting these tests if mw.ustring.find( url, "https://web.http", 1, true ) then -- track bug - TODO: IAbot bug; not known if the bug has been fixed; deferred track[categories.error] = 1; return inlineError (data.crit_err_msgs.iabot1); end if url == "https://web.archive.org/http:/" then -- track bug - TODO: IAbot bug; not known if the bug has been fixed; deferred track[categories.error] = 1; return inlineError (data.crit_err_msgs.iabot2); end ]] if not (url:lower():find ('^http') or url:find ('^//')) then return inlineError (data.crit_err_msgs.invalid_url ); end ulx.url1 = {} ulx.url1.url = url ulx.url1.extraurls = parseExtraArgs(args) local good = false; good, uri = pcall (mw.uri.new, ulx.url1.url); -- get a table of uri parts from this url; protected mode to prevent lua error when ulx.url1.url is malformed if not good or nil == uri.host then -- abandon when ulx.url1.url is malformed return inlineError (data.crit_err_msgs.invalid_url); end serviceName(uri.host, args.nolink) if args.date and args.date1 then -- Date argument return inlineError (data.crit_err_msgs.conflicting, {origin.date, origin.date1}); end date = args.date or args.date1; date = date and date:gsub (' +', ' '); -- replace multiple spaces with a single space if date and config.verifydates then if '*' == date then date = 'index'; ldf = 'iso'; -- set to default format elseif 'mdy' == date then date = nil; -- if date extracted from URL, ldf = 'mdy'; -- then |date=mdy overrides iso elseif 'dmy' == date then date = nil; -- if date extracted from URL, ldf = 'dmy'; -- then |date=dmy overrides iso elseif 'ymd' == date then date = nil; -- if date extracted from URL, ldf = 'ymd'; -- then |date=ymd overrides iso else date, ldf = decode_date (date); -- get an iso format date from date and get date's original format end end if 'wayback' == ulx.url1.service or 'locwebarchives' == ulx.url1.service or 'ukgwa' == ulx.url1.service then if date then if config.verifydates then if ldf then udate, msg = decodeWaybackDate (uri.path); -- get the url date in iso format and format of date in |date=; 'index' when wayback url date is * if not udate then -- this is the only 'fatal' error return return inlineError (data.crit_err_msgs[msg]); end if udate ~= date then -- date comparison using iso format dates date = udate; msg = table.concat ({ inlineRed (err_warn_msgs.mismatch, 'warning'), -- add warning message msg, -- add message if there is one }); end end end else -- no |date= udate, msg = decodeWaybackDate (uri.path); if not udate then -- this is the only 'fatal' error return return inlineError (data.crit_err_msgs[msg]); end if == udate then date = nil; -- unset else date = udate; end end elseif 'webcite' == ulx.url1.service then if date then if config.verifydates then if ldf then udate = decodeWebciteDate (uri.path); -- get the url date in iso format if 'query' ~= udate then -- skip if query if udate ~= date then -- date comparison using iso format dates date = udate; msg = table.concat ({ inlineRed (err_warn_msgs.mismatch, 'warning'), }); end end end end else date = decodeWebciteDate( uri.path, "iso" ) if date == "query" then date = nil; -- unset msg = inlineRed (err_warn_msgs.date_miss, 'warning'); elseif not date then -- invalid base62 string date = inlineRed (err_warn_msgs.date1, 'error'); end end elseif 'archiveis' == ulx.url1.service then if date then if config.verifydates then if ldf then udate, msg = decodeArchiveisDate (uri.path) -- get the url date in iso format if 'short link' ~= udate then -- skip if short link if udate ~= date then -- date comparison using iso format dates date = udate; msg = table.concat ({ inlineRed (err_warn_msgs.mismatch, 'warning'), -- add warning message msg, -- add message if there is one }); end end end end else -- no |date= udate, msg = decodeArchiveisDate( uri.path, "iso" ) if udate == "short link" then date = nil; -- unset msg = inlineRed (err_warn_msgs.date_miss, 'warning'); elseif == udate then date = nil; -- unset else date = udate; end end else -- some other service if not date then msg = inlineRed (err_warn_msgs.date_miss, 'warning'); end end if 'index' == date then ulx.url1.date = date .. (msg or ); -- create index + message (if there is one) elseif date then ulx.url1.date = makeDate (date, nil, nil, ldf) .. (msg or ); -- create a date in the wiki's local language + message (if there is one) else ulx.url1.date = msg; end format = args.format; -- Format argument if not format then format = "none" else for k, v in pairs (data.format_vals) do -- |format= accepts two specific values loop through a table of those values local found; -- declare a nil flag for _, p in ipairs (v) do -- loop through local language variants if format == p then -- when |format= value matches format = k; -- use name from table key found = true; -- declare found so that we can break out of outer for loop break; -- break out of inner for loop end end if found then break; end end if format == "addlpages" then if not ulx.url1.date then format = "none" end elseif format == "addlarchives" then format = "addlarchives" else format = "none" end end ulx.url1.format = format if args.title and args.title1 then -- Title argument return inlineError (data.crit_err_msgs.conflicting, {origin.title, origin.title1}); end ulx.url1.title = args.title or args.title1; local rend = createRendering() if not rend then return inlineError (data.crit_err_msgs.unknown); end return rend .. ((unnamed_params and inlineRed (err_warn_msgs.unnamed_params, 'warning')) or ) .. createTracking(); end --[[--------------------------< E X P O R T E D F U N C T I O N S >------------------------------------------ ]] return {webarchive = webarchive};. Indiewire. Retrieved August 15, 2017.
  71. ↑ "AFI's 100 Years... 100 Movies" --[[ ---------------------------------- Lua module implementing the Template loop detected: Template:Webarchive template. A merger of the functionality of three templates: Template:Wayback, Template:Webcite and Template:Cite archives ]] --[[--------------------------< D E P E N D E N C I E S >------------------------------------------------------ ]] require('Module:No globals'); local getArgs = require ('Module:Arguments').getArgs; --[[--------------------------< F O R W A R D D E C L A R A T I O N S >-------------------------------------- ]] local categories = {}; -- category names local config = {}; -- global configuration settings local digits = {}; -- for i18n; table that translates local-wiki digits to western digits local err_warn_msgs = {}; -- error and warning messages local excepted_pages = {}; local month_num = {}; -- for i18n; table that translates local-wiki month names to western digits local prefixes = {}; -- service provider tail string prefixes local services = {}; -- archive service provider data from local s_text = {}; -- table of static text strings used to build final rendering local uncategorized_namespaces = {}; -- list of namespaces that we should not categorize local uncategorized_subpages = {}; -- list of subpages that should not be categorized --[[--------------------------< P A G E S C O P E I D E N T I F I E R S >---------------------------------- ]] local non_western_digits; -- boolean flag set true when data.digits.enable is true local this_page = mw.title.getCurrentTitle(); local track = {}; -- Associative array to hold tracking categories local ulx = {}; -- Associative array to hold template data --[[--------------------------< S U B S T I T U T E >---------------------------------------------------------- Populates numbered arguments in a message string using an argument table. ]] local function substitute (msg, args) return args and mw.message.newRawMessage (msg, args):plain() or msg; end --[[--------------------------< tableLength >----------------------- Given a 1-D table, return number of elements ]] local function tableLength(T) local count = 0 for _ in pairs(T) do count = count + 1 end return count end --[=[-------------------------< M A K E _ W I K I L I N K >---------------------------------------------------- Makes a wikilink; when both link and display text is provided, returns a wikilink in the form D; if only link is provided, returns a wikilink in the form L; if neither are provided or link is omitted, returns an empty string. ]=] local function make_wikilink (link, display, no_link) if nil == no_link then if link and ( ~= link) then if display and ( ~= display) then return table.concat ({'', display, ''}); else return table.concat ({'', link, ''}); end end return display or ; -- link not set so return the display text else -- no_link if display and ( ~= display) then -- if there is display text return display; -- return that else return link or ; -- return the target article name or empty string end end end --[[--------------------------< createTracking >----------------------- Return data in track[] ie. tracking categories ]] local function createTracking() if not excepted_pages[this_page.fullText] then -- namespace:title/fragment is allowed to be categorized (typically this module's / template's testcases page(s)) if uncategorized_namespaces[this_page.nsText] then return ; -- this page not to be categorized so return empty string end for _,v in ipairs (uncategorized_subpages) do -- cycle through page name patterns if this_page.text:match (v) then -- test page name against each pattern return ; -- this subpage type not to be categorized so return empty string end end end local out = {}; if tableLength(track) > 0 then for key, _ in pairs(track) do -- loop through table table.insert (out, make_wikilink (key)); -- and convert category names to links end end return table.concat (out); -- concat into one big string; empty string if table is empty end --[[--------------------------< inlineError >----------------------- Critical error. Render output completely in red. Add to tracking category. This function called as the last thing before abandoning this module ]] local function inlineError (msg, args) track[categories.error] = 1 return table.concat ({ 'Error in ', -- open the error message span config.tname, -- insert the local language template name ' template: ', substitute (msg, args), -- insert the formatted error message '.', -- close the span createTracking() -- add the category }) end --[[--------------------------< inlineRed >----------------------- Render a text fragment in red, such as a warning as part of the final output. Add tracking category. ]] local function inlineRed(msg, trackmsg) if trackmsg == "warning" then track[categories.warning] = 1; elseif trackmsg == "error" then track[categories.error] = 1; end return '' .. msg .. '' end --[[--------------------------< base62 >----------------------- Convert base-62 to base-10 Credit: https://de.wikipedia.org/wiki/Modul:Expr ]] local function base62( value ) local r = 1 -- default return value is input value is malformed if value:match ('%W') then -- value must only be in the set [0-9a-zA-Z] return; -- nil return when value contains extraneous characters end local n = #value -- number of characters in value local k = 1 local c r = 0 for i = n, 1, -1 do -- loop through all characters in value from ls digit to ms digit c = value:byte( i, i ) if c >= 48 and c <= 57 then -- character is digit 0-9 c = c - 48 elseif c >= 65 and c <= 90 then -- character is ascii a-z c = c - 55 else -- must be ascii A-Z c = c - 61 end r = r + c * k -- accumulate this base62 character's value k = k * 62 -- bump for next end -- for i return r end --[[--------------------------< D E C O D E _ D A T E >-------------------------------------------------------- Given a date string, return it in iso format along with an indicator of the date's format. Except that month names must be recognizable as legitimate month names with proper capitalization, and that the date string must match one of the recognized date formats, no error checking is done here; return nil else ]] local function decode_date (date_str) local patterns = { ['dmy'] = {'^(%d%d?) +([^%s%d]+) +(%d%d%d%d)$', 'd', 'm', 'y'}, -- %a does not recognize unicode combining characters used by some languages ['mdy'] = {'^([^%s%d]+) (%d%d?), +(%d%d%d%d)$', 'm', 'd', 'y'}, ['ymd'] = {'^(%d%d%d%d) +([^%s%d]+) (%d%d?)$', 'y', 'm', 'd'}, -- not mos compliant at en.wiki but may be acceptible at other wikis }; local t = {}; if non_western_digits then -- this wiki uses non-western digits? date_str = mw.ustring.gsub (date_str, '%d', digits); -- convert this wiki's non-western digits to western digits end if date_str:match ('^%d%d%d%d%-%d%d%-%d%d$') then -- already an iso format date, return western digits form return date_str, 'iso'; end for k, v in pairs (patterns) do local c1, c2, c3 = mw.ustring.match (date_str, patterns[k][1]); -- c1 .. c3 are captured but we don't know what they hold if c1 then -- set on match t = { -- translate unspecified captures to y, m, and d [patterns[k][2]] = c1, -- fill the table of captures with the captures [patterns[k][3]] = c2, -- take index names from src_pattern table and assign sequential captures [patterns[k][4]] = c3, }; if month_num[t.m] then -- when month not already a number t.m = month_num[t.m]; -- replace valid month name with a number else return nil, 'iso'; -- not a valid date form because month not valid end return mw.ustring.format ('%.4d-%.2d-%.2d', t.y, t.m, t.d), k; -- return date in iso format end end return nil, 'iso'; -- date could not be decoded; return nil and default iso date end --[[--------------------------< makeDate >----------------------- Given year, month, day numbers, (zero-padded or not) return a full date in df format where df may be one of: mdy, dmy, iso, ymd on entry, year, month, day are presumed to be correct for the date that they represent; all are required in this module, makeDate() is sometimes given an iso-format date in year: makeDate (2018-09-20, nil, nil, df) this works because table.concat() sees only one table member ]] local function makeDate (year, month, day, df) local format = { ['dmy'] = 'j F Y', ['mdy'] = 'F j, Y', ['ymd'] = 'Y F j', ['iso'] = 'Y-m-d', }; local date = table.concat ({year, month, day}, '-'); -- assemble year-initial numeric-format date (zero padding not required here) if non_western_digits then -- this wiki uses non-western digits? date = mw.ustring.gsub (date, '%d', digits); -- convert this wiki's non-western digits to western digits end return mw.getContentLanguage():formatDate (format[df], date); end --[[--------------------------< I S _ V A L I D _ D A T E >---------------------------------------------------- Returns true if date is after 31 December 1899 (why is 1900 the min year? shouldn't the internet's date-of-birth be min year?), not after today's date, and represents a valid date (29 February 2017 is not a valid date). Applies Gregorian leapyear rules. all arguments are required ]] local function is_valid_date (year, month, day) local days_in_month = {31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31}; local month_length; local y, m, d; local today = os.date ('*t'); -- fetch a table of current date parts if not year or == year or not month or == month or not day or == day then return false; -- something missing end y = tonumber (year); m = tonumber (month); d = tonumber (day); if 1900 > y or today.year < y or 1 > m or 12 < m then -- year and month are within bounds TODO: 1900? return false; end if (2==m) then -- if February month_length = 28; -- then 28 days unless if (0==(y%4) and (0~=(y%100) or 0==(y%400))) then -- is a leap year? month_length = 29; -- if leap year then 29 days in February end else month_length=days_in_month[m]; end if 1 > d or month_length < d then -- day is within bounds return false; end -- here when date parts represent a valid date return os.time({['year']=y, ['month']=m, ['day']=d, ['hour']=0}) <= os.time(); -- date at midnight must be less than or equal to current date/time end --[[--------------------------< decodeWebciteDate >----------------------- Given a URI-path to Webcite (eg. /67xHmVFWP) return the encoded date in df format returns date string in df format - webcite date is a unix timestamp encoded as bae62 or the string 'query' ]] local function decodeWebciteDate(path, df) local dt = {}; local decode; dt = mw.text.split(path, "/") -- valid URL formats that are not base62 -- http://www.webcitation.org/query?id=1138911916587475 -- http://www.webcitation.org/query?url=http..&date=2012-06-01+21:40:03 -- http://www.webcitation.org/1138911916587475 -- http://www.webcitation.org/cache/73e53dd1f16cf8c5da298418d2a6e452870cf50e -- http://www.webcitation.org/getfile.php?fileid=1c46e791d68e89e12d0c2532cc3cf629b8bc8c8e if dt[2]:find ('query', 1, true) or dt[2]:find ('cache', 1, true) or dt[2]:find ('getfile', 1, true) or tonumber(dt[2]) then return 'query'; end decode = base62(dt[2]); -- base62 string -> exponential number if not decode then return nil; -- nil return when dt[2] contains characters not in %w end dt = os.date('*t', string.format("%d", decode):sub(1,10)) -- exponential number -> text -> first 10 characters (a unix timestamp) -> a table of date parts decode = makeDate (dt.year, dt.month, dt.day, 'iso'); -- date comparisons are all done in iso format with western digits if non_western_digits then -- this wiki uses non-western digits? decode = mw.ustring.gsub (decode, '%d', digits); -- convert this wiki's non-western digits to western digits end return decode; end --[[--------------------------< decodeWaybackDate >----------------------- Given a URI-path to Wayback (eg. /web/20160901010101/http://example.com ) or Library of Congress Web Archives (eg. /all/20160901010101/http://example.com) or UK Government Web Archive (eg. /ukgwa/20160901010101/http://example.com or /tna/20160901010101/http://example.com) return the formatted date eg. "September 1, 2016" in df format Handle non-digits in snapshot ID such as "re_" and "-" and "*" returns two values: first value is one of these: valid date string in df format - wayback date is valid (including the text string 'index' when date is '/*/') empty string - wayback date is malformed (less than 8 digits, not a valid date) nil - wayback date is '/save/' or otherwise not a number second return value is an appropriate 'message' may or may not be formatted ]] local function decodeWaybackDate(path, df) local msg, snapdate; snapdate = path:gsub ('^/web/', ):gsub ('^/all/', ):gsub ('^/ukgwa/', ):gsub ('^/tna/', ):gsub ('^/', ); -- remove leading /web/, /all/, /ukgwa/, /tna/, or / snapdate = snapdate:match ('^[^/]+'); -- get timestamp if snapdate == "*" then -- eg. /web/*/http.., etc. return 'index'; -- return indicator that this url has an index date end snapdate = snapdate:gsub ('%a%a_%d?$', ):gsub ('%-', ); -- from date, remove any trailing "re_", dashes msg = ; if snapdate:match ('%*$') then -- a trailing '*' causes calendar display at archive .org snapdate = snapdate:gsub ('%*$', ); -- remove so not part of length calc later msg = inlineRed (err_warn_msgs.ts_cal, 'warning'); -- make a message end if not tonumber(snapdate) then return nil, 'ts_nan'; -- return nil (fatal error flag) and message selector end local dlen = snapdate:len(); if dlen < 8 then -- we need 8 digits TODO: but shouldn't this be testing for 14 digits? return , inlineRed (err_warn_msgs.ts_short, 'error'); -- return empty string and error message end local year, month, day = snapdate:match ('(%d%d%d%d)(%d%d)(%d%d)'); -- no need for snapdatelong here if not is_valid_date (year, month, day) then return , inlineRed (err_warn_msgs.ts_date, 'error'); -- return empty string and error message end snapdate = table.concat ({year, month, day}, '-'); -- date comparisons are all done in iso format if 14 == dlen then return snapdate, msg; -- return date with message if any else return snapdate, msg .. inlineRed (err_warn_msgs.ts_len, 'warning'); -- return date with warning message(s) end end --[[--------------------------< decodeArchiveisDate >----------------------- Given an Archive.is "long link" URI-path (e.g. /2016.08.28-144552/http://example.com) return the date in df format (e.g. if df = dmy, return 28 August 2016) Handles "." and "-" in snapshot date, so 2016.08.28-144552 is same as 20160828144552 returns two values: first value is one of these: valid date string in df format - archive.is date is valid (including the text string 'short link' when url is the short form) empty string - wayback date is malformed (not a number, less than 8 digits, not a valid date) nil - wayback date is '/save/' second return value is an appropriate 'message' may or may not be formatted ]] local function decodeArchiveisDate(path, df) local snapdate if path:match ('^/%w+$') then -- short form url path is '/' followed by some number of base 62 digits and nothing else return "short link" -- e.g. http://archive.is/hD1qz end snapdate = mw.text.split (path, '/')[2]:gsub('[%.%-]', ); -- get snapshot date, e.g. 2016.08.28-144552; remove periods and hyphens local dlen = string.len(snapdate) if dlen < 8 then -- we need 8 digits TODO: but shouldn't this be testing for 14 digits? return , inlineRed (err_warn_msgs.ts_short, 'error'); -- return empty string and error message end local year, month, day = snapdate:match ('(%d%d%d%d)(%d%d)(%d%d)'); -- no need for snapdatelong here if not is_valid_date (year, month, day) then return , inlineRed (err_warn_msgs.ts_date, 'error'); -- return empty string and error message end snapdate = table.concat ({year, month, day}, '-'); -- date comparisons are all done in iso format if 14 == dlen then return snapdate; -- return date else return snapdate, inlineRed (err_warn_msgs.ts_len, 'warning'); -- return date with warning message end end --[[--------------------------< serviceName >----------------------- Given a domain extracted by mw.uri.new() (eg. web.archive.org) set tail string and service ID ]] local function serviceName(host, no_link) local tracking; local index; host = host:lower():gsub ('^web%.(.+)', '%1'):gsub ('^www%.(.+)', '%1'); -- lowercase, remove web. and www. subdomains if services[host] then index = host; else for k, _ in pairs (services) do if host:find ('%f[%a]'..k:gsub ('([%.%-])', '%%%1')) then index = k; break; end end end if index then local out = {}; -- empty string in [1] so that concatenated result has leading single space ulx.url1.service = services[index][4] or 'other'; tracking = services[index][5] or categories.other; -- build tail string if false == services[index][1] then -- select prefix table.insert (out, prefixes.at); elseif true == services[index][1] then table.insert (out, prefixes.atthe); else table.insert (out, services[index][1]); end table.insert (out, make_wikilink (services[index][2], services[index][3], no_link)); -- add article wikilink if services[index][6] then -- add tail postfix if it exists table.insert (out, services[index][6]); end ulx.url1.tail = table.concat (out, ' '); -- put it all together; result has leading space character else -- here when unknown archive ulx.url1.service = 'other'; tracking = categories.unknown; ulx.url1.tail = table.concat ({, prefixes.at, host, inlineRed (err_warn_msgs.unknown_url, error)}, ' '); end track[tracking] = 1 end --[[--------------------------< parseExtraArgs >----------------------- Parse numbered arguments starting at 2, such as url2..url10, date2..date10, title2..title10 For example: Template loop detected: Template:Webarchive Three url arguments not in numeric sequence (1..4..7). Function only processes arguments numbered 2 or greater (in this case 4 and 7) It creates numeric sequenced table entries like: urlx.url2.url = <argument value for url4> urlx.url3.url = <argument value for url7> Returns the number of URL arguments found numbered 2 or greater (in this case returns "2") ]] local function parseExtraArgs(args) local i, j, argurl, argurl2, argdate, argtitle j = 2 for i = 2, config.maxurls do argurl = "url" .. i if args[argurl] then argurl2 = "url" .. j ulx[argurl2] = {} ulx[argurl2]["url"] = args[argurl] argdate = "date" .. i if args[argdate] then ulx[argurl2]["date"] = args[argdate] else ulx[argurl2]["date"] = inlineRed (err_warn_msgs.date_miss, 'warning'); end argtitle = "title" .. i if args[argtitle] then ulx[argurl2]["title"] = args[argtitle] else ulx[argurl2]["title"] = nil end j = j + 1 end end if j == 2 then return 0 else return j - 2 end end --[[--------------------------< comma >----------------------- Given a date string, return "," if it's MDY ]] local function comma(date) return (date and date:match ('%a+ +%d%d?(,) +%d%d%d%d')) or ; end --[[--------------------------< createRendering >----------------------- Return a rendering of the data in ulx[][] ]] local function createRendering() local displayfield local out = {}; local index_date, msg = ulx.url1.date:match ('(index)(.*)'); -- when ulx.url1.date extract 'index' text and message text (if there is a message) ulx.url1.date = ulx.url1.date:gsub ('index.*', 'index'); -- remove message if 'none' == ulx.url1.format then -- For Template:Wayback, Template:Webcite table.insert (out, '['); -- open extlink markup table.insert (out, ulx.url1.url); -- add url if ulx.url1.title then table.insert (out, ' ') -- the required space table.insert (out, ulx.url1.title) -- the title table.insert (out, ']'); -- close extlink markup table.insert (out, ulx.url1.tail); -- tail text if ulx.url1.date then table.insert (out, ' ('); -- open date text; TODO: why the html entity? replace with regular space? table.insert (out, 'index' == ulx.url1.date and s_text.archive or s_text.archived); -- add text table.insert (out, ' '); -- insert a space table.insert (out, ulx.url1.date); -- add date table.insert (out, ')'); -- close date text end else -- no title if index_date then -- when url date is 'index' table.insert (out, table.concat ({' ', s_text.Archive_index, ']'})); -- add the index link label table.insert (out, msg or ); -- add date mismatch message when url date is /*/ and |date= has valid date else table.insert (out, table.concat ({' ', s_text.Archived, '] '})); -- add link label for url has timestamp date (will include mismatch message if there is one) end if ulx.url1.date then if 'index' ~= ulx.url1.date then table.insert (out, ulx.url1.date); -- add date when data is not 'index' end table.insert (out, comma(ulx.url1.date)); -- add ',' if date format is mdy table.insert (out, ulx.url1.tail); -- add tail text else -- no date table.insert (out, ulx.url1.tail); -- add tail text end end if 0 < ulx.url1.extraurls then -- For multiple archive URLs local tot = ulx.url1.extraurls + 1 table.insert (out, '.') -- terminate first url table.insert (out, table.concat ({' ', s_text.addlarchives, ': '})); -- add header text for i=2, tot do -- loop through the additionals local index = table.concat ({'url', i}); -- make an index displayfield = ulx[index]['title'] and 'title' or 'date'; -- choose display text table.insert (out, '['); -- open extlink markup table.insert (out, ulx[index]['url']); -- add the url table.insert (out, ' '); -- the required space table.insert (out, ulx[index][displayfield]); -- add the label table.insert (out, ']'); -- close extlink markup table.insert (out, i==tot and '.' or ', '); -- add terminator end end return table.concat (out); -- make a big string and done else -- For Template:Cite archives if 'addlarchives' == ulx.url1.format then -- Multiple archive services table.insert (out, table.concat ({s_text.addlarchives, ': '})); -- add header text else -- Multiple pages from the same archive table.insert (out, table.concat ({s_text.addlpages, ' '})); -- add header text table.insert (out, ulx.url1.date); -- add date to header text table.insert (out, ': '); -- close header text end local tot = ulx.url1.extraurls + 1; for i=1, tot do -- loop through the additionals local index = table.concat ({'url', i}); -- make an index table.insert (out, '['); -- open extlink markup table.insert (out, ulx[index]['url']); -- add url table.insert (out, ' '); -- add required space displayfield = ulx[index]['title']; if 'addlarchives' == ulx.url1.format then if not displayfield then displayfield = ulx[index]['date'] end else -- must be addlpages if not displayfield then displayfield = table.concat ({s_text.Page, ' ', i}); end end table.insert (out, displayfield); -- add title, date, page label text table.insert (out, ']'); -- close extlink markup table.insert (out, (i==tot and '.' or ', ')); -- add terminator end return table.concat (out); -- make a big string and done end end --[[--------------------------< P A R A M E T E R _ N A M E _ X L A T E >-------------------------------------- for internaltionalization, translate local-language parameter names to their English equivalents TODO: return error message if multiple aliases of the same canonical parameter name are found? returns two tables: new_args - holds canonical form parameters and their values either from translation or because the parameter was already in canonical form origin - maps canonical-form parameter names to their untranslated (local language) form for error messaging in the local language unrecognized parameters are ignored ]] local function parameter_name_xlate (args, params, enum_params) local name; -- holds modifiable name of the parameter name during evaluation local enum; -- for enumerated parameters, holds the enumerator during evaluation local found = false; -- flag used to break out of nested for loops local new_args = {}; -- a table that holds canonical and translated parameter k/v pairs local origin = {}; -- a table that maps original (local language) parameter names to their canonical name for local language error messaging local unnamed_params; -- set true when unsupported positional parameters are detected for k, v in pairs (args) do -- loop through all of the arguments in the args table name = k; -- copy of original parameter name if 'string' == type (k) then if non_western_digits then -- true when non-western digits supported at this wiki name = mw.ustring.gsub (name, '%d', digits); -- convert this wiki's non-western digits to western digits end enum = name:match ('%d+$'); -- get parameter enumerator if it exists; nil else if not enum then -- no enumerator so looking for non-enumnerated parameters -- TODO: insert shortcut here? if params[name] then name holds the canonical parameter name; no need to search further for pname, aliases in pairs (params) do -- loop through each parameter the params table for _, alias in ipairs (aliases) do -- loop through each alias in the parameter's aliases table if name == alias then new_args[pname] = v; -- create a new entry in the new_args table origin [pname] = k; -- create an entry to make canonical parameter name to original local language parameter name found = true; -- flag so that we can break out of these nested for loops break; -- no need to search the rest of the aliases table for name so go on to the next k, v pair end end if found then -- true when we found an alias that matched name found = false; -- reset the flag break; -- go do next args k/v pair end end else -- enumerated parameters name = name:gsub ('%d$', '#'); -- replace enumeration digits with place holder for table search -- TODO: insert shortcut here? if num_params[name] then name holds the canonical parameter name; no need to search further for pname, aliases in pairs (enum_params) do -- loop through each parameter the num_params table for _, alias in ipairs (aliases) do -- loop through each alias in the parameter's aliases table if name == alias then pname = pname:gsub ('#$', enum); -- replace the '#' place holder with the actual enumerator new_args[pname] = v; -- create a new entry in the new_args table origin [pname] = k; -- create an entry to make canonical parameter name to original local language parameter name found = true; -- flag so that we can break out of these nested for loops break; -- no need to search the rest of the aliases table for name so go on to the next k, v pair end end if found then -- true when we found an alias that matched name found = false; -- reset the flag break; -- go do next args k/v pair end end end else unnamed_params = true; -- flag for unsupported positional parameters end end -- for k, v return new_args, origin, unnamed_params; end --[[--------------------------< W E B A R C H I V E >---------------------------------------------------------- template entry point ]] local function webarchive(frame) local args = getArgs (frame); local data = mw.loadData (table.concat ({ -- make a data module name; sandbox or live 'Module:Webarchive/data', frame:getTitle():find('sandbox', 1, true) and '/sandbox' or -- this instance is ./sandbox then append /sandbox })); categories = data.categories; -- fill in the forward declarations config = data.config; if data.digits.enable then digits = data.digits; -- for i18n; table of digits in the local wiki's language non_western_digits = true; -- use_non_western_digits end err_warn_msgs = data.err_warn_msgs; excepted_pages = data.excepted_pages; month_num = data.month_num; -- for i18n; table of month names in the local wiki's language prefixes = data.prefixes; services = data.services; s_text = data.s_text; uncategorized_namespaces = data.uncategorized_namespaces; uncategorized_subpages = data.uncategorized_subpages; local origin = {}; -- holds a map of English to local language parameter names used in the current template; not currently used local unnamed_params; -- boolean set to true when template call has unnamed parameters args, origin, unnamed_params = parameter_name_xlate (args, data.params, data.enum_params); -- translate parameter names in args to English local date, format, msg, udate, uri, url; local ldf = 'iso'; -- when there is no |date= parameter, render url dates in iso format if args.url and args.url1 then -- URL argument (first) return inlineError (data.crit_err_msgs.conflicting, {origin.url, origin.url1}); end url = args.url or args.url1; if not url then return inlineError (data.crit_err_msgs.empty); end -- these iabot bugs perportedly fixed; removing these causes lua script error --[[ -- at Template:Webarchive/testcases/Production; resolve that before deleting these tests if mw.ustring.find( url, "https://web.http", 1, true ) then -- track bug - TODO: IAbot bug; not known if the bug has been fixed; deferred track[categories.error] = 1; return inlineError (data.crit_err_msgs.iabot1); end if url == "https://web.archive.org/http:/" then -- track bug - TODO: IAbot bug; not known if the bug has been fixed; deferred track[categories.error] = 1; return inlineError (data.crit_err_msgs.iabot2); end ]] if not (url:lower():find ('^http') or url:find ('^//')) then return inlineError (data.crit_err_msgs.invalid_url ); end ulx.url1 = {} ulx.url1.url = url ulx.url1.extraurls = parseExtraArgs(args) local good = false; good, uri = pcall (mw.uri.new, ulx.url1.url); -- get a table of uri parts from this url; protected mode to prevent lua error when ulx.url1.url is malformed if not good or nil == uri.host then -- abandon when ulx.url1.url is malformed return inlineError (data.crit_err_msgs.invalid_url); end serviceName(uri.host, args.nolink) if args.date and args.date1 then -- Date argument return inlineError (data.crit_err_msgs.conflicting, {origin.date, origin.date1}); end date = args.date or args.date1; date = date and date:gsub (' +', ' '); -- replace multiple spaces with a single space if date and config.verifydates then if '*' == date then date = 'index'; ldf = 'iso'; -- set to default format elseif 'mdy' == date then date = nil; -- if date extracted from URL, ldf = 'mdy'; -- then |date=mdy overrides iso elseif 'dmy' == date then date = nil; -- if date extracted from URL, ldf = 'dmy'; -- then |date=dmy overrides iso elseif 'ymd' == date then date = nil; -- if date extracted from URL, ldf = 'ymd'; -- then |date=ymd overrides iso else date, ldf = decode_date (date); -- get an iso format date from date and get date's original format end end if 'wayback' == ulx.url1.service or 'locwebarchives' == ulx.url1.service or 'ukgwa' == ulx.url1.service then if date then if config.verifydates then if ldf then udate, msg = decodeWaybackDate (uri.path); -- get the url date in iso format and format of date in |date=; 'index' when wayback url date is * if not udate then -- this is the only 'fatal' error return return inlineError (data.crit_err_msgs[msg]); end if udate ~= date then -- date comparison using iso format dates date = udate; msg = table.concat ({ inlineRed (err_warn_msgs.mismatch, 'warning'), -- add warning message msg, -- add message if there is one }); end end end else -- no |date= udate, msg = decodeWaybackDate (uri.path); if not udate then -- this is the only 'fatal' error return return inlineError (data.crit_err_msgs[msg]); end if == udate then date = nil; -- unset else date = udate; end end elseif 'webcite' == ulx.url1.service then if date then if config.verifydates then if ldf then udate = decodeWebciteDate (uri.path); -- get the url date in iso format if 'query' ~= udate then -- skip if query if udate ~= date then -- date comparison using iso format dates date = udate; msg = table.concat ({ inlineRed (err_warn_msgs.mismatch, 'warning'), }); end end end end else date = decodeWebciteDate( uri.path, "iso" ) if date == "query" then date = nil; -- unset msg = inlineRed (err_warn_msgs.date_miss, 'warning'); elseif not date then -- invalid base62 string date = inlineRed (err_warn_msgs.date1, 'error'); end end elseif 'archiveis' == ulx.url1.service then if date then if config.verifydates then if ldf then udate, msg = decodeArchiveisDate (uri.path) -- get the url date in iso format if 'short link' ~= udate then -- skip if short link if udate ~= date then -- date comparison using iso format dates date = udate; msg = table.concat ({ inlineRed (err_warn_msgs.mismatch, 'warning'), -- add warning message msg, -- add message if there is one }); end end end end else -- no |date= udate, msg = decodeArchiveisDate( uri.path, "iso" ) if udate == "short link" then date = nil; -- unset msg = inlineRed (err_warn_msgs.date_miss, 'warning'); elseif == udate then date = nil; -- unset else date = udate; end end else -- some other service if not date then msg = inlineRed (err_warn_msgs.date_miss, 'warning'); end end if 'index' == date then ulx.url1.date = date .. (msg or ); -- create index + message (if there is one) elseif date then ulx.url1.date = makeDate (date, nil, nil, ldf) .. (msg or ); -- create a date in the wiki's local language + message (if there is one) else ulx.url1.date = msg; end format = args.format; -- Format argument if not format then format = "none" else for k, v in pairs (data.format_vals) do -- |format= accepts two specific values loop through a table of those values local found; -- declare a nil flag for _, p in ipairs (v) do -- loop through local language variants if format == p then -- when |format= value matches format = k; -- use name from table key found = true; -- declare found so that we can break out of outer for loop break; -- break out of inner for loop end end if found then break; end end if format == "addlpages" then if not ulx.url1.date then format = "none" end elseif format == "addlarchives" then format = "addlarchives" else format = "none" end end ulx.url1.format = format if args.title and args.title1 then -- Title argument return inlineError (data.crit_err_msgs.conflicting, {origin.title, origin.title1}); end ulx.url1.title = args.title or args.title1; local rend = createRendering() if not rend then return inlineError (data.crit_err_msgs.unknown); end return rend .. ((unnamed_params and inlineRed (err_warn_msgs.unnamed_params, 'warning')) or ) .. createTracking(); end --[[--------------------------< E X P O R T E D F U N C T I O N S >------------------------------------------ ]] return {webarchive = webarchive};. American Film Institute. Retrieved October 23, 2010.
  72. ↑ "AFI's 100 Years... 100 Thrills" --[[ ---------------------------------- Lua module implementing the Template loop detected: Template:Webarchive template. A merger of the functionality of three templates: Template:Wayback, Template:Webcite and Template:Cite archives ]] --[[--------------------------< D E P E N D E N C I E S >------------------------------------------------------ ]] require('Module:No globals'); local getArgs = require ('Module:Arguments').getArgs; --[[--------------------------< F O R W A R D D E C L A R A T I O N S >-------------------------------------- ]] local categories = {}; -- category names local config = {}; -- global configuration settings local digits = {}; -- for i18n; table that translates local-wiki digits to western digits local err_warn_msgs = {}; -- error and warning messages local excepted_pages = {}; local month_num = {}; -- for i18n; table that translates local-wiki month names to western digits local prefixes = {}; -- service provider tail string prefixes local services = {}; -- archive service provider data from local s_text = {}; -- table of static text strings used to build final rendering local uncategorized_namespaces = {}; -- list of namespaces that we should not categorize local uncategorized_subpages = {}; -- list of subpages that should not be categorized --[[--------------------------< P A G E S C O P E I D E N T I F I E R S >---------------------------------- ]] local non_western_digits; -- boolean flag set true when data.digits.enable is true local this_page = mw.title.getCurrentTitle(); local track = {}; -- Associative array to hold tracking categories local ulx = {}; -- Associative array to hold template data --[[--------------------------< S U B S T I T U T E >---------------------------------------------------------- Populates numbered arguments in a message string using an argument table. ]] local function substitute (msg, args) return args and mw.message.newRawMessage (msg, args):plain() or msg; end --[[--------------------------< tableLength >----------------------- Given a 1-D table, return number of elements ]] local function tableLength(T) local count = 0 for _ in pairs(T) do count = count + 1 end return count end --[=[-------------------------< M A K E _ W I K I L I N K >---------------------------------------------------- Makes a wikilink; when both link and display text is provided, returns a wikilink in the form D; if only link is provided, returns a wikilink in the form L; if neither are provided or link is omitted, returns an empty string. ]=] local function make_wikilink (link, display, no_link) if nil == no_link then if link and ( ~= link) then if display and ( ~= display) then return table.concat ({'', display, ''}); else return table.concat ({'', link, ''}); end end return display or ; -- link not set so return the display text else -- no_link if display and ( ~= display) then -- if there is display text return display; -- return that else return link or ; -- return the target article name or empty string end end end --[[--------------------------< createTracking >----------------------- Return data in track[] ie. tracking categories ]] local function createTracking() if not excepted_pages[this_page.fullText] then -- namespace:title/fragment is allowed to be categorized (typically this module's / template's testcases page(s)) if uncategorized_namespaces[this_page.nsText] then return ; -- this page not to be categorized so return empty string end for _,v in ipairs (uncategorized_subpages) do -- cycle through page name patterns if this_page.text:match (v) then -- test page name against each pattern return ; -- this subpage type not to be categorized so return empty string end end end local out = {}; if tableLength(track) > 0 then for key, _ in pairs(track) do -- loop through table table.insert (out, make_wikilink (key)); -- and convert category names to links end end return table.concat (out); -- concat into one big string; empty string if table is empty end --[[--------------------------< inlineError >----------------------- Critical error. Render output completely in red. Add to tracking category. This function called as the last thing before abandoning this module ]] local function inlineError (msg, args) track[categories.error] = 1 return table.concat ({ 'Error in ', -- open the error message span config.tname, -- insert the local language template name ' template: ', substitute (msg, args), -- insert the formatted error message '.', -- close the span createTracking() -- add the category }) end --[[--------------------------< inlineRed >----------------------- Render a text fragment in red, such as a warning as part of the final output. Add tracking category. ]] local function inlineRed(msg, trackmsg) if trackmsg == "warning" then track[categories.warning] = 1; elseif trackmsg == "error" then track[categories.error] = 1; end return '' .. msg .. '' end --[[--------------------------< base62 >----------------------- Convert base-62 to base-10 Credit: https://de.wikipedia.org/wiki/Modul:Expr ]] local function base62( value ) local r = 1 -- default return value is input value is malformed if value:match ('%W') then -- value must only be in the set [0-9a-zA-Z] return; -- nil return when value contains extraneous characters end local n = #value -- number of characters in value local k = 1 local c r = 0 for i = n, 1, -1 do -- loop through all characters in value from ls digit to ms digit c = value:byte( i, i ) if c >= 48 and c <= 57 then -- character is digit 0-9 c = c - 48 elseif c >= 65 and c <= 90 then -- character is ascii a-z c = c - 55 else -- must be ascii A-Z c = c - 61 end r = r + c * k -- accumulate this base62 character's value k = k * 62 -- bump for next end -- for i return r end --[[--------------------------< D E C O D E _ D A T E >-------------------------------------------------------- Given a date string, return it in iso format along with an indicator of the date's format. Except that month names must be recognizable as legitimate month names with proper capitalization, and that the date string must match one of the recognized date formats, no error checking is done here; return nil else ]] local function decode_date (date_str) local patterns = { ['dmy'] = {'^(%d%d?) +([^%s%d]+) +(%d%d%d%d)$', 'd', 'm', 'y'}, -- %a does not recognize unicode combining characters used by some languages ['mdy'] = {'^([^%s%d]+) (%d%d?), +(%d%d%d%d)$', 'm', 'd', 'y'}, ['ymd'] = {'^(%d%d%d%d) +([^%s%d]+) (%d%d?)$', 'y', 'm', 'd'}, -- not mos compliant at en.wiki but may be acceptible at other wikis }; local t = {}; if non_western_digits then -- this wiki uses non-western digits? date_str = mw.ustring.gsub (date_str, '%d', digits); -- convert this wiki's non-western digits to western digits end if date_str:match ('^%d%d%d%d%-%d%d%-%d%d$') then -- already an iso format date, return western digits form return date_str, 'iso'; end for k, v in pairs (patterns) do local c1, c2, c3 = mw.ustring.match (date_str, patterns[k][1]); -- c1 .. c3 are captured but we don't know what they hold if c1 then -- set on match t = { -- translate unspecified captures to y, m, and d [patterns[k][2]] = c1, -- fill the table of captures with the captures [patterns[k][3]] = c2, -- take index names from src_pattern table and assign sequential captures [patterns[k][4]] = c3, }; if month_num[t.m] then -- when month not already a number t.m = month_num[t.m]; -- replace valid month name with a number else return nil, 'iso'; -- not a valid date form because month not valid end return mw.ustring.format ('%.4d-%.2d-%.2d', t.y, t.m, t.d), k; -- return date in iso format end end return nil, 'iso'; -- date could not be decoded; return nil and default iso date end --[[--------------------------< makeDate >----------------------- Given year, month, day numbers, (zero-padded or not) return a full date in df format where df may be one of: mdy, dmy, iso, ymd on entry, year, month, day are presumed to be correct for the date that they represent; all are required in this module, makeDate() is sometimes given an iso-format date in year: makeDate (2018-09-20, nil, nil, df) this works because table.concat() sees only one table member ]] local function makeDate (year, month, day, df) local format = { ['dmy'] = 'j F Y', ['mdy'] = 'F j, Y', ['ymd'] = 'Y F j', ['iso'] = 'Y-m-d', }; local date = table.concat ({year, month, day}, '-'); -- assemble year-initial numeric-format date (zero padding not required here) if non_western_digits then -- this wiki uses non-western digits? date = mw.ustring.gsub (date, '%d', digits); -- convert this wiki's non-western digits to western digits end return mw.getContentLanguage():formatDate (format[df], date); end --[[--------------------------< I S _ V A L I D _ D A T E >---------------------------------------------------- Returns true if date is after 31 December 1899 (why is 1900 the min year? shouldn't the internet's date-of-birth be min year?), not after today's date, and represents a valid date (29 February 2017 is not a valid date). Applies Gregorian leapyear rules. all arguments are required ]] local function is_valid_date (year, month, day) local days_in_month = {31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31}; local month_length; local y, m, d; local today = os.date ('*t'); -- fetch a table of current date parts if not year or == year or not month or == month or not day or == day then return false; -- something missing end y = tonumber (year); m = tonumber (month); d = tonumber (day); if 1900 > y or today.year < y or 1 > m or 12 < m then -- year and month are within bounds TODO: 1900? return false; end if (2==m) then -- if February month_length = 28; -- then 28 days unless if (0==(y%4) and (0~=(y%100) or 0==(y%400))) then -- is a leap year? month_length = 29; -- if leap year then 29 days in February end else month_length=days_in_month[m]; end if 1 > d or month_length < d then -- day is within bounds return false; end -- here when date parts represent a valid date return os.time({['year']=y, ['month']=m, ['day']=d, ['hour']=0}) <= os.time(); -- date at midnight must be less than or equal to current date/time end --[[--------------------------< decodeWebciteDate >----------------------- Given a URI-path to Webcite (eg. /67xHmVFWP) return the encoded date in df format returns date string in df format - webcite date is a unix timestamp encoded as bae62 or the string 'query' ]] local function decodeWebciteDate(path, df) local dt = {}; local decode; dt = mw.text.split(path, "/") -- valid URL formats that are not base62 -- http://www.webcitation.org/query?id=1138911916587475 -- http://www.webcitation.org/query?url=http..&date=2012-06-01+21:40:03 -- http://www.webcitation.org/1138911916587475 -- http://www.webcitation.org/cache/73e53dd1f16cf8c5da298418d2a6e452870cf50e -- http://www.webcitation.org/getfile.php?fileid=1c46e791d68e89e12d0c2532cc3cf629b8bc8c8e if dt[2]:find ('query', 1, true) or dt[2]:find ('cache', 1, true) or dt[2]:find ('getfile', 1, true) or tonumber(dt[2]) then return 'query'; end decode = base62(dt[2]); -- base62 string -> exponential number if not decode then return nil; -- nil return when dt[2] contains characters not in %w end dt = os.date('*t', string.format("%d", decode):sub(1,10)) -- exponential number -> text -> first 10 characters (a unix timestamp) -> a table of date parts decode = makeDate (dt.year, dt.month, dt.day, 'iso'); -- date comparisons are all done in iso format with western digits if non_western_digits then -- this wiki uses non-western digits? decode = mw.ustring.gsub (decode, '%d', digits); -- convert this wiki's non-western digits to western digits end return decode; end --[[--------------------------< decodeWaybackDate >----------------------- Given a URI-path to Wayback (eg. /web/20160901010101/http://example.com ) or Library of Congress Web Archives (eg. /all/20160901010101/http://example.com) or UK Government Web Archive (eg. /ukgwa/20160901010101/http://example.com or /tna/20160901010101/http://example.com) return the formatted date eg. "September 1, 2016" in df format Handle non-digits in snapshot ID such as "re_" and "-" and "*" returns two values: first value is one of these: valid date string in df format - wayback date is valid (including the text string 'index' when date is '/*/') empty string - wayback date is malformed (less than 8 digits, not a valid date) nil - wayback date is '/save/' or otherwise not a number second return value is an appropriate 'message' may or may not be formatted ]] local function decodeWaybackDate(path, df) local msg, snapdate; snapdate = path:gsub ('^/web/', ):gsub ('^/all/', ):gsub ('^/ukgwa/', ):gsub ('^/tna/', ):gsub ('^/', ); -- remove leading /web/, /all/, /ukgwa/, /tna/, or / snapdate = snapdate:match ('^[^/]+'); -- get timestamp if snapdate == "*" then -- eg. /web/*/http.., etc. return 'index'; -- return indicator that this url has an index date end snapdate = snapdate:gsub ('%a%a_%d?$', ):gsub ('%-', ); -- from date, remove any trailing "re_", dashes msg = ; if snapdate:match ('%*$') then -- a trailing '*' causes calendar display at archive .org snapdate = snapdate:gsub ('%*$', ); -- remove so not part of length calc later msg = inlineRed (err_warn_msgs.ts_cal, 'warning'); -- make a message end if not tonumber(snapdate) then return nil, 'ts_nan'; -- return nil (fatal error flag) and message selector end local dlen = snapdate:len(); if dlen < 8 then -- we need 8 digits TODO: but shouldn't this be testing for 14 digits? return , inlineRed (err_warn_msgs.ts_short, 'error'); -- return empty string and error message end local year, month, day = snapdate:match ('(%d%d%d%d)(%d%d)(%d%d)'); -- no need for snapdatelong here if not is_valid_date (year, month, day) then return , inlineRed (err_warn_msgs.ts_date, 'error'); -- return empty string and error message end snapdate = table.concat ({year, month, day}, '-'); -- date comparisons are all done in iso format if 14 == dlen then return snapdate, msg; -- return date with message if any else return snapdate, msg .. inlineRed (err_warn_msgs.ts_len, 'warning'); -- return date with warning message(s) end end --[[--------------------------< decodeArchiveisDate >----------------------- Given an Archive.is "long link" URI-path (e.g. /2016.08.28-144552/http://example.com) return the date in df format (e.g. if df = dmy, return 28 August 2016) Handles "." and "-" in snapshot date, so 2016.08.28-144552 is same as 20160828144552 returns two values: first value is one of these: valid date string in df format - archive.is date is valid (including the text string 'short link' when url is the short form) empty string - wayback date is malformed (not a number, less than 8 digits, not a valid date) nil - wayback date is '/save/' second return value is an appropriate 'message' may or may not be formatted ]] local function decodeArchiveisDate(path, df) local snapdate if path:match ('^/%w+$') then -- short form url path is '/' followed by some number of base 62 digits and nothing else return "short link" -- e.g. http://archive.is/hD1qz end snapdate = mw.text.split (path, '/')[2]:gsub('[%.%-]', ); -- get snapshot date, e.g. 2016.08.28-144552; remove periods and hyphens local dlen = string.len(snapdate) if dlen < 8 then -- we need 8 digits TODO: but shouldn't this be testing for 14 digits? return , inlineRed (err_warn_msgs.ts_short, 'error'); -- return empty string and error message end local year, month, day = snapdate:match ('(%d%d%d%d)(%d%d)(%d%d)'); -- no need for snapdatelong here if not is_valid_date (year, month, day) then return , inlineRed (err_warn_msgs.ts_date, 'error'); -- return empty string and error message end snapdate = table.concat ({year, month, day}, '-'); -- date comparisons are all done in iso format if 14 == dlen then return snapdate; -- return date else return snapdate, inlineRed (err_warn_msgs.ts_len, 'warning'); -- return date with warning message end end --[[--------------------------< serviceName >----------------------- Given a domain extracted by mw.uri.new() (eg. web.archive.org) set tail string and service ID ]] local function serviceName(host, no_link) local tracking; local index; host = host:lower():gsub ('^web%.(.+)', '%1'):gsub ('^www%.(.+)', '%1'); -- lowercase, remove web. and www. subdomains if services[host] then index = host; else for k, _ in pairs (services) do if host:find ('%f[%a]'..k:gsub ('([%.%-])', '%%%1')) then index = k; break; end end end if index then local out = {}; -- empty string in [1] so that concatenated result has leading single space ulx.url1.service = services[index][4] or 'other'; tracking = services[index][5] or categories.other; -- build tail string if false == services[index][1] then -- select prefix table.insert (out, prefixes.at); elseif true == services[index][1] then table.insert (out, prefixes.atthe); else table.insert (out, services[index][1]); end table.insert (out, make_wikilink (services[index][2], services[index][3], no_link)); -- add article wikilink if services[index][6] then -- add tail postfix if it exists table.insert (out, services[index][6]); end ulx.url1.tail = table.concat (out, ' '); -- put it all together; result has leading space character else -- here when unknown archive ulx.url1.service = 'other'; tracking = categories.unknown; ulx.url1.tail = table.concat ({, prefixes.at, host, inlineRed (err_warn_msgs.unknown_url, error)}, ' '); end track[tracking] = 1 end --[[--------------------------< parseExtraArgs >----------------------- Parse numbered arguments starting at 2, such as url2..url10, date2..date10, title2..title10 For example: Template loop detected: Template:Webarchive Three url arguments not in numeric sequence (1..4..7). Function only processes arguments numbered 2 or greater (in this case 4 and 7) It creates numeric sequenced table entries like: urlx.url2.url = <argument value for url4> urlx.url3.url = <argument value for url7> Returns the number of URL arguments found numbered 2 or greater (in this case returns "2") ]] local function parseExtraArgs(args) local i, j, argurl, argurl2, argdate, argtitle j = 2 for i = 2, config.maxurls do argurl = "url" .. i if args[argurl] then argurl2 = "url" .. j ulx[argurl2] = {} ulx[argurl2]["url"] = args[argurl] argdate = "date" .. i if args[argdate] then ulx[argurl2]["date"] = args[argdate] else ulx[argurl2]["date"] = inlineRed (err_warn_msgs.date_miss, 'warning'); end argtitle = "title" .. i if args[argtitle] then ulx[argurl2]["title"] = args[argtitle] else ulx[argurl2]["title"] = nil end j = j + 1 end end if j == 2 then return 0 else return j - 2 end end --[[--------------------------< comma >----------------------- Given a date string, return "," if it's MDY ]] local function comma(date) return (date and date:match ('%a+ +%d%d?(,) +%d%d%d%d')) or ; end --[[--------------------------< createRendering >----------------------- Return a rendering of the data in ulx[][] ]] local function createRendering() local displayfield local out = {}; local index_date, msg = ulx.url1.date:match ('(index)(.*)'); -- when ulx.url1.date extract 'index' text and message text (if there is a message) ulx.url1.date = ulx.url1.date:gsub ('index.*', 'index'); -- remove message if 'none' == ulx.url1.format then -- For Template:Wayback, Template:Webcite table.insert (out, '['); -- open extlink markup table.insert (out, ulx.url1.url); -- add url if ulx.url1.title then table.insert (out, ' ') -- the required space table.insert (out, ulx.url1.title) -- the title table.insert (out, ']'); -- close extlink markup table.insert (out, ulx.url1.tail); -- tail text if ulx.url1.date then table.insert (out, ' ('); -- open date text; TODO: why the html entity? replace with regular space? table.insert (out, 'index' == ulx.url1.date and s_text.archive or s_text.archived); -- add text table.insert (out, ' '); -- insert a space table.insert (out, ulx.url1.date); -- add date table.insert (out, ')'); -- close date text end else -- no title if index_date then -- when url date is 'index' table.insert (out, table.concat ({' ', s_text.Archive_index, ']'})); -- add the index link label table.insert (out, msg or ); -- add date mismatch message when url date is /*/ and |date= has valid date else table.insert (out, table.concat ({' ', s_text.Archived, '] '})); -- add link label for url has timestamp date (will include mismatch message if there is one) end if ulx.url1.date then if 'index' ~= ulx.url1.date then table.insert (out, ulx.url1.date); -- add date when data is not 'index' end table.insert (out, comma(ulx.url1.date)); -- add ',' if date format is mdy table.insert (out, ulx.url1.tail); -- add tail text else -- no date table.insert (out, ulx.url1.tail); -- add tail text end end if 0 < ulx.url1.extraurls then -- For multiple archive URLs local tot = ulx.url1.extraurls + 1 table.insert (out, '.') -- terminate first url table.insert (out, table.concat ({' ', s_text.addlarchives, ': '})); -- add header text for i=2, tot do -- loop through the additionals local index = table.concat ({'url', i}); -- make an index displayfield = ulx[index]['title'] and 'title' or 'date'; -- choose display text table.insert (out, '['); -- open extlink markup table.insert (out, ulx[index]['url']); -- add the url table.insert (out, ' '); -- the required space table.insert (out, ulx[index][displayfield]); -- add the label table.insert (out, ']'); -- close extlink markup table.insert (out, i==tot and '.' or ', '); -- add terminator end end return table.concat (out); -- make a big string and done else -- For Template:Cite archives if 'addlarchives' == ulx.url1.format then -- Multiple archive services table.insert (out, table.concat ({s_text.addlarchives, ': '})); -- add header text else -- Multiple pages from the same archive table.insert (out, table.concat ({s_text.addlpages, ' '})); -- add header text table.insert (out, ulx.url1.date); -- add date to header text table.insert (out, ': '); -- close header text end local tot = ulx.url1.extraurls + 1; for i=1, tot do -- loop through the additionals local index = table.concat ({'url', i}); -- make an index table.insert (out, '['); -- open extlink markup table.insert (out, ulx[index]['url']); -- add url table.insert (out, ' '); -- add required space displayfield = ulx[index]['title']; if 'addlarchives' == ulx.url1.format then if not displayfield then displayfield = ulx[index]['date'] end else -- must be addlpages if not displayfield then displayfield = table.concat ({s_text.Page, ' ', i}); end end table.insert (out, displayfield); -- add title, date, page label text table.insert (out, ']'); -- close extlink markup table.insert (out, (i==tot and '.' or ', ')); -- add terminator end return table.concat (out); -- make a big string and done end end --[[--------------------------< P A R A M E T E R _ N A M E _ X L A T E >-------------------------------------- for internaltionalization, translate local-language parameter names to their English equivalents TODO: return error message if multiple aliases of the same canonical parameter name are found? returns two tables: new_args - holds canonical form parameters and their values either from translation or because the parameter was already in canonical form origin - maps canonical-form parameter names to their untranslated (local language) form for error messaging in the local language unrecognized parameters are ignored ]] local function parameter_name_xlate (args, params, enum_params) local name; -- holds modifiable name of the parameter name during evaluation local enum; -- for enumerated parameters, holds the enumerator during evaluation local found = false; -- flag used to break out of nested for loops local new_args = {}; -- a table that holds canonical and translated parameter k/v pairs local origin = {}; -- a table that maps original (local language) parameter names to their canonical name for local language error messaging local unnamed_params; -- set true when unsupported positional parameters are detected for k, v in pairs (args) do -- loop through all of the arguments in the args table name = k; -- copy of original parameter name if 'string' == type (k) then if non_western_digits then -- true when non-western digits supported at this wiki name = mw.ustring.gsub (name, '%d', digits); -- convert this wiki's non-western digits to western digits end enum = name:match ('%d+$'); -- get parameter enumerator if it exists; nil else if not enum then -- no enumerator so looking for non-enumnerated parameters -- TODO: insert shortcut here? if params[name] then name holds the canonical parameter name; no need to search further for pname, aliases in pairs (params) do -- loop through each parameter the params table for _, alias in ipairs (aliases) do -- loop through each alias in the parameter's aliases table if name == alias then new_args[pname] = v; -- create a new entry in the new_args table origin [pname] = k; -- create an entry to make canonical parameter name to original local language parameter name found = true; -- flag so that we can break out of these nested for loops break; -- no need to search the rest of the aliases table for name so go on to the next k, v pair end end if found then -- true when we found an alias that matched name found = false; -- reset the flag break; -- go do next args k/v pair end end else -- enumerated parameters name = name:gsub ('%d$', '#'); -- replace enumeration digits with place holder for table search -- TODO: insert shortcut here? if num_params[name] then name holds the canonical parameter name; no need to search further for pname, aliases in pairs (enum_params) do -- loop through each parameter the num_params table for _, alias in ipairs (aliases) do -- loop through each alias in the parameter's aliases table if name == alias then pname = pname:gsub ('#$', enum); -- replace the '#' place holder with the actual enumerator new_args[pname] = v; -- create a new entry in the new_args table origin [pname] = k; -- create an entry to make canonical parameter name to original local language parameter name found = true; -- flag so that we can break out of these nested for loops break; -- no need to search the rest of the aliases table for name so go on to the next k, v pair end end if found then -- true when we found an alias that matched name found = false; -- reset the flag break; -- go do next args k/v pair end end end else unnamed_params = true; -- flag for unsupported positional parameters end end -- for k, v return new_args, origin, unnamed_params; end --[[--------------------------< W E B A R C H I V E >---------------------------------------------------------- template entry point ]] local function webarchive(frame) local args = getArgs (frame); local data = mw.loadData (table.concat ({ -- make a data module name; sandbox or live 'Module:Webarchive/data', frame:getTitle():find('sandbox', 1, true) and '/sandbox' or -- this instance is ./sandbox then append /sandbox })); categories = data.categories; -- fill in the forward declarations config = data.config; if data.digits.enable then digits = data.digits; -- for i18n; table of digits in the local wiki's language non_western_digits = true; -- use_non_western_digits end err_warn_msgs = data.err_warn_msgs; excepted_pages = data.excepted_pages; month_num = data.month_num; -- for i18n; table of month names in the local wiki's language prefixes = data.prefixes; services = data.services; s_text = data.s_text; uncategorized_namespaces = data.uncategorized_namespaces; uncategorized_subpages = data.uncategorized_subpages; local origin = {}; -- holds a map of English to local language parameter names used in the current template; not currently used local unnamed_params; -- boolean set to true when template call has unnamed parameters args, origin, unnamed_params = parameter_name_xlate (args, data.params, data.enum_params); -- translate parameter names in args to English local date, format, msg, udate, uri, url; local ldf = 'iso'; -- when there is no |date= parameter, render url dates in iso format if args.url and args.url1 then -- URL argument (first) return inlineError (data.crit_err_msgs.conflicting, {origin.url, origin.url1}); end url = args.url or args.url1; if not url then return inlineError (data.crit_err_msgs.empty); end -- these iabot bugs perportedly fixed; removing these causes lua script error --[[ -- at Template:Webarchive/testcases/Production; resolve that before deleting these tests if mw.ustring.find( url, "https://web.http", 1, true ) then -- track bug - TODO: IAbot bug; not known if the bug has been fixed; deferred track[categories.error] = 1; return inlineError (data.crit_err_msgs.iabot1); end if url == "https://web.archive.org/http:/" then -- track bug - TODO: IAbot bug; not known if the bug has been fixed; deferred track[categories.error] = 1; return inlineError (data.crit_err_msgs.iabot2); end ]] if not (url:lower():find ('^http') or url:find ('^//')) then return inlineError (data.crit_err_msgs.invalid_url ); end ulx.url1 = {} ulx.url1.url = url ulx.url1.extraurls = parseExtraArgs(args) local good = false; good, uri = pcall (mw.uri.new, ulx.url1.url); -- get a table of uri parts from this url; protected mode to prevent lua error when ulx.url1.url is malformed if not good or nil == uri.host then -- abandon when ulx.url1.url is malformed return inlineError (data.crit_err_msgs.invalid_url); end serviceName(uri.host, args.nolink) if args.date and args.date1 then -- Date argument return inlineError (data.crit_err_msgs.conflicting, {origin.date, origin.date1}); end date = args.date or args.date1; date = date and date:gsub (' +', ' '); -- replace multiple spaces with a single space if date and config.verifydates then if '*' == date then date = 'index'; ldf = 'iso'; -- set to default format elseif 'mdy' == date then date = nil; -- if date extracted from URL, ldf = 'mdy'; -- then |date=mdy overrides iso elseif 'dmy' == date then date = nil; -- if date extracted from URL, ldf = 'dmy'; -- then |date=dmy overrides iso elseif 'ymd' == date then date = nil; -- if date extracted from URL, ldf = 'ymd'; -- then |date=ymd overrides iso else date, ldf = decode_date (date); -- get an iso format date from date and get date's original format end end if 'wayback' == ulx.url1.service or 'locwebarchives' == ulx.url1.service or 'ukgwa' == ulx.url1.service then if date then if config.verifydates then if ldf then udate, msg = decodeWaybackDate (uri.path); -- get the url date in iso format and format of date in |date=; 'index' when wayback url date is * if not udate then -- this is the only 'fatal' error return return inlineError (data.crit_err_msgs[msg]); end if udate ~= date then -- date comparison using iso format dates date = udate; msg = table.concat ({ inlineRed (err_warn_msgs.mismatch, 'warning'), -- add warning message msg, -- add message if there is one }); end end end else -- no |date= udate, msg = decodeWaybackDate (uri.path); if not udate then -- this is the only 'fatal' error return return inlineError (data.crit_err_msgs[msg]); end if == udate then date = nil; -- unset else date = udate; end end elseif 'webcite' == ulx.url1.service then if date then if config.verifydates then if ldf then udate = decodeWebciteDate (uri.path); -- get the url date in iso format if 'query' ~= udate then -- skip if query if udate ~= date then -- date comparison using iso format dates date = udate; msg = table.concat ({ inlineRed (err_warn_msgs.mismatch, 'warning'), }); end end end end else date = decodeWebciteDate( uri.path, "iso" ) if date == "query" then date = nil; -- unset msg = inlineRed (err_warn_msgs.date_miss, 'warning'); elseif not date then -- invalid base62 string date = inlineRed (err_warn_msgs.date1, 'error'); end end elseif 'archiveis' == ulx.url1.service then if date then if config.verifydates then if ldf then udate, msg = decodeArchiveisDate (uri.path) -- get the url date in iso format if 'short link' ~= udate then -- skip if short link if udate ~= date then -- date comparison using iso format dates date = udate; msg = table.concat ({ inlineRed (err_warn_msgs.mismatch, 'warning'), -- add warning message msg, -- add message if there is one }); end end end end else -- no |date= udate, msg = decodeArchiveisDate( uri.path, "iso" ) if udate == "short link" then date = nil; -- unset msg = inlineRed (err_warn_msgs.date_miss, 'warning'); elseif == udate then date = nil; -- unset else date = udate; end end else -- some other service if not date then msg = inlineRed (err_warn_msgs.date_miss, 'warning'); end end if 'index' == date then ulx.url1.date = date .. (msg or ); -- create index + message (if there is one) elseif date then ulx.url1.date = makeDate (date, nil, nil, ldf) .. (msg or ); -- create a date in the wiki's local language + message (if there is one) else ulx.url1.date = msg; end format = args.format; -- Format argument if not format then format = "none" else for k, v in pairs (data.format_vals) do -- |format= accepts two specific values loop through a table of those values local found; -- declare a nil flag for _, p in ipairs (v) do -- loop through local language variants if format == p then -- when |format= value matches format = k; -- use name from table key found = true; -- declare found so that we can break out of outer for loop break; -- break out of inner for loop end end if found then break; end end if format == "addlpages" then if not ulx.url1.date then format = "none" end elseif format == "addlarchives" then format = "addlarchives" else format = "none" end end ulx.url1.format = format if args.title and args.title1 then -- Title argument return inlineError (data.crit_err_msgs.conflicting, {origin.title, origin.title1}); end ulx.url1.title = args.title or args.title1; local rend = createRendering() if not rend then return inlineError (data.crit_err_msgs.unknown); end return rend .. ((unnamed_params and inlineRed (err_warn_msgs.unnamed_params, 'warning')) or ) .. createTracking(); end --[[--------------------------< E X P O R T E D F U N C T I O N S >------------------------------------------ ]] return {webarchive = webarchive};. American Film Institute. Retrieved October 23, 2010.
  73. ↑ "AFI's 100 Years... 100 Movies (10th Anniversary Edition)" --[[ ---------------------------------- Lua module implementing the Template loop detected: Template:Webarchive template. A merger of the functionality of three templates: Template:Wayback, Template:Webcite and Template:Cite archives ]] --[[--------------------------< D E P E N D E N C I E S >------------------------------------------------------ ]] require('Module:No globals'); local getArgs = require ('Module:Arguments').getArgs; --[[--------------------------< F O R W A R D D E C L A R A T I O N S >-------------------------------------- ]] local categories = {}; -- category names local config = {}; -- global configuration settings local digits = {}; -- for i18n; table that translates local-wiki digits to western digits local err_warn_msgs = {}; -- error and warning messages local excepted_pages = {}; local month_num = {}; -- for i18n; table that translates local-wiki month names to western digits local prefixes = {}; -- service provider tail string prefixes local services = {}; -- archive service provider data from local s_text = {}; -- table of static text strings used to build final rendering local uncategorized_namespaces = {}; -- list of namespaces that we should not categorize local uncategorized_subpages = {}; -- list of subpages that should not be categorized --[[--------------------------< P A G E S C O P E I D E N T I F I E R S >---------------------------------- ]] local non_western_digits; -- boolean flag set true when data.digits.enable is true local this_page = mw.title.getCurrentTitle(); local track = {}; -- Associative array to hold tracking categories local ulx = {}; -- Associative array to hold template data --[[--------------------------< S U B S T I T U T E >---------------------------------------------------------- Populates numbered arguments in a message string using an argument table. ]] local function substitute (msg, args) return args and mw.message.newRawMessage (msg, args):plain() or msg; end --[[--------------------------< tableLength >----------------------- Given a 1-D table, return number of elements ]] local function tableLength(T) local count = 0 for _ in pairs(T) do count = count + 1 end return count end --[=[-------------------------< M A K E _ W I K I L I N K >---------------------------------------------------- Makes a wikilink; when both link and display text is provided, returns a wikilink in the form D; if only link is provided, returns a wikilink in the form L; if neither are provided or link is omitted, returns an empty string. ]=] local function make_wikilink (link, display, no_link) if nil == no_link then if link and ( ~= link) then if display and ( ~= display) then return table.concat ({'', display, ''}); else return table.concat ({'', link, ''}); end end return display or ; -- link not set so return the display text else -- no_link if display and ( ~= display) then -- if there is display text return display; -- return that else return link or ; -- return the target article name or empty string end end end --[[--------------------------< createTracking >----------------------- Return data in track[] ie. tracking categories ]] local function createTracking() if not excepted_pages[this_page.fullText] then -- namespace:title/fragment is allowed to be categorized (typically this module's / template's testcases page(s)) if uncategorized_namespaces[this_page.nsText] then return ; -- this page not to be categorized so return empty string end for _,v in ipairs (uncategorized_subpages) do -- cycle through page name patterns if this_page.text:match (v) then -- test page name against each pattern return ; -- this subpage type not to be categorized so return empty string end end end local out = {}; if tableLength(track) > 0 then for key, _ in pairs(track) do -- loop through table table.insert (out, make_wikilink (key)); -- and convert category names to links end end return table.concat (out); -- concat into one big string; empty string if table is empty end --[[--------------------------< inlineError >----------------------- Critical error. Render output completely in red. Add to tracking category. This function called as the last thing before abandoning this module ]] local function inlineError (msg, args) track[categories.error] = 1 return table.concat ({ 'Error in ', -- open the error message span config.tname, -- insert the local language template name ' template: ', substitute (msg, args), -- insert the formatted error message '.', -- close the span createTracking() -- add the category }) end --[[--------------------------< inlineRed >----------------------- Render a text fragment in red, such as a warning as part of the final output. Add tracking category. ]] local function inlineRed(msg, trackmsg) if trackmsg == "warning" then track[categories.warning] = 1; elseif trackmsg == "error" then track[categories.error] = 1; end return '' .. msg .. '' end --[[--------------------------< base62 >----------------------- Convert base-62 to base-10 Credit: https://de.wikipedia.org/wiki/Modul:Expr ]] local function base62( value ) local r = 1 -- default return value is input value is malformed if value:match ('%W') then -- value must only be in the set [0-9a-zA-Z] return; -- nil return when value contains extraneous characters end local n = #value -- number of characters in value local k = 1 local c r = 0 for i = n, 1, -1 do -- loop through all characters in value from ls digit to ms digit c = value:byte( i, i ) if c >= 48 and c <= 57 then -- character is digit 0-9 c = c - 48 elseif c >= 65 and c <= 90 then -- character is ascii a-z c = c - 55 else -- must be ascii A-Z c = c - 61 end r = r + c * k -- accumulate this base62 character's value k = k * 62 -- bump for next end -- for i return r end --[[--------------------------< D E C O D E _ D A T E >-------------------------------------------------------- Given a date string, return it in iso format along with an indicator of the date's format. Except that month names must be recognizable as legitimate month names with proper capitalization, and that the date string must match one of the recognized date formats, no error checking is done here; return nil else ]] local function decode_date (date_str) local patterns = { ['dmy'] = {'^(%d%d?) +([^%s%d]+) +(%d%d%d%d)$', 'd', 'm', 'y'}, -- %a does not recognize unicode combining characters used by some languages ['mdy'] = {'^([^%s%d]+) (%d%d?), +(%d%d%d%d)$', 'm', 'd', 'y'}, ['ymd'] = {'^(%d%d%d%d) +([^%s%d]+) (%d%d?)$', 'y', 'm', 'd'}, -- not mos compliant at en.wiki but may be acceptible at other wikis }; local t = {}; if non_western_digits then -- this wiki uses non-western digits? date_str = mw.ustring.gsub (date_str, '%d', digits); -- convert this wiki's non-western digits to western digits end if date_str:match ('^%d%d%d%d%-%d%d%-%d%d$') then -- already an iso format date, return western digits form return date_str, 'iso'; end for k, v in pairs (patterns) do local c1, c2, c3 = mw.ustring.match (date_str, patterns[k][1]); -- c1 .. c3 are captured but we don't know what they hold if c1 then -- set on match t = { -- translate unspecified captures to y, m, and d [patterns[k][2]] = c1, -- fill the table of captures with the captures [patterns[k][3]] = c2, -- take index names from src_pattern table and assign sequential captures [patterns[k][4]] = c3, }; if month_num[t.m] then -- when month not already a number t.m = month_num[t.m]; -- replace valid month name with a number else return nil, 'iso'; -- not a valid date form because month not valid end return mw.ustring.format ('%.4d-%.2d-%.2d', t.y, t.m, t.d), k; -- return date in iso format end end return nil, 'iso'; -- date could not be decoded; return nil and default iso date end --[[--------------------------< makeDate >----------------------- Given year, month, day numbers, (zero-padded or not) return a full date in df format where df may be one of: mdy, dmy, iso, ymd on entry, year, month, day are presumed to be correct for the date that they represent; all are required in this module, makeDate() is sometimes given an iso-format date in year: makeDate (2018-09-20, nil, nil, df) this works because table.concat() sees only one table member ]] local function makeDate (year, month, day, df) local format = { ['dmy'] = 'j F Y', ['mdy'] = 'F j, Y', ['ymd'] = 'Y F j', ['iso'] = 'Y-m-d', }; local date = table.concat ({year, month, day}, '-'); -- assemble year-initial numeric-format date (zero padding not required here) if non_western_digits then -- this wiki uses non-western digits? date = mw.ustring.gsub (date, '%d', digits); -- convert this wiki's non-western digits to western digits end return mw.getContentLanguage():formatDate (format[df], date); end --[[--------------------------< I S _ V A L I D _ D A T E >---------------------------------------------------- Returns true if date is after 31 December 1899 (why is 1900 the min year? shouldn't the internet's date-of-birth be min year?), not after today's date, and represents a valid date (29 February 2017 is not a valid date). Applies Gregorian leapyear rules. all arguments are required ]] local function is_valid_date (year, month, day) local days_in_month = {31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31}; local month_length; local y, m, d; local today = os.date ('*t'); -- fetch a table of current date parts if not year or == year or not month or == month or not day or == day then return false; -- something missing end y = tonumber (year); m = tonumber (month); d = tonumber (day); if 1900 > y or today.year < y or 1 > m or 12 < m then -- year and month are within bounds TODO: 1900? return false; end if (2==m) then -- if February month_length = 28; -- then 28 days unless if (0==(y%4) and (0~=(y%100) or 0==(y%400))) then -- is a leap year? month_length = 29; -- if leap year then 29 days in February end else month_length=days_in_month[m]; end if 1 > d or month_length < d then -- day is within bounds return false; end -- here when date parts represent a valid date return os.time({['year']=y, ['month']=m, ['day']=d, ['hour']=0}) <= os.time(); -- date at midnight must be less than or equal to current date/time end --[[--------------------------< decodeWebciteDate >----------------------- Given a URI-path to Webcite (eg. /67xHmVFWP) return the encoded date in df format returns date string in df format - webcite date is a unix timestamp encoded as bae62 or the string 'query' ]] local function decodeWebciteDate(path, df) local dt = {}; local decode; dt = mw.text.split(path, "/") -- valid URL formats that are not base62 -- http://www.webcitation.org/query?id=1138911916587475 -- http://www.webcitation.org/query?url=http..&date=2012-06-01+21:40:03 -- http://www.webcitation.org/1138911916587475 -- http://www.webcitation.org/cache/73e53dd1f16cf8c5da298418d2a6e452870cf50e -- http://www.webcitation.org/getfile.php?fileid=1c46e791d68e89e12d0c2532cc3cf629b8bc8c8e if dt[2]:find ('query', 1, true) or dt[2]:find ('cache', 1, true) or dt[2]:find ('getfile', 1, true) or tonumber(dt[2]) then return 'query'; end decode = base62(dt[2]); -- base62 string -> exponential number if not decode then return nil; -- nil return when dt[2] contains characters not in %w end dt = os.date('*t', string.format("%d", decode):sub(1,10)) -- exponential number -> text -> first 10 characters (a unix timestamp) -> a table of date parts decode = makeDate (dt.year, dt.month, dt.day, 'iso'); -- date comparisons are all done in iso format with western digits if non_western_digits then -- this wiki uses non-western digits? decode = mw.ustring.gsub (decode, '%d', digits); -- convert this wiki's non-western digits to western digits end return decode; end --[[--------------------------< decodeWaybackDate >----------------------- Given a URI-path to Wayback (eg. /web/20160901010101/http://example.com ) or Library of Congress Web Archives (eg. /all/20160901010101/http://example.com) or UK Government Web Archive (eg. /ukgwa/20160901010101/http://example.com or /tna/20160901010101/http://example.com) return the formatted date eg. "September 1, 2016" in df format Handle non-digits in snapshot ID such as "re_" and "-" and "*" returns two values: first value is one of these: valid date string in df format - wayback date is valid (including the text string 'index' when date is '/*/') empty string - wayback date is malformed (less than 8 digits, not a valid date) nil - wayback date is '/save/' or otherwise not a number second return value is an appropriate 'message' may or may not be formatted ]] local function decodeWaybackDate(path, df) local msg, snapdate; snapdate = path:gsub ('^/web/', ):gsub ('^/all/', ):gsub ('^/ukgwa/', ):gsub ('^/tna/', ):gsub ('^/', ); -- remove leading /web/, /all/, /ukgwa/, /tna/, or / snapdate = snapdate:match ('^[^/]+'); -- get timestamp if snapdate == "*" then -- eg. /web/*/http.., etc. return 'index'; -- return indicator that this url has an index date end snapdate = snapdate:gsub ('%a%a_%d?$', ):gsub ('%-', ); -- from date, remove any trailing "re_", dashes msg = ; if snapdate:match ('%*$') then -- a trailing '*' causes calendar display at archive .org snapdate = snapdate:gsub ('%*$', ); -- remove so not part of length calc later msg = inlineRed (err_warn_msgs.ts_cal, 'warning'); -- make a message end if not tonumber(snapdate) then return nil, 'ts_nan'; -- return nil (fatal error flag) and message selector end local dlen = snapdate:len(); if dlen < 8 then -- we need 8 digits TODO: but shouldn't this be testing for 14 digits? return , inlineRed (err_warn_msgs.ts_short, 'error'); -- return empty string and error message end local year, month, day = snapdate:match ('(%d%d%d%d)(%d%d)(%d%d)'); -- no need for snapdatelong here if not is_valid_date (year, month, day) then return , inlineRed (err_warn_msgs.ts_date, 'error'); -- return empty string and error message end snapdate = table.concat ({year, month, day}, '-'); -- date comparisons are all done in iso format if 14 == dlen then return snapdate, msg; -- return date with message if any else return snapdate, msg .. inlineRed (err_warn_msgs.ts_len, 'warning'); -- return date with warning message(s) end end --[[--------------------------< decodeArchiveisDate >----------------------- Given an Archive.is "long link" URI-path (e.g. /2016.08.28-144552/http://example.com) return the date in df format (e.g. if df = dmy, return 28 August 2016) Handles "." and "-" in snapshot date, so 2016.08.28-144552 is same as 20160828144552 returns two values: first value is one of these: valid date string in df format - archive.is date is valid (including the text string 'short link' when url is the short form) empty string - wayback date is malformed (not a number, less than 8 digits, not a valid date) nil - wayback date is '/save/' second return value is an appropriate 'message' may or may not be formatted ]] local function decodeArchiveisDate(path, df) local snapdate if path:match ('^/%w+$') then -- short form url path is '/' followed by some number of base 62 digits and nothing else return "short link" -- e.g. http://archive.is/hD1qz end snapdate = mw.text.split (path, '/')[2]:gsub('[%.%-]', ); -- get snapshot date, e.g. 2016.08.28-144552; remove periods and hyphens local dlen = string.len(snapdate) if dlen < 8 then -- we need 8 digits TODO: but shouldn't this be testing for 14 digits? return , inlineRed (err_warn_msgs.ts_short, 'error'); -- return empty string and error message end local year, month, day = snapdate:match ('(%d%d%d%d)(%d%d)(%d%d)'); -- no need for snapdatelong here if not is_valid_date (year, month, day) then return , inlineRed (err_warn_msgs.ts_date, 'error'); -- return empty string and error message end snapdate = table.concat ({year, month, day}, '-'); -- date comparisons are all done in iso format if 14 == dlen then return snapdate; -- return date else return snapdate, inlineRed (err_warn_msgs.ts_len, 'warning'); -- return date with warning message end end --[[--------------------------< serviceName >----------------------- Given a domain extracted by mw.uri.new() (eg. web.archive.org) set tail string and service ID ]] local function serviceName(host, no_link) local tracking; local index; host = host:lower():gsub ('^web%.(.+)', '%1'):gsub ('^www%.(.+)', '%1'); -- lowercase, remove web. and www. subdomains if services[host] then index = host; else for k, _ in pairs (services) do if host:find ('%f[%a]'..k:gsub ('([%.%-])', '%%%1')) then index = k; break; end end end if index then local out = {}; -- empty string in [1] so that concatenated result has leading single space ulx.url1.service = services[index][4] or 'other'; tracking = services[index][5] or categories.other; -- build tail string if false == services[index][1] then -- select prefix table.insert (out, prefixes.at); elseif true == services[index][1] then table.insert (out, prefixes.atthe); else table.insert (out, services[index][1]); end table.insert (out, make_wikilink (services[index][2], services[index][3], no_link)); -- add article wikilink if services[index][6] then -- add tail postfix if it exists table.insert (out, services[index][6]); end ulx.url1.tail = table.concat (out, ' '); -- put it all together; result has leading space character else -- here when unknown archive ulx.url1.service = 'other'; tracking = categories.unknown; ulx.url1.tail = table.concat ({, prefixes.at, host, inlineRed (err_warn_msgs.unknown_url, error)}, ' '); end track[tracking] = 1 end --[[--------------------------< parseExtraArgs >----------------------- Parse numbered arguments starting at 2, such as url2..url10, date2..date10, title2..title10 For example: Template loop detected: Template:Webarchive Three url arguments not in numeric sequence (1..4..7). Function only processes arguments numbered 2 or greater (in this case 4 and 7) It creates numeric sequenced table entries like: urlx.url2.url = <argument value for url4> urlx.url3.url = <argument value for url7> Returns the number of URL arguments found numbered 2 or greater (in this case returns "2") ]] local function parseExtraArgs(args) local i, j, argurl, argurl2, argdate, argtitle j = 2 for i = 2, config.maxurls do argurl = "url" .. i if args[argurl] then argurl2 = "url" .. j ulx[argurl2] = {} ulx[argurl2]["url"] = args[argurl] argdate = "date" .. i if args[argdate] then ulx[argurl2]["date"] = args[argdate] else ulx[argurl2]["date"] = inlineRed (err_warn_msgs.date_miss, 'warning'); end argtitle = "title" .. i if args[argtitle] then ulx[argurl2]["title"] = args[argtitle] else ulx[argurl2]["title"] = nil end j = j + 1 end end if j == 2 then return 0 else return j - 2 end end --[[--------------------------< comma >----------------------- Given a date string, return "," if it's MDY ]] local function comma(date) return (date and date:match ('%a+ +%d%d?(,) +%d%d%d%d')) or ; end --[[--------------------------< createRendering >----------------------- Return a rendering of the data in ulx[][] ]] local function createRendering() local displayfield local out = {}; local index_date, msg = ulx.url1.date:match ('(index)(.*)'); -- when ulx.url1.date extract 'index' text and message text (if there is a message) ulx.url1.date = ulx.url1.date:gsub ('index.*', 'index'); -- remove message if 'none' == ulx.url1.format then -- For Template:Wayback, Template:Webcite table.insert (out, '['); -- open extlink markup table.insert (out, ulx.url1.url); -- add url if ulx.url1.title then table.insert (out, ' ') -- the required space table.insert (out, ulx.url1.title) -- the title table.insert (out, ']'); -- close extlink markup table.insert (out, ulx.url1.tail); -- tail text if ulx.url1.date then table.insert (out, ' ('); -- open date text; TODO: why the html entity? replace with regular space? table.insert (out, 'index' == ulx.url1.date and s_text.archive or s_text.archived); -- add text table.insert (out, ' '); -- insert a space table.insert (out, ulx.url1.date); -- add date table.insert (out, ')'); -- close date text end else -- no title if index_date then -- when url date is 'index' table.insert (out, table.concat ({' ', s_text.Archive_index, ']'})); -- add the index link label table.insert (out, msg or ); -- add date mismatch message when url date is /*/ and |date= has valid date else table.insert (out, table.concat ({' ', s_text.Archived, '] '})); -- add link label for url has timestamp date (will include mismatch message if there is one) end if ulx.url1.date then if 'index' ~= ulx.url1.date then table.insert (out, ulx.url1.date); -- add date when data is not 'index' end table.insert (out, comma(ulx.url1.date)); -- add ',' if date format is mdy table.insert (out, ulx.url1.tail); -- add tail text else -- no date table.insert (out, ulx.url1.tail); -- add tail text end end if 0 < ulx.url1.extraurls then -- For multiple archive URLs local tot = ulx.url1.extraurls + 1 table.insert (out, '.') -- terminate first url table.insert (out, table.concat ({' ', s_text.addlarchives, ': '})); -- add header text for i=2, tot do -- loop through the additionals local index = table.concat ({'url', i}); -- make an index displayfield = ulx[index]['title'] and 'title' or 'date'; -- choose display text table.insert (out, '['); -- open extlink markup table.insert (out, ulx[index]['url']); -- add the url table.insert (out, ' '); -- the required space table.insert (out, ulx[index][displayfield]); -- add the label table.insert (out, ']'); -- close extlink markup table.insert (out, i==tot and '.' or ', '); -- add terminator end end return table.concat (out); -- make a big string and done else -- For Template:Cite archives if 'addlarchives' == ulx.url1.format then -- Multiple archive services table.insert (out, table.concat ({s_text.addlarchives, ': '})); -- add header text else -- Multiple pages from the same archive table.insert (out, table.concat ({s_text.addlpages, ' '})); -- add header text table.insert (out, ulx.url1.date); -- add date to header text table.insert (out, ': '); -- close header text end local tot = ulx.url1.extraurls + 1; for i=1, tot do -- loop through the additionals local index = table.concat ({'url', i}); -- make an index table.insert (out, '['); -- open extlink markup table.insert (out, ulx[index]['url']); -- add url table.insert (out, ' '); -- add required space displayfield = ulx[index]['title']; if 'addlarchives' == ulx.url1.format then if not displayfield then displayfield = ulx[index]['date'] end else -- must be addlpages if not displayfield then displayfield = table.concat ({s_text.Page, ' ', i}); end end table.insert (out, displayfield); -- add title, date, page label text table.insert (out, ']'); -- close extlink markup table.insert (out, (i==tot and '.' or ', ')); -- add terminator end return table.concat (out); -- make a big string and done end end --[[--------------------------< P A R A M E T E R _ N A M E _ X L A T E >-------------------------------------- for internaltionalization, translate local-language parameter names to their English equivalents TODO: return error message if multiple aliases of the same canonical parameter name are found? returns two tables: new_args - holds canonical form parameters and their values either from translation or because the parameter was already in canonical form origin - maps canonical-form parameter names to their untranslated (local language) form for error messaging in the local language unrecognized parameters are ignored ]] local function parameter_name_xlate (args, params, enum_params) local name; -- holds modifiable name of the parameter name during evaluation local enum; -- for enumerated parameters, holds the enumerator during evaluation local found = false; -- flag used to break out of nested for loops local new_args = {}; -- a table that holds canonical and translated parameter k/v pairs local origin = {}; -- a table that maps original (local language) parameter names to their canonical name for local language error messaging local unnamed_params; -- set true when unsupported positional parameters are detected for k, v in pairs (args) do -- loop through all of the arguments in the args table name = k; -- copy of original parameter name if 'string' == type (k) then if non_western_digits then -- true when non-western digits supported at this wiki name = mw.ustring.gsub (name, '%d', digits); -- convert this wiki's non-western digits to western digits end enum = name:match ('%d+$'); -- get parameter enumerator if it exists; nil else if not enum then -- no enumerator so looking for non-enumnerated parameters -- TODO: insert shortcut here? if params[name] then name holds the canonical parameter name; no need to search further for pname, aliases in pairs (params) do -- loop through each parameter the params table for _, alias in ipairs (aliases) do -- loop through each alias in the parameter's aliases table if name == alias then new_args[pname] = v; -- create a new entry in the new_args table origin [pname] = k; -- create an entry to make canonical parameter name to original local language parameter name found = true; -- flag so that we can break out of these nested for loops break; -- no need to search the rest of the aliases table for name so go on to the next k, v pair end end if found then -- true when we found an alias that matched name found = false; -- reset the flag break; -- go do next args k/v pair end end else -- enumerated parameters name = name:gsub ('%d$', '#'); -- replace enumeration digits with place holder for table search -- TODO: insert shortcut here? if num_params[name] then name holds the canonical parameter name; no need to search further for pname, aliases in pairs (enum_params) do -- loop through each parameter the num_params table for _, alias in ipairs (aliases) do -- loop through each alias in the parameter's aliases table if name == alias then pname = pname:gsub ('#$', enum); -- replace the '#' place holder with the actual enumerator new_args[pname] = v; -- create a new entry in the new_args table origin [pname] = k; -- create an entry to make canonical parameter name to original local language parameter name found = true; -- flag so that we can break out of these nested for loops break; -- no need to search the rest of the aliases table for name so go on to the next k, v pair end end if found then -- true when we found an alias that matched name found = false; -- reset the flag break; -- go do next args k/v pair end end end else unnamed_params = true; -- flag for unsupported positional parameters end end -- for k, v return new_args, origin, unnamed_params; end --[[--------------------------< W E B A R C H I V E >---------------------------------------------------------- template entry point ]] local function webarchive(frame) local args = getArgs (frame); local data = mw.loadData (table.concat ({ -- make a data module name; sandbox or live 'Module:Webarchive/data', frame:getTitle():find('sandbox', 1, true) and '/sandbox' or -- this instance is ./sandbox then append /sandbox })); categories = data.categories; -- fill in the forward declarations config = data.config; if data.digits.enable then digits = data.digits; -- for i18n; table of digits in the local wiki's language non_western_digits = true; -- use_non_western_digits end err_warn_msgs = data.err_warn_msgs; excepted_pages = data.excepted_pages; month_num = data.month_num; -- for i18n; table of month names in the local wiki's language prefixes = data.prefixes; services = data.services; s_text = data.s_text; uncategorized_namespaces = data.uncategorized_namespaces; uncategorized_subpages = data.uncategorized_subpages; local origin = {}; -- holds a map of English to local language parameter names used in the current template; not currently used local unnamed_params; -- boolean set to true when template call has unnamed parameters args, origin, unnamed_params = parameter_name_xlate (args, data.params, data.enum_params); -- translate parameter names in args to English local date, format, msg, udate, uri, url; local ldf = 'iso'; -- when there is no |date= parameter, render url dates in iso format if args.url and args.url1 then -- URL argument (first) return inlineError (data.crit_err_msgs.conflicting, {origin.url, origin.url1}); end url = args.url or args.url1; if not url then return inlineError (data.crit_err_msgs.empty); end -- these iabot bugs perportedly fixed; removing these causes lua script error --[[ -- at Template:Webarchive/testcases/Production; resolve that before deleting these tests if mw.ustring.find( url, "https://web.http", 1, true ) then -- track bug - TODO: IAbot bug; not known if the bug has been fixed; deferred track[categories.error] = 1; return inlineError (data.crit_err_msgs.iabot1); end if url == "https://web.archive.org/http:/" then -- track bug - TODO: IAbot bug; not known if the bug has been fixed; deferred track[categories.error] = 1; return inlineError (data.crit_err_msgs.iabot2); end ]] if not (url:lower():find ('^http') or url:find ('^//')) then return inlineError (data.crit_err_msgs.invalid_url ); end ulx.url1 = {} ulx.url1.url = url ulx.url1.extraurls = parseExtraArgs(args) local good = false; good, uri = pcall (mw.uri.new, ulx.url1.url); -- get a table of uri parts from this url; protected mode to prevent lua error when ulx.url1.url is malformed if not good or nil == uri.host then -- abandon when ulx.url1.url is malformed return inlineError (data.crit_err_msgs.invalid_url); end serviceName(uri.host, args.nolink) if args.date and args.date1 then -- Date argument return inlineError (data.crit_err_msgs.conflicting, {origin.date, origin.date1}); end date = args.date or args.date1; date = date and date:gsub (' +', ' '); -- replace multiple spaces with a single space if date and config.verifydates then if '*' == date then date = 'index'; ldf = 'iso'; -- set to default format elseif 'mdy' == date then date = nil; -- if date extracted from URL, ldf = 'mdy'; -- then |date=mdy overrides iso elseif 'dmy' == date then date = nil; -- if date extracted from URL, ldf = 'dmy'; -- then |date=dmy overrides iso elseif 'ymd' == date then date = nil; -- if date extracted from URL, ldf = 'ymd'; -- then |date=ymd overrides iso else date, ldf = decode_date (date); -- get an iso format date from date and get date's original format end end if 'wayback' == ulx.url1.service or 'locwebarchives' == ulx.url1.service or 'ukgwa' == ulx.url1.service then if date then if config.verifydates then if ldf then udate, msg = decodeWaybackDate (uri.path); -- get the url date in iso format and format of date in |date=; 'index' when wayback url date is * if not udate then -- this is the only 'fatal' error return return inlineError (data.crit_err_msgs[msg]); end if udate ~= date then -- date comparison using iso format dates date = udate; msg = table.concat ({ inlineRed (err_warn_msgs.mismatch, 'warning'), -- add warning message msg, -- add message if there is one }); end end end else -- no |date= udate, msg = decodeWaybackDate (uri.path); if not udate then -- this is the only 'fatal' error return return inlineError (data.crit_err_msgs[msg]); end if == udate then date = nil; -- unset else date = udate; end end elseif 'webcite' == ulx.url1.service then if date then if config.verifydates then if ldf then udate = decodeWebciteDate (uri.path); -- get the url date in iso format if 'query' ~= udate then -- skip if query if udate ~= date then -- date comparison using iso format dates date = udate; msg = table.concat ({ inlineRed (err_warn_msgs.mismatch, 'warning'), }); end end end end else date = decodeWebciteDate( uri.path, "iso" ) if date == "query" then date = nil; -- unset msg = inlineRed (err_warn_msgs.date_miss, 'warning'); elseif not date then -- invalid base62 string date = inlineRed (err_warn_msgs.date1, 'error'); end end elseif 'archiveis' == ulx.url1.service then if date then if config.verifydates then if ldf then udate, msg = decodeArchiveisDate (uri.path) -- get the url date in iso format if 'short link' ~= udate then -- skip if short link if udate ~= date then -- date comparison using iso format dates date = udate; msg = table.concat ({ inlineRed (err_warn_msgs.mismatch, 'warning'), -- add warning message msg, -- add message if there is one }); end end end end else -- no |date= udate, msg = decodeArchiveisDate( uri.path, "iso" ) if udate == "short link" then date = nil; -- unset msg = inlineRed (err_warn_msgs.date_miss, 'warning'); elseif == udate then date = nil; -- unset else date = udate; end end else -- some other service if not date then msg = inlineRed (err_warn_msgs.date_miss, 'warning'); end end if 'index' == date then ulx.url1.date = date .. (msg or ); -- create index + message (if there is one) elseif date then ulx.url1.date = makeDate (date, nil, nil, ldf) .. (msg or ); -- create a date in the wiki's local language + message (if there is one) else ulx.url1.date = msg; end format = args.format; -- Format argument if not format then format = "none" else for k, v in pairs (data.format_vals) do -- |format= accepts two specific values loop through a table of those values local found; -- declare a nil flag for _, p in ipairs (v) do -- loop through local language variants if format == p then -- when |format= value matches format = k; -- use name from table key found = true; -- declare found so that we can break out of outer for loop break; -- break out of inner for loop end end if found then break; end end if format == "addlpages" then if not ulx.url1.date then format = "none" end elseif format == "addlarchives" then format = "addlarchives" else format = "none" end end ulx.url1.format = format if args.title and args.title1 then -- Title argument return inlineError (data.crit_err_msgs.conflicting, {origin.title, origin.title1}); end ulx.url1.title = args.title or args.title1; local rend = createRendering() if not rend then return inlineError (data.crit_err_msgs.unknown); end return rend .. ((unnamed_params and inlineRed (err_warn_msgs.unnamed_params, 'warning')) or ) .. createTracking(); end --[[--------------------------< E X P O R T E D F U N C T I O N S >------------------------------------------ ]] return {webarchive = webarchive};. American Film Institute. Retrieved October 23, 2010.
  74. ↑ AFI's 10 Top 10. American Film Institute. Retrieved on October 23, 2010.
  75. ↑ Thompson, David and Christie, Ian Scorsese on Scorsese p. 83.
  76. ↑ 76.0 76.1 Evans, Mike The Making of Raging Bull p. 88.
  77. ↑ FAQ 9. What is that nice music in Raging Bull?. mascagni.org. Retrieved on February 26, 2009.
  78. ↑ Pamela McClintock. Sunset Pictures in shape. Variety.
  79. ↑ 'Raging Bull II' is Shooting Now – /Film. Slashfilm.
  80. ↑ "Raging Bull II" headed into production. www.ifc.com.
  81. ↑ The Deadline Team. Cannes: Main Street Films Sets U.S. Dates For 'Great Expectations' And 'Bronx Bull' – Deadline. Deadline.
  82. ↑ Patten, D. "MGM Files ‘Raging Bull 2′ Lawsuit Against Jake LaMotta & Sequel Producers." --[[ ---------------------------------- Lua module implementing the Template loop detected: Template:Webarchive template. A merger of the functionality of three templates: Template:Wayback, Template:Webcite and Template:Cite archives ]] --[[--------------------------< D E P E N D E N C I E S >------------------------------------------------------ ]] require('Module:No globals'); local getArgs = require ('Module:Arguments').getArgs; --[[--------------------------< F O R W A R D D E C L A R A T I O N S >-------------------------------------- ]] local categories = {}; -- category names local config = {}; -- global configuration settings local digits = {}; -- for i18n; table that translates local-wiki digits to western digits local err_warn_msgs = {}; -- error and warning messages local excepted_pages = {}; local month_num = {}; -- for i18n; table that translates local-wiki month names to western digits local prefixes = {}; -- service provider tail string prefixes local services = {}; -- archive service provider data from local s_text = {}; -- table of static text strings used to build final rendering local uncategorized_namespaces = {}; -- list of namespaces that we should not categorize local uncategorized_subpages = {}; -- list of subpages that should not be categorized --[[--------------------------< P A G E S C O P E I D E N T I F I E R S >---------------------------------- ]] local non_western_digits; -- boolean flag set true when data.digits.enable is true local this_page = mw.title.getCurrentTitle(); local track = {}; -- Associative array to hold tracking categories local ulx = {}; -- Associative array to hold template data --[[--------------------------< S U B S T I T U T E >---------------------------------------------------------- Populates numbered arguments in a message string using an argument table. ]] local function substitute (msg, args) return args and mw.message.newRawMessage (msg, args):plain() or msg; end --[[--------------------------< tableLength >----------------------- Given a 1-D table, return number of elements ]] local function tableLength(T) local count = 0 for _ in pairs(T) do count = count + 1 end return count end --[=[-------------------------< M A K E _ W I K I L I N K >---------------------------------------------------- Makes a wikilink; when both link and display text is provided, returns a wikilink in the form D; if only link is provided, returns a wikilink in the form L; if neither are provided or link is omitted, returns an empty string. ]=] local function make_wikilink (link, display, no_link) if nil == no_link then if link and ( ~= link) then if display and ( ~= display) then return table.concat ({'', display, ''}); else return table.concat ({'', link, ''}); end end return display or ; -- link not set so return the display text else -- no_link if display and ( ~= display) then -- if there is display text return display; -- return that else return link or ; -- return the target article name or empty string end end end --[[--------------------------< createTracking >----------------------- Return data in track[] ie. tracking categories ]] local function createTracking() if not excepted_pages[this_page.fullText] then -- namespace:title/fragment is allowed to be categorized (typically this module's / template's testcases page(s)) if uncategorized_namespaces[this_page.nsText] then return ; -- this page not to be categorized so return empty string end for _,v in ipairs (uncategorized_subpages) do -- cycle through page name patterns if this_page.text:match (v) then -- test page name against each pattern return ; -- this subpage type not to be categorized so return empty string end end end local out = {}; if tableLength(track) > 0 then for key, _ in pairs(track) do -- loop through table table.insert (out, make_wikilink (key)); -- and convert category names to links end end return table.concat (out); -- concat into one big string; empty string if table is empty end --[[--------------------------< inlineError >----------------------- Critical error. Render output completely in red. Add to tracking category. This function called as the last thing before abandoning this module ]] local function inlineError (msg, args) track[categories.error] = 1 return table.concat ({ 'Error in ', -- open the error message span config.tname, -- insert the local language template name ' template: ', substitute (msg, args), -- insert the formatted error message '.', -- close the span createTracking() -- add the category }) end --[[--------------------------< inlineRed >----------------------- Render a text fragment in red, such as a warning as part of the final output. Add tracking category. ]] local function inlineRed(msg, trackmsg) if trackmsg == "warning" then track[categories.warning] = 1; elseif trackmsg == "error" then track[categories.error] = 1; end return '' .. msg .. '' end --[[--------------------------< base62 >----------------------- Convert base-62 to base-10 Credit: https://de.wikipedia.org/wiki/Modul:Expr ]] local function base62( value ) local r = 1 -- default return value is input value is malformed if value:match ('%W') then -- value must only be in the set [0-9a-zA-Z] return; -- nil return when value contains extraneous characters end local n = #value -- number of characters in value local k = 1 local c r = 0 for i = n, 1, -1 do -- loop through all characters in value from ls digit to ms digit c = value:byte( i, i ) if c >= 48 and c <= 57 then -- character is digit 0-9 c = c - 48 elseif c >= 65 and c <= 90 then -- character is ascii a-z c = c - 55 else -- must be ascii A-Z c = c - 61 end r = r + c * k -- accumulate this base62 character's value k = k * 62 -- bump for next end -- for i return r end --[[--------------------------< D E C O D E _ D A T E >-------------------------------------------------------- Given a date string, return it in iso format along with an indicator of the date's format. Except that month names must be recognizable as legitimate month names with proper capitalization, and that the date string must match one of the recognized date formats, no error checking is done here; return nil else ]] local function decode_date (date_str) local patterns = { ['dmy'] = {'^(%d%d?) +([^%s%d]+) +(%d%d%d%d)$', 'd', 'm', 'y'}, -- %a does not recognize unicode combining characters used by some languages ['mdy'] = {'^([^%s%d]+) (%d%d?), +(%d%d%d%d)$', 'm', 'd', 'y'}, ['ymd'] = {'^(%d%d%d%d) +([^%s%d]+) (%d%d?)$', 'y', 'm', 'd'}, -- not mos compliant at en.wiki but may be acceptible at other wikis }; local t = {}; if non_western_digits then -- this wiki uses non-western digits? date_str = mw.ustring.gsub (date_str, '%d', digits); -- convert this wiki's non-western digits to western digits end if date_str:match ('^%d%d%d%d%-%d%d%-%d%d$') then -- already an iso format date, return western digits form return date_str, 'iso'; end for k, v in pairs (patterns) do local c1, c2, c3 = mw.ustring.match (date_str, patterns[k][1]); -- c1 .. c3 are captured but we don't know what they hold if c1 then -- set on match t = { -- translate unspecified captures to y, m, and d [patterns[k][2]] = c1, -- fill the table of captures with the captures [patterns[k][3]] = c2, -- take index names from src_pattern table and assign sequential captures [patterns[k][4]] = c3, }; if month_num[t.m] then -- when month not already a number t.m = month_num[t.m]; -- replace valid month name with a number else return nil, 'iso'; -- not a valid date form because month not valid end return mw.ustring.format ('%.4d-%.2d-%.2d', t.y, t.m, t.d), k; -- return date in iso format end end return nil, 'iso'; -- date could not be decoded; return nil and default iso date end --[[--------------------------< makeDate >----------------------- Given year, month, day numbers, (zero-padded or not) return a full date in df format where df may be one of: mdy, dmy, iso, ymd on entry, year, month, day are presumed to be correct for the date that they represent; all are required in this module, makeDate() is sometimes given an iso-format date in year: makeDate (2018-09-20, nil, nil, df) this works because table.concat() sees only one table member ]] local function makeDate (year, month, day, df) local format = { ['dmy'] = 'j F Y', ['mdy'] = 'F j, Y', ['ymd'] = 'Y F j', ['iso'] = 'Y-m-d', }; local date = table.concat ({year, month, day}, '-'); -- assemble year-initial numeric-format date (zero padding not required here) if non_western_digits then -- this wiki uses non-western digits? date = mw.ustring.gsub (date, '%d', digits); -- convert this wiki's non-western digits to western digits end return mw.getContentLanguage():formatDate (format[df], date); end --[[--------------------------< I S _ V A L I D _ D A T E >---------------------------------------------------- Returns true if date is after 31 December 1899 (why is 1900 the min year? shouldn't the internet's date-of-birth be min year?), not after today's date, and represents a valid date (29 February 2017 is not a valid date). Applies Gregorian leapyear rules. all arguments are required ]] local function is_valid_date (year, month, day) local days_in_month = {31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31}; local month_length; local y, m, d; local today = os.date ('*t'); -- fetch a table of current date parts if not year or == year or not month or == month or not day or == day then return false; -- something missing end y = tonumber (year); m = tonumber (month); d = tonumber (day); if 1900 > y or today.year < y or 1 > m or 12 < m then -- year and month are within bounds TODO: 1900? return false; end if (2==m) then -- if February month_length = 28; -- then 28 days unless if (0==(y%4) and (0~=(y%100) or 0==(y%400))) then -- is a leap year? month_length = 29; -- if leap year then 29 days in February end else month_length=days_in_month[m]; end if 1 > d or month_length < d then -- day is within bounds return false; end -- here when date parts represent a valid date return os.time({['year']=y, ['month']=m, ['day']=d, ['hour']=0}) <= os.time(); -- date at midnight must be less than or equal to current date/time end --[[--------------------------< decodeWebciteDate >----------------------- Given a URI-path to Webcite (eg. /67xHmVFWP) return the encoded date in df format returns date string in df format - webcite date is a unix timestamp encoded as bae62 or the string 'query' ]] local function decodeWebciteDate(path, df) local dt = {}; local decode; dt = mw.text.split(path, "/") -- valid URL formats that are not base62 -- http://www.webcitation.org/query?id=1138911916587475 -- http://www.webcitation.org/query?url=http..&date=2012-06-01+21:40:03 -- http://www.webcitation.org/1138911916587475 -- http://www.webcitation.org/cache/73e53dd1f16cf8c5da298418d2a6e452870cf50e -- http://www.webcitation.org/getfile.php?fileid=1c46e791d68e89e12d0c2532cc3cf629b8bc8c8e if dt[2]:find ('query', 1, true) or dt[2]:find ('cache', 1, true) or dt[2]:find ('getfile', 1, true) or tonumber(dt[2]) then return 'query'; end decode = base62(dt[2]); -- base62 string -> exponential number if not decode then return nil; -- nil return when dt[2] contains characters not in %w end dt = os.date('*t', string.format("%d", decode):sub(1,10)) -- exponential number -> text -> first 10 characters (a unix timestamp) -> a table of date parts decode = makeDate (dt.year, dt.month, dt.day, 'iso'); -- date comparisons are all done in iso format with western digits if non_western_digits then -- this wiki uses non-western digits? decode = mw.ustring.gsub (decode, '%d', digits); -- convert this wiki's non-western digits to western digits end return decode; end --[[--------------------------< decodeWaybackDate >----------------------- Given a URI-path to Wayback (eg. /web/20160901010101/http://example.com ) or Library of Congress Web Archives (eg. /all/20160901010101/http://example.com) or UK Government Web Archive (eg. /ukgwa/20160901010101/http://example.com or /tna/20160901010101/http://example.com) return the formatted date eg. "September 1, 2016" in df format Handle non-digits in snapshot ID such as "re_" and "-" and "*" returns two values: first value is one of these: valid date string in df format - wayback date is valid (including the text string 'index' when date is '/*/') empty string - wayback date is malformed (less than 8 digits, not a valid date) nil - wayback date is '/save/' or otherwise not a number second return value is an appropriate 'message' may or may not be formatted ]] local function decodeWaybackDate(path, df) local msg, snapdate; snapdate = path:gsub ('^/web/', ):gsub ('^/all/', ):gsub ('^/ukgwa/', ):gsub ('^/tna/', ):gsub ('^/', ); -- remove leading /web/, /all/, /ukgwa/, /tna/, or / snapdate = snapdate:match ('^[^/]+'); -- get timestamp if snapdate == "*" then -- eg. /web/*/http.., etc. return 'index'; -- return indicator that this url has an index date end snapdate = snapdate:gsub ('%a%a_%d?$', ):gsub ('%-', ); -- from date, remove any trailing "re_", dashes msg = ; if snapdate:match ('%*$') then -- a trailing '*' causes calendar display at archive .org snapdate = snapdate:gsub ('%*$', ); -- remove so not part of length calc later msg = inlineRed (err_warn_msgs.ts_cal, 'warning'); -- make a message end if not tonumber(snapdate) then return nil, 'ts_nan'; -- return nil (fatal error flag) and message selector end local dlen = snapdate:len(); if dlen < 8 then -- we need 8 digits TODO: but shouldn't this be testing for 14 digits? return , inlineRed (err_warn_msgs.ts_short, 'error'); -- return empty string and error message end local year, month, day = snapdate:match ('(%d%d%d%d)(%d%d)(%d%d)'); -- no need for snapdatelong here if not is_valid_date (year, month, day) then return , inlineRed (err_warn_msgs.ts_date, 'error'); -- return empty string and error message end snapdate = table.concat ({year, month, day}, '-'); -- date comparisons are all done in iso format if 14 == dlen then return snapdate, msg; -- return date with message if any else return snapdate, msg .. inlineRed (err_warn_msgs.ts_len, 'warning'); -- return date with warning message(s) end end --[[--------------------------< decodeArchiveisDate >----------------------- Given an Archive.is "long link" URI-path (e.g. /2016.08.28-144552/http://example.com) return the date in df format (e.g. if df = dmy, return 28 August 2016) Handles "." and "-" in snapshot date, so 2016.08.28-144552 is same as 20160828144552 returns two values: first value is one of these: valid date string in df format - archive.is date is valid (including the text string 'short link' when url is the short form) empty string - wayback date is malformed (not a number, less than 8 digits, not a valid date) nil - wayback date is '/save/' second return value is an appropriate 'message' may or may not be formatted ]] local function decodeArchiveisDate(path, df) local snapdate if path:match ('^/%w+$') then -- short form url path is '/' followed by some number of base 62 digits and nothing else return "short link" -- e.g. http://archive.is/hD1qz end snapdate = mw.text.split (path, '/')[2]:gsub('[%.%-]', ); -- get snapshot date, e.g. 2016.08.28-144552; remove periods and hyphens local dlen = string.len(snapdate) if dlen < 8 then -- we need 8 digits TODO: but shouldn't this be testing for 14 digits? return , inlineRed (err_warn_msgs.ts_short, 'error'); -- return empty string and error message end local year, month, day = snapdate:match ('(%d%d%d%d)(%d%d)(%d%d)'); -- no need for snapdatelong here if not is_valid_date (year, month, day) then return , inlineRed (err_warn_msgs.ts_date, 'error'); -- return empty string and error message end snapdate = table.concat ({year, month, day}, '-'); -- date comparisons are all done in iso format if 14 == dlen then return snapdate; -- return date else return snapdate, inlineRed (err_warn_msgs.ts_len, 'warning'); -- return date with warning message end end --[[--------------------------< serviceName >----------------------- Given a domain extracted by mw.uri.new() (eg. web.archive.org) set tail string and service ID ]] local function serviceName(host, no_link) local tracking; local index; host = host:lower():gsub ('^web%.(.+)', '%1'):gsub ('^www%.(.+)', '%1'); -- lowercase, remove web. and www. subdomains if services[host] then index = host; else for k, _ in pairs (services) do if host:find ('%f[%a]'..k:gsub ('([%.%-])', '%%%1')) then index = k; break; end end end if index then local out = {}; -- empty string in [1] so that concatenated result has leading single space ulx.url1.service = services[index][4] or 'other'; tracking = services[index][5] or categories.other; -- build tail string if false == services[index][1] then -- select prefix table.insert (out, prefixes.at); elseif true == services[index][1] then table.insert (out, prefixes.atthe); else table.insert (out, services[index][1]); end table.insert (out, make_wikilink (services[index][2], services[index][3], no_link)); -- add article wikilink if services[index][6] then -- add tail postfix if it exists table.insert (out, services[index][6]); end ulx.url1.tail = table.concat (out, ' '); -- put it all together; result has leading space character else -- here when unknown archive ulx.url1.service = 'other'; tracking = categories.unknown; ulx.url1.tail = table.concat ({, prefixes.at, host, inlineRed (err_warn_msgs.unknown_url, error)}, ' '); end track[tracking] = 1 end --[[--------------------------< parseExtraArgs >----------------------- Parse numbered arguments starting at 2, such as url2..url10, date2..date10, title2..title10 For example: Template loop detected: Template:Webarchive Three url arguments not in numeric sequence (1..4..7). Function only processes arguments numbered 2 or greater (in this case 4 and 7) It creates numeric sequenced table entries like: urlx.url2.url = <argument value for url4> urlx.url3.url = <argument value for url7> Returns the number of URL arguments found numbered 2 or greater (in this case returns "2") ]] local function parseExtraArgs(args) local i, j, argurl, argurl2, argdate, argtitle j = 2 for i = 2, config.maxurls do argurl = "url" .. i if args[argurl] then argurl2 = "url" .. j ulx[argurl2] = {} ulx[argurl2]["url"] = args[argurl] argdate = "date" .. i if args[argdate] then ulx[argurl2]["date"] = args[argdate] else ulx[argurl2]["date"] = inlineRed (err_warn_msgs.date_miss, 'warning'); end argtitle = "title" .. i if args[argtitle] then ulx[argurl2]["title"] = args[argtitle] else ulx[argurl2]["title"] = nil end j = j + 1 end end if j == 2 then return 0 else return j - 2 end end --[[--------------------------< comma >----------------------- Given a date string, return "," if it's MDY ]] local function comma(date) return (date and date:match ('%a+ +%d%d?(,) +%d%d%d%d')) or ; end --[[--------------------------< createRendering >----------------------- Return a rendering of the data in ulx[][] ]] local function createRendering() local displayfield local out = {}; local index_date, msg = ulx.url1.date:match ('(index)(.*)'); -- when ulx.url1.date extract 'index' text and message text (if there is a message) ulx.url1.date = ulx.url1.date:gsub ('index.*', 'index'); -- remove message if 'none' == ulx.url1.format then -- For Template:Wayback, Template:Webcite table.insert (out, '['); -- open extlink markup table.insert (out, ulx.url1.url); -- add url if ulx.url1.title then table.insert (out, ' ') -- the required space table.insert (out, ulx.url1.title) -- the title table.insert (out, ']'); -- close extlink markup table.insert (out, ulx.url1.tail); -- tail text if ulx.url1.date then table.insert (out, ' ('); -- open date text; TODO: why the html entity? replace with regular space? table.insert (out, 'index' == ulx.url1.date and s_text.archive or s_text.archived); -- add text table.insert (out, ' '); -- insert a space table.insert (out, ulx.url1.date); -- add date table.insert (out, ')'); -- close date text end else -- no title if index_date then -- when url date is 'index' table.insert (out, table.concat ({' ', s_text.Archive_index, ']'})); -- add the index link label table.insert (out, msg or ); -- add date mismatch message when url date is /*/ and |date= has valid date else table.insert (out, table.concat ({' ', s_text.Archived, '] '})); -- add link label for url has timestamp date (will include mismatch message if there is one) end if ulx.url1.date then if 'index' ~= ulx.url1.date then table.insert (out, ulx.url1.date); -- add date when data is not 'index' end table.insert (out, comma(ulx.url1.date)); -- add ',' if date format is mdy table.insert (out, ulx.url1.tail); -- add tail text else -- no date table.insert (out, ulx.url1.tail); -- add tail text end end if 0 < ulx.url1.extraurls then -- For multiple archive URLs local tot = ulx.url1.extraurls + 1 table.insert (out, '.') -- terminate first url table.insert (out, table.concat ({' ', s_text.addlarchives, ': '})); -- add header text for i=2, tot do -- loop through the additionals local index = table.concat ({'url', i}); -- make an index displayfield = ulx[index]['title'] and 'title' or 'date'; -- choose display text table.insert (out, '['); -- open extlink markup table.insert (out, ulx[index]['url']); -- add the url table.insert (out, ' '); -- the required space table.insert (out, ulx[index][displayfield]); -- add the label table.insert (out, ']'); -- close extlink markup table.insert (out, i==tot and '.' or ', '); -- add terminator end end return table.concat (out); -- make a big string and done else -- For Template:Cite archives if 'addlarchives' == ulx.url1.format then -- Multiple archive services table.insert (out, table.concat ({s_text.addlarchives, ': '})); -- add header text else -- Multiple pages from the same archive table.insert (out, table.concat ({s_text.addlpages, ' '})); -- add header text table.insert (out, ulx.url1.date); -- add date to header text table.insert (out, ': '); -- close header text end local tot = ulx.url1.extraurls + 1; for i=1, tot do -- loop through the additionals local index = table.concat ({'url', i}); -- make an index table.insert (out, '['); -- open extlink markup table.insert (out, ulx[index]['url']); -- add url table.insert (out, ' '); -- add required space displayfield = ulx[index]['title']; if 'addlarchives' == ulx.url1.format then if not displayfield then displayfield = ulx[index]['date'] end else -- must be addlpages if not displayfield then displayfield = table.concat ({s_text.Page, ' ', i}); end end table.insert (out, displayfield); -- add title, date, page label text table.insert (out, ']'); -- close extlink markup table.insert (out, (i==tot and '.' or ', ')); -- add terminator end return table.concat (out); -- make a big string and done end end --[[--------------------------< P A R A M E T E R _ N A M E _ X L A T E >-------------------------------------- for internaltionalization, translate local-language parameter names to their English equivalents TODO: return error message if multiple aliases of the same canonical parameter name are found? returns two tables: new_args - holds canonical form parameters and their values either from translation or because the parameter was already in canonical form origin - maps canonical-form parameter names to their untranslated (local language) form for error messaging in the local language unrecognized parameters are ignored ]] local function parameter_name_xlate (args, params, enum_params) local name; -- holds modifiable name of the parameter name during evaluation local enum; -- for enumerated parameters, holds the enumerator during evaluation local found = false; -- flag used to break out of nested for loops local new_args = {}; -- a table that holds canonical and translated parameter k/v pairs local origin = {}; -- a table that maps original (local language) parameter names to their canonical name for local language error messaging local unnamed_params; -- set true when unsupported positional parameters are detected for k, v in pairs (args) do -- loop through all of the arguments in the args table name = k; -- copy of original parameter name if 'string' == type (k) then if non_western_digits then -- true when non-western digits supported at this wiki name = mw.ustring.gsub (name, '%d', digits); -- convert this wiki's non-western digits to western digits end enum = name:match ('%d+$'); -- get parameter enumerator if it exists; nil else if not enum then -- no enumerator so looking for non-enumnerated parameters -- TODO: insert shortcut here? if params[name] then name holds the canonical parameter name; no need to search further for pname, aliases in pairs (params) do -- loop through each parameter the params table for _, alias in ipairs (aliases) do -- loop through each alias in the parameter's aliases table if name == alias then new_args[pname] = v; -- create a new entry in the new_args table origin [pname] = k; -- create an entry to make canonical parameter name to original local language parameter name found = true; -- flag so that we can break out of these nested for loops break; -- no need to search the rest of the aliases table for name so go on to the next k, v pair end end if found then -- true when we found an alias that matched name found = false; -- reset the flag break; -- go do next args k/v pair end end else -- enumerated parameters name = name:gsub ('%d$', '#'); -- replace enumeration digits with place holder for table search -- TODO: insert shortcut here? if num_params[name] then name holds the canonical parameter name; no need to search further for pname, aliases in pairs (enum_params) do -- loop through each parameter the num_params table for _, alias in ipairs (aliases) do -- loop through each alias in the parameter's aliases table if name == alias then pname = pname:gsub ('#$', enum); -- replace the '#' place holder with the actual enumerator new_args[pname] = v; -- create a new entry in the new_args table origin [pname] = k; -- create an entry to make canonical parameter name to original local language parameter name found = true; -- flag so that we can break out of these nested for loops break; -- no need to search the rest of the aliases table for name so go on to the next k, v pair end end if found then -- true when we found an alias that matched name found = false; -- reset the flag break; -- go do next args k/v pair end end end else unnamed_params = true; -- flag for unsupported positional parameters end end -- for k, v return new_args, origin, unnamed_params; end --[[--------------------------< W E B A R C H I V E >---------------------------------------------------------- template entry point ]] local function webarchive(frame) local args = getArgs (frame); local data = mw.loadData (table.concat ({ -- make a data module name; sandbox or live 'Module:Webarchive/data', frame:getTitle():find('sandbox', 1, true) and '/sandbox' or -- this instance is ./sandbox then append /sandbox })); categories = data.categories; -- fill in the forward declarations config = data.config; if data.digits.enable then digits = data.digits; -- for i18n; table of digits in the local wiki's language non_western_digits = true; -- use_non_western_digits end err_warn_msgs = data.err_warn_msgs; excepted_pages = data.excepted_pages; month_num = data.month_num; -- for i18n; table of month names in the local wiki's language prefixes = data.prefixes; services = data.services; s_text = data.s_text; uncategorized_namespaces = data.uncategorized_namespaces; uncategorized_subpages = data.uncategorized_subpages; local origin = {}; -- holds a map of English to local language parameter names used in the current template; not currently used local unnamed_params; -- boolean set to true when template call has unnamed parameters args, origin, unnamed_params = parameter_name_xlate (args, data.params, data.enum_params); -- translate parameter names in args to English local date, format, msg, udate, uri, url; local ldf = 'iso'; -- when there is no |date= parameter, render url dates in iso format if args.url and args.url1 then -- URL argument (first) return inlineError (data.crit_err_msgs.conflicting, {origin.url, origin.url1}); end url = args.url or args.url1; if not url then return inlineError (data.crit_err_msgs.empty); end -- these iabot bugs perportedly fixed; removing these causes lua script error --[[ -- at Template:Webarchive/testcases/Production; resolve that before deleting these tests if mw.ustring.find( url, "https://web.http", 1, true ) then -- track bug - TODO: IAbot bug; not known if the bug has been fixed; deferred track[categories.error] = 1; return inlineError (data.crit_err_msgs.iabot1); end if url == "https://web.archive.org/http:/" then -- track bug - TODO: IAbot bug; not known if the bug has been fixed; deferred track[categories.error] = 1; return inlineError (data.crit_err_msgs.iabot2); end ]] if not (url:lower():find ('^http') or url:find ('^//')) then return inlineError (data.crit_err_msgs.invalid_url ); end ulx.url1 = {} ulx.url1.url = url ulx.url1.extraurls = parseExtraArgs(args) local good = false; good, uri = pcall (mw.uri.new, ulx.url1.url); -- get a table of uri parts from this url; protected mode to prevent lua error when ulx.url1.url is malformed if not good or nil == uri.host then -- abandon when ulx.url1.url is malformed return inlineError (data.crit_err_msgs.invalid_url); end serviceName(uri.host, args.nolink) if args.date and args.date1 then -- Date argument return inlineError (data.crit_err_msgs.conflicting, {origin.date, origin.date1}); end date = args.date or args.date1; date = date and date:gsub (' +', ' '); -- replace multiple spaces with a single space if date and config.verifydates then if '*' == date then date = 'index'; ldf = 'iso'; -- set to default format elseif 'mdy' == date then date = nil; -- if date extracted from URL, ldf = 'mdy'; -- then |date=mdy overrides iso elseif 'dmy' == date then date = nil; -- if date extracted from URL, ldf = 'dmy'; -- then |date=dmy overrides iso elseif 'ymd' == date then date = nil; -- if date extracted from URL, ldf = 'ymd'; -- then |date=ymd overrides iso else date, ldf = decode_date (date); -- get an iso format date from date and get date's original format end end if 'wayback' == ulx.url1.service or 'locwebarchives' == ulx.url1.service or 'ukgwa' == ulx.url1.service then if date then if config.verifydates then if ldf then udate, msg = decodeWaybackDate (uri.path); -- get the url date in iso format and format of date in |date=; 'index' when wayback url date is * if not udate then -- this is the only 'fatal' error return return inlineError (data.crit_err_msgs[msg]); end if udate ~= date then -- date comparison using iso format dates date = udate; msg = table.concat ({ inlineRed (err_warn_msgs.mismatch, 'warning'), -- add warning message msg, -- add message if there is one }); end end end else -- no |date= udate, msg = decodeWaybackDate (uri.path); if not udate then -- this is the only 'fatal' error return return inlineError (data.crit_err_msgs[msg]); end if == udate then date = nil; -- unset else date = udate; end end elseif 'webcite' == ulx.url1.service then if date then if config.verifydates then if ldf then udate = decodeWebciteDate (uri.path); -- get the url date in iso format if 'query' ~= udate then -- skip if query if udate ~= date then -- date comparison using iso format dates date = udate; msg = table.concat ({ inlineRed (err_warn_msgs.mismatch, 'warning'), }); end end end end else date = decodeWebciteDate( uri.path, "iso" ) if date == "query" then date = nil; -- unset msg = inlineRed (err_warn_msgs.date_miss, 'warning'); elseif not date then -- invalid base62 string date = inlineRed (err_warn_msgs.date1, 'error'); end end elseif 'archiveis' == ulx.url1.service then if date then if config.verifydates then if ldf then udate, msg = decodeArchiveisDate (uri.path) -- get the url date in iso format if 'short link' ~= udate then -- skip if short link if udate ~= date then -- date comparison using iso format dates date = udate; msg = table.concat ({ inlineRed (err_warn_msgs.mismatch, 'warning'), -- add warning message msg, -- add message if there is one }); end end end end else -- no |date= udate, msg = decodeArchiveisDate( uri.path, "iso" ) if udate == "short link" then date = nil; -- unset msg = inlineRed (err_warn_msgs.date_miss, 'warning'); elseif == udate then date = nil; -- unset else date = udate; end end else -- some other service if not date then msg = inlineRed (err_warn_msgs.date_miss, 'warning'); end end if 'index' == date then ulx.url1.date = date .. (msg or ); -- create index + message (if there is one) elseif date then ulx.url1.date = makeDate (date, nil, nil, ldf) .. (msg or ); -- create a date in the wiki's local language + message (if there is one) else ulx.url1.date = msg; end format = args.format; -- Format argument if not format then format = "none" else for k, v in pairs (data.format_vals) do -- |format= accepts two specific values loop through a table of those values local found; -- declare a nil flag for _, p in ipairs (v) do -- loop through local language variants if format == p then -- when |format= value matches format = k; -- use name from table key found = true; -- declare found so that we can break out of outer for loop break; -- break out of inner for loop end end if found then break; end end if format == "addlpages" then if not ulx.url1.date then format = "none" end elseif format == "addlarchives" then format = "addlarchives" else format = "none" end end ulx.url1.format = format if args.title and args.title1 then -- Title argument return inlineError (data.crit_err_msgs.conflicting, {origin.title, origin.title1}); end ulx.url1.title = args.title or args.title1; local rend = createRendering() if not rend then return inlineError (data.crit_err_msgs.unknown); end return rend .. ((unnamed_params and inlineRed (err_warn_msgs.unnamed_params, 'warning')) or ) .. createTracking(); end --[[--------------------------< E X P O R T E D F U N C T I O N S >------------------------------------------ ]] return {webarchive = webarchive}; Deadline Hollywood (July 3, 2012). Retrieved July 5, 2012.
  83. ↑ Patten, Dominic. "MGM Settles ‘Raging Bull II’ Lawsuit With Movie Name Change." --[[ ---------------------------------- Lua module implementing the Template loop detected: Template:Webarchive template. A merger of the functionality of three templates: Template:Wayback, Template:Webcite and Template:Cite archives ]] --[[--------------------------< D E P E N D E N C I E S >------------------------------------------------------ ]] require('Module:No globals'); local getArgs = require ('Module:Arguments').getArgs; --[[--------------------------< F O R W A R D D E C L A R A T I O N S >-------------------------------------- ]] local categories = {}; -- category names local config = {}; -- global configuration settings local digits = {}; -- for i18n; table that translates local-wiki digits to western digits local err_warn_msgs = {}; -- error and warning messages local excepted_pages = {}; local month_num = {}; -- for i18n; table that translates local-wiki month names to western digits local prefixes = {}; -- service provider tail string prefixes local services = {}; -- archive service provider data from local s_text = {}; -- table of static text strings used to build final rendering local uncategorized_namespaces = {}; -- list of namespaces that we should not categorize local uncategorized_subpages = {}; -- list of subpages that should not be categorized --[[--------------------------< P A G E S C O P E I D E N T I F I E R S >---------------------------------- ]] local non_western_digits; -- boolean flag set true when data.digits.enable is true local this_page = mw.title.getCurrentTitle(); local track = {}; -- Associative array to hold tracking categories local ulx = {}; -- Associative array to hold template data --[[--------------------------< S U B S T I T U T E >---------------------------------------------------------- Populates numbered arguments in a message string using an argument table. ]] local function substitute (msg, args) return args and mw.message.newRawMessage (msg, args):plain() or msg; end --[[--------------------------< tableLength >----------------------- Given a 1-D table, return number of elements ]] local function tableLength(T) local count = 0 for _ in pairs(T) do count = count + 1 end return count end --[=[-------------------------< M A K E _ W I K I L I N K >---------------------------------------------------- Makes a wikilink; when both link and display text is provided, returns a wikilink in the form D; if only link is provided, returns a wikilink in the form L; if neither are provided or link is omitted, returns an empty string. ]=] local function make_wikilink (link, display, no_link) if nil == no_link then if link and ( ~= link) then if display and ( ~= display) then return table.concat ({'', display, ''}); else return table.concat ({'', link, ''}); end end return display or ; -- link not set so return the display text else -- no_link if display and ( ~= display) then -- if there is display text return display; -- return that else return link or ; -- return the target article name or empty string end end end --[[--------------------------< createTracking >----------------------- Return data in track[] ie. tracking categories ]] local function createTracking() if not excepted_pages[this_page.fullText] then -- namespace:title/fragment is allowed to be categorized (typically this module's / template's testcases page(s)) if uncategorized_namespaces[this_page.nsText] then return ; -- this page not to be categorized so return empty string end for _,v in ipairs (uncategorized_subpages) do -- cycle through page name patterns if this_page.text:match (v) then -- test page name against each pattern return ; -- this subpage type not to be categorized so return empty string end end end local out = {}; if tableLength(track) > 0 then for key, _ in pairs(track) do -- loop through table table.insert (out, make_wikilink (key)); -- and convert category names to links end end return table.concat (out); -- concat into one big string; empty string if table is empty end --[[--------------------------< inlineError >----------------------- Critical error. Render output completely in red. Add to tracking category. This function called as the last thing before abandoning this module ]] local function inlineError (msg, args) track[categories.error] = 1 return table.concat ({ 'Error in ', -- open the error message span config.tname, -- insert the local language template name ' template: ', substitute (msg, args), -- insert the formatted error message '.', -- close the span createTracking() -- add the category }) end --[[--------------------------< inlineRed >----------------------- Render a text fragment in red, such as a warning as part of the final output. Add tracking category. ]] local function inlineRed(msg, trackmsg) if trackmsg == "warning" then track[categories.warning] = 1; elseif trackmsg == "error" then track[categories.error] = 1; end return '' .. msg .. '' end --[[--------------------------< base62 >----------------------- Convert base-62 to base-10 Credit: https://de.wikipedia.org/wiki/Modul:Expr ]] local function base62( value ) local r = 1 -- default return value is input value is malformed if value:match ('%W') then -- value must only be in the set [0-9a-zA-Z] return; -- nil return when value contains extraneous characters end local n = #value -- number of characters in value local k = 1 local c r = 0 for i = n, 1, -1 do -- loop through all characters in value from ls digit to ms digit c = value:byte( i, i ) if c >= 48 and c <= 57 then -- character is digit 0-9 c = c - 48 elseif c >= 65 and c <= 90 then -- character is ascii a-z c = c - 55 else -- must be ascii A-Z c = c - 61 end r = r + c * k -- accumulate this base62 character's value k = k * 62 -- bump for next end -- for i return r end --[[--------------------------< D E C O D E _ D A T E >-------------------------------------------------------- Given a date string, return it in iso format along with an indicator of the date's format. Except that month names must be recognizable as legitimate month names with proper capitalization, and that the date string must match one of the recognized date formats, no error checking is done here; return nil else ]] local function decode_date (date_str) local patterns = { ['dmy'] = {'^(%d%d?) +([^%s%d]+) +(%d%d%d%d)$', 'd', 'm', 'y'}, -- %a does not recognize unicode combining characters used by some languages ['mdy'] = {'^([^%s%d]+) (%d%d?), +(%d%d%d%d)$', 'm', 'd', 'y'}, ['ymd'] = {'^(%d%d%d%d) +([^%s%d]+) (%d%d?)$', 'y', 'm', 'd'}, -- not mos compliant at en.wiki but may be acceptible at other wikis }; local t = {}; if non_western_digits then -- this wiki uses non-western digits? date_str = mw.ustring.gsub (date_str, '%d', digits); -- convert this wiki's non-western digits to western digits end if date_str:match ('^%d%d%d%d%-%d%d%-%d%d$') then -- already an iso format date, return western digits form return date_str, 'iso'; end for k, v in pairs (patterns) do local c1, c2, c3 = mw.ustring.match (date_str, patterns[k][1]); -- c1 .. c3 are captured but we don't know what they hold if c1 then -- set on match t = { -- translate unspecified captures to y, m, and d [patterns[k][2]] = c1, -- fill the table of captures with the captures [patterns[k][3]] = c2, -- take index names from src_pattern table and assign sequential captures [patterns[k][4]] = c3, }; if month_num[t.m] then -- when month not already a number t.m = month_num[t.m]; -- replace valid month name with a number else return nil, 'iso'; -- not a valid date form because month not valid end return mw.ustring.format ('%.4d-%.2d-%.2d', t.y, t.m, t.d), k; -- return date in iso format end end return nil, 'iso'; -- date could not be decoded; return nil and default iso date end --[[--------------------------< makeDate >----------------------- Given year, month, day numbers, (zero-padded or not) return a full date in df format where df may be one of: mdy, dmy, iso, ymd on entry, year, month, day are presumed to be correct for the date that they represent; all are required in this module, makeDate() is sometimes given an iso-format date in year: makeDate (2018-09-20, nil, nil, df) this works because table.concat() sees only one table member ]] local function makeDate (year, month, day, df) local format = { ['dmy'] = 'j F Y', ['mdy'] = 'F j, Y', ['ymd'] = 'Y F j', ['iso'] = 'Y-m-d', }; local date = table.concat ({year, month, day}, '-'); -- assemble year-initial numeric-format date (zero padding not required here) if non_western_digits then -- this wiki uses non-western digits? date = mw.ustring.gsub (date, '%d', digits); -- convert this wiki's non-western digits to western digits end return mw.getContentLanguage():formatDate (format[df], date); end --[[--------------------------< I S _ V A L I D _ D A T E >---------------------------------------------------- Returns true if date is after 31 December 1899 (why is 1900 the min year? shouldn't the internet's date-of-birth be min year?), not after today's date, and represents a valid date (29 February 2017 is not a valid date). Applies Gregorian leapyear rules. all arguments are required ]] local function is_valid_date (year, month, day) local days_in_month = {31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31}; local month_length; local y, m, d; local today = os.date ('*t'); -- fetch a table of current date parts if not year or == year or not month or == month or not day or == day then return false; -- something missing end y = tonumber (year); m = tonumber (month); d = tonumber (day); if 1900 > y or today.year < y or 1 > m or 12 < m then -- year and month are within bounds TODO: 1900? return false; end if (2==m) then -- if February month_length = 28; -- then 28 days unless if (0==(y%4) and (0~=(y%100) or 0==(y%400))) then -- is a leap year? month_length = 29; -- if leap year then 29 days in February end else month_length=days_in_month[m]; end if 1 > d or month_length < d then -- day is within bounds return false; end -- here when date parts represent a valid date return os.time({['year']=y, ['month']=m, ['day']=d, ['hour']=0}) <= os.time(); -- date at midnight must be less than or equal to current date/time end --[[--------------------------< decodeWebciteDate >----------------------- Given a URI-path to Webcite (eg. /67xHmVFWP) return the encoded date in df format returns date string in df format - webcite date is a unix timestamp encoded as bae62 or the string 'query' ]] local function decodeWebciteDate(path, df) local dt = {}; local decode; dt = mw.text.split(path, "/") -- valid URL formats that are not base62 -- http://www.webcitation.org/query?id=1138911916587475 -- http://www.webcitation.org/query?url=http..&date=2012-06-01+21:40:03 -- http://www.webcitation.org/1138911916587475 -- http://www.webcitation.org/cache/73e53dd1f16cf8c5da298418d2a6e452870cf50e -- http://www.webcitation.org/getfile.php?fileid=1c46e791d68e89e12d0c2532cc3cf629b8bc8c8e if dt[2]:find ('query', 1, true) or dt[2]:find ('cache', 1, true) or dt[2]:find ('getfile', 1, true) or tonumber(dt[2]) then return 'query'; end decode = base62(dt[2]); -- base62 string -> exponential number if not decode then return nil; -- nil return when dt[2] contains characters not in %w end dt = os.date('*t', string.format("%d", decode):sub(1,10)) -- exponential number -> text -> first 10 characters (a unix timestamp) -> a table of date parts decode = makeDate (dt.year, dt.month, dt.day, 'iso'); -- date comparisons are all done in iso format with western digits if non_western_digits then -- this wiki uses non-western digits? decode = mw.ustring.gsub (decode, '%d', digits); -- convert this wiki's non-western digits to western digits end return decode; end --[[--------------------------< decodeWaybackDate >----------------------- Given a URI-path to Wayback (eg. /web/20160901010101/http://example.com ) or Library of Congress Web Archives (eg. /all/20160901010101/http://example.com) or UK Government Web Archive (eg. /ukgwa/20160901010101/http://example.com or /tna/20160901010101/http://example.com) return the formatted date eg. "September 1, 2016" in df format Handle non-digits in snapshot ID such as "re_" and "-" and "*" returns two values: first value is one of these: valid date string in df format - wayback date is valid (including the text string 'index' when date is '/*/') empty string - wayback date is malformed (less than 8 digits, not a valid date) nil - wayback date is '/save/' or otherwise not a number second return value is an appropriate 'message' may or may not be formatted ]] local function decodeWaybackDate(path, df) local msg, snapdate; snapdate = path:gsub ('^/web/', ):gsub ('^/all/', ):gsub ('^/ukgwa/', ):gsub ('^/tna/', ):gsub ('^/', ); -- remove leading /web/, /all/, /ukgwa/, /tna/, or / snapdate = snapdate:match ('^[^/]+'); -- get timestamp if snapdate == "*" then -- eg. /web/*/http.., etc. return 'index'; -- return indicator that this url has an index date end snapdate = snapdate:gsub ('%a%a_%d?$', ):gsub ('%-', ); -- from date, remove any trailing "re_", dashes msg = ; if snapdate:match ('%*$') then -- a trailing '*' causes calendar display at archive .org snapdate = snapdate:gsub ('%*$', ); -- remove so not part of length calc later msg = inlineRed (err_warn_msgs.ts_cal, 'warning'); -- make a message end if not tonumber(snapdate) then return nil, 'ts_nan'; -- return nil (fatal error flag) and message selector end local dlen = snapdate:len(); if dlen < 8 then -- we need 8 digits TODO: but shouldn't this be testing for 14 digits? return , inlineRed (err_warn_msgs.ts_short, 'error'); -- return empty string and error message end local year, month, day = snapdate:match ('(%d%d%d%d)(%d%d)(%d%d)'); -- no need for snapdatelong here if not is_valid_date (year, month, day) then return , inlineRed (err_warn_msgs.ts_date, 'error'); -- return empty string and error message end snapdate = table.concat ({year, month, day}, '-'); -- date comparisons are all done in iso format if 14 == dlen then return snapdate, msg; -- return date with message if any else return snapdate, msg .. inlineRed (err_warn_msgs.ts_len, 'warning'); -- return date with warning message(s) end end --[[--------------------------< decodeArchiveisDate >----------------------- Given an Archive.is "long link" URI-path (e.g. /2016.08.28-144552/http://example.com) return the date in df format (e.g. if df = dmy, return 28 August 2016) Handles "." and "-" in snapshot date, so 2016.08.28-144552 is same as 20160828144552 returns two values: first value is one of these: valid date string in df format - archive.is date is valid (including the text string 'short link' when url is the short form) empty string - wayback date is malformed (not a number, less than 8 digits, not a valid date) nil - wayback date is '/save/' second return value is an appropriate 'message' may or may not be formatted ]] local function decodeArchiveisDate(path, df) local snapdate if path:match ('^/%w+$') then -- short form url path is '/' followed by some number of base 62 digits and nothing else return "short link" -- e.g. http://archive.is/hD1qz end snapdate = mw.text.split (path, '/')[2]:gsub('[%.%-]', ); -- get snapshot date, e.g. 2016.08.28-144552; remove periods and hyphens local dlen = string.len(snapdate) if dlen < 8 then -- we need 8 digits TODO: but shouldn't this be testing for 14 digits? return , inlineRed (err_warn_msgs.ts_short, 'error'); -- return empty string and error message end local year, month, day = snapdate:match ('(%d%d%d%d)(%d%d)(%d%d)'); -- no need for snapdatelong here if not is_valid_date (year, month, day) then return , inlineRed (err_warn_msgs.ts_date, 'error'); -- return empty string and error message end snapdate = table.concat ({year, month, day}, '-'); -- date comparisons are all done in iso format if 14 == dlen then return snapdate; -- return date else return snapdate, inlineRed (err_warn_msgs.ts_len, 'warning'); -- return date with warning message end end --[[--------------------------< serviceName >----------------------- Given a domain extracted by mw.uri.new() (eg. web.archive.org) set tail string and service ID ]] local function serviceName(host, no_link) local tracking; local index; host = host:lower():gsub ('^web%.(.+)', '%1'):gsub ('^www%.(.+)', '%1'); -- lowercase, remove web. and www. subdomains if services[host] then index = host; else for k, _ in pairs (services) do if host:find ('%f[%a]'..k:gsub ('([%.%-])', '%%%1')) then index = k; break; end end end if index then local out = {}; -- empty string in [1] so that concatenated result has leading single space ulx.url1.service = services[index][4] or 'other'; tracking = services[index][5] or categories.other; -- build tail string if false == services[index][1] then -- select prefix table.insert (out, prefixes.at); elseif true == services[index][1] then table.insert (out, prefixes.atthe); else table.insert (out, services[index][1]); end table.insert (out, make_wikilink (services[index][2], services[index][3], no_link)); -- add article wikilink if services[index][6] then -- add tail postfix if it exists table.insert (out, services[index][6]); end ulx.url1.tail = table.concat (out, ' '); -- put it all together; result has leading space character else -- here when unknown archive ulx.url1.service = 'other'; tracking = categories.unknown; ulx.url1.tail = table.concat ({, prefixes.at, host, inlineRed (err_warn_msgs.unknown_url, error)}, ' '); end track[tracking] = 1 end --[[--------------------------< parseExtraArgs >----------------------- Parse numbered arguments starting at 2, such as url2..url10, date2..date10, title2..title10 For example: Template loop detected: Template:Webarchive Three url arguments not in numeric sequence (1..4..7). Function only processes arguments numbered 2 or greater (in this case 4 and 7) It creates numeric sequenced table entries like: urlx.url2.url = <argument value for url4> urlx.url3.url = <argument value for url7> Returns the number of URL arguments found numbered 2 or greater (in this case returns "2") ]] local function parseExtraArgs(args) local i, j, argurl, argurl2, argdate, argtitle j = 2 for i = 2, config.maxurls do argurl = "url" .. i if args[argurl] then argurl2 = "url" .. j ulx[argurl2] = {} ulx[argurl2]["url"] = args[argurl] argdate = "date" .. i if args[argdate] then ulx[argurl2]["date"] = args[argdate] else ulx[argurl2]["date"] = inlineRed (err_warn_msgs.date_miss, 'warning'); end argtitle = "title" .. i if args[argtitle] then ulx[argurl2]["title"] = args[argtitle] else ulx[argurl2]["title"] = nil end j = j + 1 end end if j == 2 then return 0 else return j - 2 end end --[[--------------------------< comma >----------------------- Given a date string, return "," if it's MDY ]] local function comma(date) return (date and date:match ('%a+ +%d%d?(,) +%d%d%d%d')) or ; end --[[--------------------------< createRendering >----------------------- Return a rendering of the data in ulx[][] ]] local function createRendering() local displayfield local out = {}; local index_date, msg = ulx.url1.date:match ('(index)(.*)'); -- when ulx.url1.date extract 'index' text and message text (if there is a message) ulx.url1.date = ulx.url1.date:gsub ('index.*', 'index'); -- remove message if 'none' == ulx.url1.format then -- For Template:Wayback, Template:Webcite table.insert (out, '['); -- open extlink markup table.insert (out, ulx.url1.url); -- add url if ulx.url1.title then table.insert (out, ' ') -- the required space table.insert (out, ulx.url1.title) -- the title table.insert (out, ']'); -- close extlink markup table.insert (out, ulx.url1.tail); -- tail text if ulx.url1.date then table.insert (out, ' ('); -- open date text; TODO: why the html entity? replace with regular space? table.insert (out, 'index' == ulx.url1.date and s_text.archive or s_text.archived); -- add text table.insert (out, ' '); -- insert a space table.insert (out, ulx.url1.date); -- add date table.insert (out, ')'); -- close date text end else -- no title if index_date then -- when url date is 'index' table.insert (out, table.concat ({' ', s_text.Archive_index, ']'})); -- add the index link label table.insert (out, msg or ); -- add date mismatch message when url date is /*/ and |date= has valid date else table.insert (out, table.concat ({' ', s_text.Archived, '] '})); -- add link label for url has timestamp date (will include mismatch message if there is one) end if ulx.url1.date then if 'index' ~= ulx.url1.date then table.insert (out, ulx.url1.date); -- add date when data is not 'index' end table.insert (out, comma(ulx.url1.date)); -- add ',' if date format is mdy table.insert (out, ulx.url1.tail); -- add tail text else -- no date table.insert (out, ulx.url1.tail); -- add tail text end end if 0 < ulx.url1.extraurls then -- For multiple archive URLs local tot = ulx.url1.extraurls + 1 table.insert (out, '.') -- terminate first url table.insert (out, table.concat ({' ', s_text.addlarchives, ': '})); -- add header text for i=2, tot do -- loop through the additionals local index = table.concat ({'url', i}); -- make an index displayfield = ulx[index]['title'] and 'title' or 'date'; -- choose display text table.insert (out, '['); -- open extlink markup table.insert (out, ulx[index]['url']); -- add the url table.insert (out, ' '); -- the required space table.insert (out, ulx[index][displayfield]); -- add the label table.insert (out, ']'); -- close extlink markup table.insert (out, i==tot and '.' or ', '); -- add terminator end end return table.concat (out); -- make a big string and done else -- For Template:Cite archives if 'addlarchives' == ulx.url1.format then -- Multiple archive services table.insert (out, table.concat ({s_text.addlarchives, ': '})); -- add header text else -- Multiple pages from the same archive table.insert (out, table.concat ({s_text.addlpages, ' '})); -- add header text table.insert (out, ulx.url1.date); -- add date to header text table.insert (out, ': '); -- close header text end local tot = ulx.url1.extraurls + 1; for i=1, tot do -- loop through the additionals local index = table.concat ({'url', i}); -- make an index table.insert (out, '['); -- open extlink markup table.insert (out, ulx[index]['url']); -- add url table.insert (out, ' '); -- add required space displayfield = ulx[index]['title']; if 'addlarchives' == ulx.url1.format then if not displayfield then displayfield = ulx[index]['date'] end else -- must be addlpages if not displayfield then displayfield = table.concat ({s_text.Page, ' ', i}); end end table.insert (out, displayfield); -- add title, date, page label text table.insert (out, ']'); -- close extlink markup table.insert (out, (i==tot and '.' or ', ')); -- add terminator end return table.concat (out); -- make a big string and done end end --[[--------------------------< P A R A M E T E R _ N A M E _ X L A T E >-------------------------------------- for internaltionalization, translate local-language parameter names to their English equivalents TODO: return error message if multiple aliases of the same canonical parameter name are found? returns two tables: new_args - holds canonical form parameters and their values either from translation or because the parameter was already in canonical form origin - maps canonical-form parameter names to their untranslated (local language) form for error messaging in the local language unrecognized parameters are ignored ]] local function parameter_name_xlate (args, params, enum_params) local name; -- holds modifiable name of the parameter name during evaluation local enum; -- for enumerated parameters, holds the enumerator during evaluation local found = false; -- flag used to break out of nested for loops local new_args = {}; -- a table that holds canonical and translated parameter k/v pairs local origin = {}; -- a table that maps original (local language) parameter names to their canonical name for local language error messaging local unnamed_params; -- set true when unsupported positional parameters are detected for k, v in pairs (args) do -- loop through all of the arguments in the args table name = k; -- copy of original parameter name if 'string' == type (k) then if non_western_digits then -- true when non-western digits supported at this wiki name = mw.ustring.gsub (name, '%d', digits); -- convert this wiki's non-western digits to western digits end enum = name:match ('%d+$'); -- get parameter enumerator if it exists; nil else if not enum then -- no enumerator so looking for non-enumnerated parameters -- TODO: insert shortcut here? if params[name] then name holds the canonical parameter name; no need to search further for pname, aliases in pairs (params) do -- loop through each parameter the params table for _, alias in ipairs (aliases) do -- loop through each alias in the parameter's aliases table if name == alias then new_args[pname] = v; -- create a new entry in the new_args table origin [pname] = k; -- create an entry to make canonical parameter name to original local language parameter name found = true; -- flag so that we can break out of these nested for loops break; -- no need to search the rest of the aliases table for name so go on to the next k, v pair end end if found then -- true when we found an alias that matched name found = false; -- reset the flag break; -- go do next args k/v pair end end else -- enumerated parameters name = name:gsub ('%d$', '#'); -- replace enumeration digits with place holder for table search -- TODO: insert shortcut here? if num_params[name] then name holds the canonical parameter name; no need to search further for pname, aliases in pairs (enum_params) do -- loop through each parameter the num_params table for _, alias in ipairs (aliases) do -- loop through each alias in the parameter's aliases table if name == alias then pname = pname:gsub ('#$', enum); -- replace the '#' place holder with the actual enumerator new_args[pname] = v; -- create a new entry in the new_args table origin [pname] = k; -- create an entry to make canonical parameter name to original local language parameter name found = true; -- flag so that we can break out of these nested for loops break; -- no need to search the rest of the aliases table for name so go on to the next k, v pair end end if found then -- true when we found an alias that matched name found = false; -- reset the flag break; -- go do next args k/v pair end end end else unnamed_params = true; -- flag for unsupported positional parameters end end -- for k, v return new_args, origin, unnamed_params; end --[[--------------------------< W E B A R C H I V E >---------------------------------------------------------- template entry point ]] local function webarchive(frame) local args = getArgs (frame); local data = mw.loadData (table.concat ({ -- make a data module name; sandbox or live 'Module:Webarchive/data', frame:getTitle():find('sandbox', 1, true) and '/sandbox' or -- this instance is ./sandbox then append /sandbox })); categories = data.categories; -- fill in the forward declarations config = data.config; if data.digits.enable then digits = data.digits; -- for i18n; table of digits in the local wiki's language non_western_digits = true; -- use_non_western_digits end err_warn_msgs = data.err_warn_msgs; excepted_pages = data.excepted_pages; month_num = data.month_num; -- for i18n; table of month names in the local wiki's language prefixes = data.prefixes; services = data.services; s_text = data.s_text; uncategorized_namespaces = data.uncategorized_namespaces; uncategorized_subpages = data.uncategorized_subpages; local origin = {}; -- holds a map of English to local language parameter names used in the current template; not currently used local unnamed_params; -- boolean set to true when template call has unnamed parameters args, origin, unnamed_params = parameter_name_xlate (args, data.params, data.enum_params); -- translate parameter names in args to English local date, format, msg, udate, uri, url; local ldf = 'iso'; -- when there is no |date= parameter, render url dates in iso format if args.url and args.url1 then -- URL argument (first) return inlineError (data.crit_err_msgs.conflicting, {origin.url, origin.url1}); end url = args.url or args.url1; if not url then return inlineError (data.crit_err_msgs.empty); end -- these iabot bugs perportedly fixed; removing these causes lua script error --[[ -- at Template:Webarchive/testcases/Production; resolve that before deleting these tests if mw.ustring.find( url, "https://web.http", 1, true ) then -- track bug - TODO: IAbot bug; not known if the bug has been fixed; deferred track[categories.error] = 1; return inlineError (data.crit_err_msgs.iabot1); end if url == "https://web.archive.org/http:/" then -- track bug - TODO: IAbot bug; not known if the bug has been fixed; deferred track[categories.error] = 1; return inlineError (data.crit_err_msgs.iabot2); end ]] if not (url:lower():find ('^http') or url:find ('^//')) then return inlineError (data.crit_err_msgs.invalid_url ); end ulx.url1 = {} ulx.url1.url = url ulx.url1.extraurls = parseExtraArgs(args) local good = false; good, uri = pcall (mw.uri.new, ulx.url1.url); -- get a table of uri parts from this url; protected mode to prevent lua error when ulx.url1.url is malformed if not good or nil == uri.host then -- abandon when ulx.url1.url is malformed return inlineError (data.crit_err_msgs.invalid_url); end serviceName(uri.host, args.nolink) if args.date and args.date1 then -- Date argument return inlineError (data.crit_err_msgs.conflicting, {origin.date, origin.date1}); end date = args.date or args.date1; date = date and date:gsub (' +', ' '); -- replace multiple spaces with a single space if date and config.verifydates then if '*' == date then date = 'index'; ldf = 'iso'; -- set to default format elseif 'mdy' == date then date = nil; -- if date extracted from URL, ldf = 'mdy'; -- then |date=mdy overrides iso elseif 'dmy' == date then date = nil; -- if date extracted from URL, ldf = 'dmy'; -- then |date=dmy overrides iso elseif 'ymd' == date then date = nil; -- if date extracted from URL, ldf = 'ymd'; -- then |date=ymd overrides iso else date, ldf = decode_date (date); -- get an iso format date from date and get date's original format end end if 'wayback' == ulx.url1.service or 'locwebarchives' == ulx.url1.service or 'ukgwa' == ulx.url1.service then if date then if config.verifydates then if ldf then udate, msg = decodeWaybackDate (uri.path); -- get the url date in iso format and format of date in |date=; 'index' when wayback url date is * if not udate then -- this is the only 'fatal' error return return inlineError (data.crit_err_msgs[msg]); end if udate ~= date then -- date comparison using iso format dates date = udate; msg = table.concat ({ inlineRed (err_warn_msgs.mismatch, 'warning'), -- add warning message msg, -- add message if there is one }); end end end else -- no |date= udate, msg = decodeWaybackDate (uri.path); if not udate then -- this is the only 'fatal' error return return inlineError (data.crit_err_msgs[msg]); end if == udate then date = nil; -- unset else date = udate; end end elseif 'webcite' == ulx.url1.service then if date then if config.verifydates then if ldf then udate = decodeWebciteDate (uri.path); -- get the url date in iso format if 'query' ~= udate then -- skip if query if udate ~= date then -- date comparison using iso format dates date = udate; msg = table.concat ({ inlineRed (err_warn_msgs.mismatch, 'warning'), }); end end end end else date = decodeWebciteDate( uri.path, "iso" ) if date == "query" then date = nil; -- unset msg = inlineRed (err_warn_msgs.date_miss, 'warning'); elseif not date then -- invalid base62 string date = inlineRed (err_warn_msgs.date1, 'error'); end end elseif 'archiveis' == ulx.url1.service then if date then if config.verifydates then if ldf then udate, msg = decodeArchiveisDate (uri.path) -- get the url date in iso format if 'short link' ~= udate then -- skip if short link if udate ~= date then -- date comparison using iso format dates date = udate; msg = table.concat ({ inlineRed (err_warn_msgs.mismatch, 'warning'), -- add warning message msg, -- add message if there is one }); end end end end else -- no |date= udate, msg = decodeArchiveisDate( uri.path, "iso" ) if udate == "short link" then date = nil; -- unset msg = inlineRed (err_warn_msgs.date_miss, 'warning'); elseif == udate then date = nil; -- unset else date = udate; end end else -- some other service if not date then msg = inlineRed (err_warn_msgs.date_miss, 'warning'); end end if 'index' == date then ulx.url1.date = date .. (msg or ); -- create index + message (if there is one) elseif date then ulx.url1.date = makeDate (date, nil, nil, ldf) .. (msg or ); -- create a date in the wiki's local language + message (if there is one) else ulx.url1.date = msg; end format = args.format; -- Format argument if not format then format = "none" else for k, v in pairs (data.format_vals) do -- |format= accepts two specific values loop through a table of those values local found; -- declare a nil flag for _, p in ipairs (v) do -- loop through local language variants if format == p then -- when |format= value matches format = k; -- use name from table key found = true; -- declare found so that we can break out of outer for loop break; -- break out of inner for loop end end if found then break; end end if format == "addlpages" then if not ulx.url1.date then format = "none" end elseif format == "addlarchives" then format = "addlarchives" else format = "none" end end ulx.url1.format = format if args.title and args.title1 then -- Title argument return inlineError (data.crit_err_msgs.conflicting, {origin.title, origin.title1}); end ulx.url1.title = args.title or args.title1; local rend = createRendering() if not rend then return inlineError (data.crit_err_msgs.unknown); end return rend .. ((unnamed_params and inlineRed (err_warn_msgs.unnamed_params, 'warning')) or ) .. createTracking(); end --[[--------------------------< E X P O R T E D F U N C T I O N S >------------------------------------------ ]] return {webarchive = webarchive}; Deadline Hollywood (August 1, 2012). Retrieved on August 2, 2012.

References[]

  • Biskind, Peter (1998). Easy Riders, Raging Bulls. New York: Simon & Schuster. ISBN 0-684-80996-6. 
  • Evans, Mike (2006). The Making of Raging Bull. London: Unanimous Ltd. ISBN 1-903318-83-1. 


External links[]

Template:Wikiquote

  • Raging Bull at the Internet Movie Database (IMDb)
  • Raging Bull at the American Film Institute Catalog of Motion Pictures
  • Template:Tcmdb title
  • Raging Bull at FilmSite.org
  • Template:Mojo title
  • Template:Rotten-tomatoes
  • Template:Metacritic film
  • Template:Allmovie title

Template:Martin Scorsese Template:Paul Schrader Template:Sight and Sound Poll Template:Los Angeles Film Critics Association Award for Best Film

Community content is available under CC-BY-SA unless otherwise noted.