git-remote-mediawiki: actually send empty comment when they're empty
[git/git.git] / contrib / mw-to-git / git-remote-mediawiki
CommitLineData
428c995c
JN
1#! /usr/bin/perl
2
3# Copyright (C) 2011
4# Jérémie Nikaes <jeremie.nikaes@ensimag.imag.fr>
5# Arnaud Lacurie <arnaud.lacurie@ensimag.imag.fr>
6# Claire Fousse <claire.fousse@ensimag.imag.fr>
7# David Amouyal <david.amouyal@ensimag.imag.fr>
8# Matthieu Moy <matthieu.moy@grenoble-inp.fr>
9# License: GPL v2 or later
10
11# Gateway between Git and MediaWiki.
12# https://github.com/Bibzball/Git-Mediawiki/wiki
13#
14# Known limitations:
15#
5a29217d
MM
16# - Several strategies are provided to fetch modifications from the
17# wiki, but no automatic heuristics is provided, the user has
18# to understand and chose which strategy is appropriate for him.
428c995c
JN
19#
20# - Git renames could be turned into MediaWiki renames (see TODO
21# below)
22#
428c995c
JN
23# - No way to import "one page, and all pages included in it"
24#
25# - Multiple remote MediaWikis have not been very well tested.
26
27use strict;
28use MediaWiki::API;
29use DateTime::Format::ISO8601;
428c995c 30
721a533f 31# By default, use UTF-8 to communicate with Git and the user
428c995c 32binmode STDERR, ":utf8";
721a533f 33binmode STDOUT, ":utf8";
428c995c
JN
34
35use URI::Escape;
a7271ad1
MM
36use IPC::Open2;
37
428c995c
JN
38use warnings;
39
40# Mediawiki filenames can contain forward slashes. This variable decides by which pattern they should be replaced
41use constant SLASH_REPLACEMENT => "%2F";
42
43# It's not always possible to delete pages (may require some
44# priviledges). Deleted pages are replaced with this content.
45use constant DELETED_CONTENT => "[[Category:Deleted]]\n";
46
47# It's not possible to create empty pages. New empty files in Git are
48# sent with this content instead.
49use constant EMPTY_CONTENT => "<!-- empty page -->\n";
50
51# used to reflect file creation or deletion in diff.
52use constant NULL_SHA1 => "0000000000000000000000000000000000000000";
53
dcb1ea62
MM
54# Used on Git's side to reflect empty edit messages on the wiki
55use constant EMPTY_MESSAGE => '*Empty MediaWiki Message*';
56
428c995c
JN
57my $remotename = $ARGV[0];
58my $url = $ARGV[1];
59
60# Accept both space-separated and multiple keys in config file.
61# Spaces should be written as _ anyway because we'll use chomp.
62my @tracked_pages = split(/[ \n]/, run_git("config --get-all remote.". $remotename .".pages"));
63chomp(@tracked_pages);
64
65# Just like @tracked_pages, but for MediaWiki categories.
66my @tracked_categories = split(/[ \n]/, run_git("config --get-all remote.". $remotename .".categories"));
67chomp(@tracked_categories);
68
6a9e55b0
PV
69# Import media files too.
70my $import_media = run_git("config --get --bool remote.". $remotename .".mediaimport");
71chomp($import_media);
72$import_media = ($import_media eq "true");
73
428c995c 74my $wiki_login = run_git("config --get remote.". $remotename .".mwLogin");
2da7830d 75# Note: mwPassword is discourraged. Use the credential system instead.
428c995c 76my $wiki_passwd = run_git("config --get remote.". $remotename .".mwPassword");
1d6abac9 77my $wiki_domain = run_git("config --get remote.". $remotename .".mwDomain");
428c995c
JN
78chomp($wiki_login);
79chomp($wiki_passwd);
1d6abac9 80chomp($wiki_domain);
428c995c
JN
81
82# Import only last revisions (both for clone and fetch)
83my $shallow_import = run_git("config --get --bool remote.". $remotename .".shallow");
84chomp($shallow_import);
85$shallow_import = ($shallow_import eq "true");
86
5a29217d
MM
87# Fetch (clone and pull) by revisions instead of by pages. This behavior
88# is more efficient when we have a wiki with lots of pages and we fetch
89# the revisions quite often so that they concern only few pages.
90# Possible values:
91# - by_rev: perform one query per new revision on the remote wiki
92# - by_page: query each tracked page for new revision
93my $fetch_strategy = run_git("config --get remote.$remotename.fetchStrategy");
94unless ($fetch_strategy) {
95 $fetch_strategy = run_git("config --get mediawiki.fetchStrategy");
96}
97chomp($fetch_strategy);
98unless ($fetch_strategy) {
99 $fetch_strategy = "by_page";
100}
101
93f0d338
MM
102# Dumb push: don't update notes and mediawiki ref to reflect the last push.
103#
104# Configurable with mediawiki.dumbPush, or per-remote with
105# remote.<remotename>.dumbPush.
106#
107# This means the user will have to re-import the just-pushed
108# revisions. On the other hand, this means that the Git revisions
109# corresponding to MediaWiki revisions are all imported from the wiki,
110# regardless of whether they were initially created in Git or from the
111# web interface, hence all users will get the same history (i.e. if
112# the push from Git to MediaWiki loses some information, everybody
113# will get the history with information lost). If the import is
114# deterministic, this means everybody gets the same sha1 for each
115# MediaWiki revision.
116my $dumb_push = run_git("config --get --bool remote.$remotename.dumbPush");
117unless ($dumb_push) {
118 $dumb_push = run_git("config --get --bool mediawiki.dumbPush");
119}
120chomp($dumb_push);
121$dumb_push = ($dumb_push eq "true");
122
428c995c
JN
123my $wiki_name = $url;
124$wiki_name =~ s/[^\/]*:\/\///;
9fb79503
MM
125# If URL is like http://user:password@example.com/, we clearly don't
126# want the password in $wiki_name. While we're there, also remove user
127# and '@' sign, to avoid author like MWUser@HTTPUser@host.com
128$wiki_name =~ s/^.*@//;
428c995c
JN
129
130# Commands parser
131my $entry;
132my @cmd;
133while (<STDIN>) {
134 chomp;
135 @cmd = split(/ /);
136 if (defined($cmd[0])) {
137 # Line not blank
138 if ($cmd[0] eq "capabilities") {
139 die("Too many arguments for capabilities") unless (!defined($cmd[1]));
140 mw_capabilities();
141 } elsif ($cmd[0] eq "list") {
142 die("Too many arguments for list") unless (!defined($cmd[2]));
143 mw_list($cmd[1]);
144 } elsif ($cmd[0] eq "import") {
145 die("Invalid arguments for import") unless ($cmd[1] ne "" && !defined($cmd[2]));
146 mw_import($cmd[1]);
147 } elsif ($cmd[0] eq "option") {
148 die("Too many arguments for option") unless ($cmd[1] ne "" && $cmd[2] ne "" && !defined($cmd[3]));
149 mw_option($cmd[1],$cmd[2]);
150 } elsif ($cmd[0] eq "push") {
151 mw_push($cmd[1]);
152 } else {
153 print STDERR "Unknown command. Aborting...\n";
154 last;
155 }
156 } else {
157 # blank line: we should terminate
158 last;
159 }
160
161 BEGIN { $| = 1 } # flush STDOUT, to make sure the previous
162 # command is fully processed.
163}
164
165########################## Functions ##############################
166
a7271ad1
MM
167## credential API management (generic functions)
168
169sub credential_from_url {
170 my $url = shift;
171 my $parsed = URI->new($url);
172 my %credential;
173
174 if ($parsed->scheme) {
175 $credential{protocol} = $parsed->scheme;
176 }
177 if ($parsed->host) {
178 $credential{host} = $parsed->host;
179 }
180 if ($parsed->path) {
181 $credential{path} = $parsed->path;
182 }
183 if ($parsed->userinfo) {
184 if ($parsed->userinfo =~ /([^:]*):(.*)/) {
185 $credential{username} = $1;
186 $credential{password} = $2;
187 } else {
188 $credential{username} = $parsed->userinfo;
189 }
190 }
191
192 return %credential;
193}
194
195sub credential_read {
196 my %credential;
197 my $reader = shift;
198 my $op = shift;
199 while (<$reader>) {
200 my ($key, $value) = /([^=]*)=(.*)/;
201 if (not defined $key) {
202 die "ERROR receiving response from git credential $op:\n$_\n";
203 }
204 $credential{$key} = $value;
205 }
206 return %credential;
207}
208
209sub credential_write {
210 my $credential = shift;
211 my $writer = shift;
212 while (my ($key, $value) = each(%$credential) ) {
213 if ($value) {
214 print $writer "$key=$value\n";
215 }
216 }
217}
218
219sub credential_run {
220 my $op = shift;
221 my $credential = shift;
222 my $pid = open2(my $reader, my $writer, "git credential $op");
223 credential_write($credential, $writer);
224 print $writer "\n";
225 close($writer);
226
227 if ($op eq "fill") {
228 %$credential = credential_read($reader, $op);
229 } else {
230 if (<$reader>) {
231 die "ERROR while running git credential $op:\n$_";
232 }
233 }
234 close($reader);
235 waitpid($pid, 0);
236 my $child_exit_status = $? >> 8;
237 if ($child_exit_status != 0) {
238 die "'git credential $op' failed with code $child_exit_status.";
239 }
240}
241
428c995c
JN
242# MediaWiki API instance, created lazily.
243my $mediawiki;
244
245sub mw_connect_maybe {
246 if ($mediawiki) {
a7271ad1 247 return;
428c995c
JN
248 }
249 $mediawiki = MediaWiki::API->new;
250 $mediawiki->{config}->{api_url} = "$url/api.php";
251 if ($wiki_login) {
a7271ad1
MM
252 my %credential = credential_from_url($url);
253 $credential{username} = $wiki_login;
254 $credential{password} = $wiki_passwd;
255 credential_run("fill", \%credential);
256 my $request = {lgname => $credential{username},
257 lgpassword => $credential{password},
258 lgdomain => $wiki_domain};
259 if ($mediawiki->login($request)) {
260 credential_run("approve", \%credential);
261 print STDERR "Logged in mediawiki user \"$credential{username}\".\n";
428c995c 262 } else {
a7271ad1
MM
263 print STDERR "Failed to log in mediawiki user \"$credential{username}\" on $url\n";
264 print STDERR " (error " .
265 $mediawiki->{error}->{code} . ': ' .
266 $mediawiki->{error}->{details} . ")\n";
267 credential_run("reject", \%credential);
268 exit 1;
428c995c
JN
269 }
270 }
271}
272
9cb74f36
MM
273## Functions for listing pages on the remote wiki
274sub get_mw_tracked_pages {
275 my $pages = shift;
6df7e0df
MM
276 get_mw_page_list(\@tracked_pages, $pages);
277}
278
279sub get_mw_page_list {
280 my $page_list = shift;
281 my $pages = shift;
282 my @some_pages = @$page_list;
9cb74f36
MM
283 while (@some_pages) {
284 my $last = 50;
285 if ($#some_pages < $last) {
286 $last = $#some_pages;
287 }
288 my @slice = @some_pages[0..$last];
289 get_mw_first_pages(\@slice, $pages);
290 @some_pages = @some_pages[51..$#some_pages];
291 }
292}
293
294sub get_mw_tracked_categories {
295 my $pages = shift;
296 foreach my $category (@tracked_categories) {
297 if (index($category, ':') < 0) {
298 # Mediawiki requires the Category
299 # prefix, but let's not force the user
300 # to specify it.
301 $category = "Category:" . $category;
302 }
303 my $mw_pages = $mediawiki->list( {
304 action => 'query',
305 list => 'categorymembers',
306 cmtitle => $category,
307 cmlimit => 'max' } )
308 || die $mediawiki->{error}->{code} . ': '
309 . $mediawiki->{error}->{details};
310 foreach my $page (@{$mw_pages}) {
311 $pages->{$page->{title}} = $page;
312 }
313 }
314}
315
316sub get_mw_all_pages {
317 my $pages = shift;
318 # No user-provided list, get the list of pages from the API.
319 my $mw_pages = $mediawiki->list({
320 action => 'query',
321 list => 'allpages',
322 aplimit => 'max'
323 });
324 if (!defined($mw_pages)) {
325 print STDERR "fatal: could not get the list of wiki pages.\n";
326 print STDERR "fatal: '$url' does not appear to be a mediawiki\n";
327 print STDERR "fatal: make sure '$url/api.php' is a valid page.\n";
328 exit 1;
329 }
330 foreach my $page (@{$mw_pages}) {
331 $pages->{$page->{title}} = $page;
332 }
333}
334
335# queries the wiki for a set of pages. Meant to be used within a loop
336# querying the wiki for slices of page list.
428c995c
JN
337sub get_mw_first_pages {
338 my $some_pages = shift;
339 my @some_pages = @{$some_pages};
340
341 my $pages = shift;
342
343 # pattern 'page1|page2|...' required by the API
344 my $titles = join('|', @some_pages);
345
346 my $mw_pages = $mediawiki->api({
347 action => 'query',
348 titles => $titles,
349 });
350 if (!defined($mw_pages)) {
351 print STDERR "fatal: could not query the list of wiki pages.\n";
352 print STDERR "fatal: '$url' does not appear to be a mediawiki\n";
353 print STDERR "fatal: make sure '$url/api.php' is a valid page.\n";
354 exit 1;
355 }
356 while (my ($id, $page) = each(%{$mw_pages->{query}->{pages}})) {
357 if ($id < 0) {
358 print STDERR "Warning: page $page->{title} not found on wiki\n";
359 } else {
360 $pages->{$page->{title}} = $page;
361 }
362 }
363}
364
9cb74f36 365# Get the list of pages to be fetched according to configuration.
428c995c
JN
366sub get_mw_pages {
367 mw_connect_maybe();
368
369 my %pages; # hash on page titles to avoid duplicates
370 my $user_defined;
371 if (@tracked_pages) {
372 $user_defined = 1;
373 # The user provided a list of pages titles, but we
374 # still need to query the API to get the page IDs.
9cb74f36 375 get_mw_tracked_pages(\%pages);
428c995c
JN
376 }
377 if (@tracked_categories) {
378 $user_defined = 1;
9cb74f36 379 get_mw_tracked_categories(\%pages);
428c995c
JN
380 }
381 if (!$user_defined) {
9cb74f36 382 get_mw_all_pages(\%pages);
428c995c 383 }
6a9e55b0
PV
384 if ($import_media) {
385 print STDERR "Getting media files for selected pages...\n";
386 if ($user_defined) {
387 get_linked_mediafiles(\%pages);
388 } else {
389 get_all_mediafiles(\%pages);
428c995c
JN
390 }
391 }
b1ede9a9 392 return %pages;
428c995c
JN
393}
394
b3d98595
NKT
395# usage: $out = run_git("command args");
396# $out = run_git("command args", "raw"); # don't interpret output as UTF-8.
428c995c 397sub run_git {
b3d98595
NKT
398 my $args = shift;
399 my $encoding = (shift || "encoding(UTF-8)");
400 open(my $git, "-|:$encoding", "git " . $args);
428c995c
JN
401 my $res = do { local $/; <$git> };
402 close($git);
403
404 return $res;
405}
406
407
6a9e55b0
PV
408sub get_all_mediafiles {
409 my $pages = shift;
410 # Attach list of all pages for media files from the API,
411 # they are in a different namespace, only one namespace
412 # can be queried at the same moment
413 my $mw_pages = $mediawiki->list({
414 action => 'query',
415 list => 'allpages',
416 apnamespace => get_mw_namespace_id("File"),
417 aplimit => 'max'
418 });
419 if (!defined($mw_pages)) {
420 print STDERR "fatal: could not get the list of pages for media files.\n";
421 print STDERR "fatal: '$url' does not appear to be a mediawiki\n";
422 print STDERR "fatal: make sure '$url/api.php' is a valid page.\n";
423 exit 1;
424 }
425 foreach my $page (@{$mw_pages}) {
426 $pages->{$page->{title}} = $page;
427 }
428}
429
430sub get_linked_mediafiles {
431 my $pages = shift;
432 my @titles = map $_->{title}, values(%{$pages});
433
434 # The query is split in small batches because of the MW API limit of
435 # the number of links to be returned (500 links max).
436 my $batch = 10;
437 while (@titles) {
438 if ($#titles < $batch) {
439 $batch = $#titles;
440 }
441 my @slice = @titles[0..$batch];
442
443 # pattern 'page1|page2|...' required by the API
444 my $mw_titles = join('|', @slice);
445
446 # Media files could be included or linked from
447 # a page, get all related
448 my $query = {
449 action => 'query',
450 prop => 'links|images',
451 titles => $mw_titles,
452 plnamespace => get_mw_namespace_id("File"),
453 pllimit => 'max'
454 };
455 my $result = $mediawiki->api($query);
456
457 while (my ($id, $page) = each(%{$result->{query}->{pages}})) {
6df7e0df 458 my @media_titles;
6a9e55b0
PV
459 if (defined($page->{links})) {
460 my @link_titles = map $_->{title}, @{$page->{links}};
6df7e0df 461 push(@media_titles, @link_titles);
6a9e55b0
PV
462 }
463 if (defined($page->{images})) {
464 my @image_titles = map $_->{title}, @{$page->{images}};
6df7e0df 465 push(@media_titles, @image_titles);
6a9e55b0 466 }
6df7e0df
MM
467 if (@media_titles) {
468 get_mw_page_list(\@media_titles, $pages);
6a9e55b0
PV
469 }
470 }
471
472 @titles = @titles[($batch+1)..$#titles];
473 }
474}
475
476sub get_mw_mediafile_for_page_revision {
477 # Name of the file on Wiki, with the prefix.
6df7e0df 478 my $filename = shift;
6a9e55b0
PV
479 my $timestamp = shift;
480 my %mediafile;
481
6df7e0df
MM
482 # Search if on a media file with given timestamp exists on
483 # MediaWiki. In that case download the file.
6a9e55b0
PV
484 my $query = {
485 action => 'query',
486 prop => 'imageinfo',
6df7e0df 487 titles => "File:" . $filename,
6a9e55b0
PV
488 iistart => $timestamp,
489 iiend => $timestamp,
490 iiprop => 'timestamp|archivename|url',
491 iilimit => 1
492 };
493 my $result = $mediawiki->api($query);
494
6df7e0df 495 my ($fileid, $file) = each( %{$result->{query}->{pages}} );
6a9e55b0
PV
496 # If not defined it means there is no revision of the file for
497 # given timestamp.
498 if (defined($file->{imageinfo})) {
6a9e55b0
PV
499 $mediafile{title} = $filename;
500
501 my $fileinfo = pop(@{$file->{imageinfo}});
502 $mediafile{timestamp} = $fileinfo->{timestamp};
6df7e0df
MM
503 # Mediawiki::API's download function doesn't support https URLs
504 # and can't download old versions of files.
505 print STDERR "\tDownloading file $mediafile{title}, version $mediafile{timestamp}\n";
506 $mediafile{content} = download_mw_mediafile($fileinfo->{url});
6a9e55b0
PV
507 }
508 return %mediafile;
509}
510
6df7e0df 511sub download_mw_mediafile {
6a9e55b0 512 my $url = shift;
6a9e55b0 513
6df7e0df
MM
514 my $response = $mediawiki->{ua}->get($url);
515 if ($response->code == 200) {
516 return $response->decoded_content;
6a9e55b0 517 } else {
6df7e0df
MM
518 print STDERR "Error downloading mediafile from :\n";
519 print STDERR "URL: $url\n";
520 print STDERR "Server response: " . $response->code . " " . $response->message . "\n";
6a9e55b0 521 exit 1;
6a9e55b0
PV
522 }
523}
524
428c995c
JN
525sub get_last_local_revision {
526 # Get note regarding last mediawiki revision
527 my $note = run_git("notes --ref=$remotename/mediawiki show refs/mediawiki/$remotename/master 2>/dev/null");
528 my @note_info = split(/ /, $note);
529
530 my $lastrevision_number;
531 if (!(defined($note_info[0]) && $note_info[0] eq "mediawiki_revision:")) {
532 print STDERR "No previous mediawiki revision found";
533 $lastrevision_number = 0;
534 } else {
535 # Notes are formatted : mediawiki_revision: #number
536 $lastrevision_number = $note_info[1];
537 chomp($lastrevision_number);
538 print STDERR "Last local mediawiki revision found is $lastrevision_number";
539 }
540 return $lastrevision_number;
541}
542
3c1ed90e
MM
543# Remember the timestamp corresponding to a revision id.
544my %basetimestamps;
545
5a29217d
MM
546# Get the last remote revision without taking in account which pages are
547# tracked or not. This function makes a single request to the wiki thus
548# avoid a loop onto all tracked pages. This is useful for the fetch-by-rev
549# option.
550sub get_last_global_remote_rev {
551 mw_connect_maybe();
552
553 my $query = {
554 action => 'query',
555 list => 'recentchanges',
556 prop => 'revisions',
557 rclimit => '1',
558 rcdir => 'older',
559 };
560 my $result = $mediawiki->api($query);
561 return $result->{query}->{recentchanges}[0]->{revid};
562}
563
564# Get the last remote revision concerning the tracked pages and the tracked
565# categories.
428c995c
JN
566sub get_last_remote_revision {
567 mw_connect_maybe();
568
b1ede9a9
MM
569 my %pages_hash = get_mw_pages();
570 my @pages = values(%pages_hash);
428c995c
JN
571
572 my $max_rev_num = 0;
573
574 foreach my $page (@pages) {
575 my $id = $page->{pageid};
576
577 my $query = {
578 action => 'query',
579 prop => 'revisions',
3c1ed90e 580 rvprop => 'ids|timestamp',
428c995c
JN
581 pageids => $id,
582 };
583
584 my $result = $mediawiki->api($query);
585
586 my $lastrev = pop(@{$result->{query}->{pages}->{$id}->{revisions}});
587
3c1ed90e
MM
588 $basetimestamps{$lastrev->{revid}} = $lastrev->{timestamp};
589
428c995c
JN
590 $max_rev_num = ($lastrev->{revid} > $max_rev_num ? $lastrev->{revid} : $max_rev_num);
591 }
592
593 print STDERR "Last remote revision found is $max_rev_num.\n";
594 return $max_rev_num;
595}
596
597# Clean content before sending it to MediaWiki
598sub mediawiki_clean {
599 my $string = shift;
600 my $page_created = shift;
601 # Mediawiki does not allow blank space at the end of a page and ends with a single \n.
602 # This function right trims a string and adds a \n at the end to follow this rule
603 $string =~ s/\s+$//;
604 if ($string eq "" && $page_created) {
605 # Creating empty pages is forbidden.
606 $string = EMPTY_CONTENT;
607 }
608 return $string."\n";
609}
610
611# Filter applied on MediaWiki data before adding them to Git
612sub mediawiki_smudge {
613 my $string = shift;
614 if ($string eq EMPTY_CONTENT) {
615 $string = "";
616 }
617 # This \n is important. This is due to mediawiki's way to handle end of files.
618 return $string."\n";
619}
620
621sub mediawiki_clean_filename {
622 my $filename = shift;
623 $filename =~ s/@{[SLASH_REPLACEMENT]}/\//g;
624 # [, ], |, {, and } are forbidden by MediaWiki, even URL-encoded.
625 # Do a variant of URL-encoding, i.e. looks like URL-encoding,
626 # but with _ added to prevent MediaWiki from thinking this is
627 # an actual special character.
628 $filename =~ s/[\[\]\{\}\|]/sprintf("_%%_%x", ord($&))/ge;
629 # If we use the uri escape before
630 # we should unescape here, before anything
631
632 return $filename;
633}
634
635sub mediawiki_smudge_filename {
636 my $filename = shift;
637 $filename =~ s/\//@{[SLASH_REPLACEMENT]}/g;
638 $filename =~ s/ /_/g;
639 # Decode forbidden characters encoded in mediawiki_clean_filename
640 $filename =~ s/_%_([0-9a-fA-F][0-9a-fA-F])/sprintf("%c", hex($1))/ge;
641 return $filename;
642}
643
644sub literal_data {
645 my ($content) = @_;
646 print STDOUT "data ", bytes::length($content), "\n", $content;
647}
648
6a9e55b0
PV
649sub literal_data_raw {
650 # Output possibly binary content.
651 my ($content) = @_;
652 # Avoid confusion between size in bytes and in characters
653 utf8::downgrade($content);
654 binmode STDOUT, ":raw";
655 print STDOUT "data ", bytes::length($content), "\n", $content;
656 binmode STDOUT, ":utf8";
657}
658
428c995c
JN
659sub mw_capabilities {
660 # Revisions are imported to the private namespace
661 # refs/mediawiki/$remotename/ by the helper and fetched into
662 # refs/remotes/$remotename later by fetch.
663 print STDOUT "refspec refs/heads/*:refs/mediawiki/$remotename/*\n";
664 print STDOUT "import\n";
665 print STDOUT "list\n";
666 print STDOUT "push\n";
667 print STDOUT "\n";
668}
669
670sub mw_list {
671 # MediaWiki do not have branches, we consider one branch arbitrarily
672 # called master, and HEAD pointing to it.
673 print STDOUT "? refs/heads/master\n";
674 print STDOUT "\@refs/heads/master HEAD\n";
675 print STDOUT "\n";
676}
677
678sub mw_option {
679 print STDERR "remote-helper command 'option $_[0]' not yet implemented\n";
680 print STDOUT "unsupported\n";
681}
682
683sub fetch_mw_revisions_for_page {
684 my $page = shift;
685 my $id = shift;
686 my $fetch_from = shift;
687 my @page_revs = ();
688 my $query = {
689 action => 'query',
690 prop => 'revisions',
691 rvprop => 'ids',
692 rvdir => 'newer',
693 rvstartid => $fetch_from,
694 rvlimit => 500,
695 pageids => $id,
696 };
697
698 my $revnum = 0;
699 # Get 500 revisions at a time due to the mediawiki api limit
700 while (1) {
701 my $result = $mediawiki->api($query);
702
703 # Parse each of those 500 revisions
704 foreach my $revision (@{$result->{query}->{pages}->{$id}->{revisions}}) {
705 my $page_rev_ids;
706 $page_rev_ids->{pageid} = $page->{pageid};
707 $page_rev_ids->{revid} = $revision->{revid};
708 push(@page_revs, $page_rev_ids);
709 $revnum++;
710 }
711 last unless $result->{'query-continue'};
712 $query->{rvstartid} = $result->{'query-continue'}->{revisions}->{rvstartid};
713 }
714 if ($shallow_import && @page_revs) {
715 print STDERR " Found 1 revision (shallow import).\n";
716 @page_revs = sort {$b->{revid} <=> $a->{revid}} (@page_revs);
717 return $page_revs[0];
718 }
719 print STDERR " Found ", $revnum, " revision(s).\n";
720 return @page_revs;
721}
722
723sub fetch_mw_revisions {
724 my $pages = shift; my @pages = @{$pages};
725 my $fetch_from = shift;
726
727 my @revisions = ();
728 my $n = 1;
729 foreach my $page (@pages) {
730 my $id = $page->{pageid};
731
732 print STDERR "page $n/", scalar(@pages), ": ". $page->{title} ."\n";
733 $n++;
734 my @page_revs = fetch_mw_revisions_for_page($page, $id, $fetch_from);
735 @revisions = (@page_revs, @revisions);
736 }
737
738 return ($n, @revisions);
739}
740
741sub import_file_revision {
742 my $commit = shift;
743 my %commit = %{$commit};
744 my $full_import = shift;
745 my $n = shift;
6a9e55b0
PV
746 my $mediafile = shift;
747 my %mediafile;
748 if ($mediafile) {
749 %mediafile = %{$mediafile};
750 }
428c995c
JN
751
752 my $title = $commit{title};
753 my $comment = $commit{comment};
754 my $content = $commit{content};
755 my $author = $commit{author};
756 my $date = $commit{date};
757
758 print STDOUT "commit refs/mediawiki/$remotename/master\n";
759 print STDOUT "mark :$n\n";
760 print STDOUT "committer $author <$author\@$wiki_name> ", $date->epoch, " +0000\n";
761 literal_data($comment);
762
763 # If it's not a clone, we need to know where to start from
764 if (!$full_import && $n == 1) {
765 print STDOUT "from refs/mediawiki/$remotename/master^0\n";
766 }
767 if ($content ne DELETED_CONTENT) {
768 print STDOUT "M 644 inline $title.mw\n";
769 literal_data($content);
6a9e55b0
PV
770 if (%mediafile) {
771 print STDOUT "M 644 inline $mediafile{title}\n";
772 literal_data_raw($mediafile{content});
773 }
428c995c
JN
774 print STDOUT "\n\n";
775 } else {
776 print STDOUT "D $title.mw\n";
777 }
778
779 # mediawiki revision number in the git note
780 if ($full_import && $n == 1) {
781 print STDOUT "reset refs/notes/$remotename/mediawiki\n";
782 }
783 print STDOUT "commit refs/notes/$remotename/mediawiki\n";
784 print STDOUT "committer $author <$author\@$wiki_name> ", $date->epoch, " +0000\n";
785 literal_data("Note added by git-mediawiki during import");
786 if (!$full_import && $n == 1) {
787 print STDOUT "from refs/notes/$remotename/mediawiki^0\n";
788 }
789 print STDOUT "N inline :$n\n";
790 literal_data("mediawiki_revision: " . $commit{mw_revision});
791 print STDOUT "\n\n";
792}
793
794# parse a sequence of
795# <cmd> <arg1>
796# <cmd> <arg2>
797# \n
798# (like batch sequence of import and sequence of push statements)
799sub get_more_refs {
800 my $cmd = shift;
801 my @refs;
802 while (1) {
803 my $line = <STDIN>;
804 if ($line =~ m/^$cmd (.*)$/) {
805 push(@refs, $1);
806 } elsif ($line eq "\n") {
807 return @refs;
808 } else {
809 die("Invalid command in a '$cmd' batch: ". $_);
810 }
811 }
812}
813
814sub mw_import {
815 # multiple import commands can follow each other.
816 my @refs = (shift, get_more_refs("import"));
817 foreach my $ref (@refs) {
818 mw_import_ref($ref);
819 }
820 print STDOUT "done\n";
821}
822
823sub mw_import_ref {
824 my $ref = shift;
825 # The remote helper will call "import HEAD" and
826 # "import refs/heads/master".
827 # Since HEAD is a symbolic ref to master (by convention,
828 # followed by the output of the command "list" that we gave),
829 # we don't need to do anything in this case.
830 if ($ref eq "HEAD") {
831 return;
832 }
833
834 mw_connect_maybe();
835
428c995c
JN
836 print STDERR "Searching revisions...\n";
837 my $last_local = get_last_local_revision();
838 my $fetch_from = $last_local + 1;
839 if ($fetch_from == 1) {
840 print STDERR ", fetching from beginning.\n";
841 } else {
842 print STDERR ", fetching from here.\n";
843 }
428c995c 844
5a29217d
MM
845 my $n = 0;
846 if ($fetch_strategy eq "by_rev") {
847 print STDERR "Fetching & writing export data by revs...\n";
848 $n = mw_import_ref_by_revs($fetch_from);
849 } elsif ($fetch_strategy eq "by_page") {
850 print STDERR "Fetching & writing export data by pages...\n";
851 $n = mw_import_ref_by_pages($fetch_from);
852 } else {
853 print STDERR "fatal: invalid fetch strategy \"$fetch_strategy\".\n";
854 print STDERR "Check your configuration variables remote.$remotename.fetchStrategy and mediawiki.fetchStrategy\n";
855 exit 1;
856 }
857
858 if ($fetch_from == 1 && $n == 0) {
859 print STDERR "You appear to have cloned an empty MediaWiki.\n";
860 # Something has to be done remote-helper side. If nothing is done, an error is
861 # thrown saying that HEAD is refering to unknown object 0000000000000000000
862 # and the clone fails.
863 }
864}
865
866sub mw_import_ref_by_pages {
867
868 my $fetch_from = shift;
869 my %pages_hash = get_mw_pages();
870 my @pages = values(%pages_hash);
871
428c995c
JN
872 my ($n, @revisions) = fetch_mw_revisions(\@pages, $fetch_from);
873
4465b6d6
MM
874 @revisions = sort {$a->{revid} <=> $b->{revid}} @revisions;
875 my @revision_ids = map $_->{revid}, @revisions;
876
312fa9aa
MM
877 return mw_import_revids($fetch_from, \@revision_ids, \%pages_hash);
878}
879
5a29217d
MM
880sub mw_import_ref_by_revs {
881
882 my $fetch_from = shift;
883 my %pages_hash = get_mw_pages();
428c995c 884
5a29217d
MM
885 my $last_remote = get_last_global_remote_rev();
886 my @revision_ids = $fetch_from..$last_remote;
887 return mw_import_revids($fetch_from, \@revision_ids, \%pages_hash);
888}
889
890# Import revisions given in second argument (array of integers).
891# Only pages appearing in the third argument (hash indexed by page titles)
892# will be imported.
312fa9aa
MM
893sub mw_import_revids {
894 my $fetch_from = shift;
895 my $revision_ids = shift;
896 my $pages = shift;
897
898 my $n = 0;
5a29217d 899 my $n_actual = 0;
428c995c
JN
900 my $last_timestamp = 0; # Placeholer in case $rev->timestamp is undefined
901
312fa9aa 902 foreach my $pagerevid (@$revision_ids) {
428c995c
JN
903 # fetch the content of the pages
904 my $query = {
905 action => 'query',
906 prop => 'revisions',
907 rvprop => 'content|timestamp|comment|user|ids',
4465b6d6 908 revids => $pagerevid,
428c995c
JN
909 };
910
911 my $result = $mediawiki->api($query);
912
ff0bfd75
MM
913 if (!$result) {
914 die "Failed to retrieve modified page for revision $pagerevid";
915 }
916
917 if (!defined($result->{query}->{pages})) {
918 die "Invalid revision $pagerevid.";
919 }
920
4465b6d6
MM
921 my @result_pages = values(%{$result->{query}->{pages}});
922 my $result_page = $result_pages[0];
923 my $rev = $result_pages[0]->{revisions}->[0];
428c995c 924
5a29217d
MM
925 # Count page even if we skip it, since we display
926 # $n/$total and $total includes skipped pages.
428c995c
JN
927 $n++;
928
4465b6d6 929 my $page_title = $result_page->{title};
5a29217d
MM
930
931 if (!exists($pages->{$page_title})) {
932 print STDERR "$n/", scalar(@$revision_ids),
933 ": Skipping revision #$rev->{revid} of $page_title\n";
934 next;
935 }
936
937 $n_actual++;
938
428c995c
JN
939 my %commit;
940 $commit{author} = $rev->{user} || 'Anonymous';
dcb1ea62 941 $commit{comment} = $rev->{comment} || EMPTY_MESSAGE;
6a9e55b0 942 $commit{title} = mediawiki_smudge_filename($page_title);
4465b6d6 943 $commit{mw_revision} = $rev->{revid};
428c995c
JN
944 $commit{content} = mediawiki_smudge($rev->{'*'});
945
946 if (!defined($rev->{timestamp})) {
947 $last_timestamp++;
948 } else {
949 $last_timestamp = $rev->{timestamp};
950 }
951 $commit{date} = DateTime::Format::ISO8601->parse_datetime($last_timestamp);
952
6a9e55b0 953 # Differentiates classic pages and media files.
6df7e0df 954 my ($namespace, $filename) = $page_title =~ /^([^:]*):(.*)$/;
6a9e55b0 955 my %mediafile;
6df7e0df 956 if ($namespace && get_mw_namespace_id($namespace) == get_mw_namespace_id("File")) {
6a9e55b0
PV
957 %mediafile = get_mw_mediafile_for_page_revision($filename, $rev->{timestamp});
958 }
959 # If this is a revision of the media page for new version
960 # of a file do one common commit for both file and media page.
961 # Else do commit only for that page.
312fa9aa 962 print STDERR "$n/", scalar(@$revision_ids), ": Revision #$rev->{revid} of $commit{title}\n";
5a29217d 963 import_file_revision(\%commit, ($fetch_from == 1), $n_actual, \%mediafile);
428c995c
JN
964 }
965
5a29217d 966 return $n_actual;
428c995c
JN
967}
968
969sub error_non_fast_forward {
fd47d7b9
MM
970 my $advice = run_git("config --bool advice.pushNonFastForward");
971 chomp($advice);
972 if ($advice ne "false") {
973 # Native git-push would show this after the summary.
974 # We can't ask it to display it cleanly, so print it
975 # ourselves before.
976 print STDERR "To prevent you from losing history, non-fast-forward updates were rejected\n";
977 print STDERR "Merge the remote changes (e.g. 'git pull') before pushing again. See the\n";
978 print STDERR "'Note about fast-forwards' section of 'git push --help' for details.\n";
979 }
428c995c
JN
980 print STDOUT "error $_[0] \"non-fast-forward\"\n";
981 return 0;
982}
983
b3d98595
NKT
984sub mw_upload_file {
985 my $complete_file_name = shift;
986 my $new_sha1 = shift;
987 my $extension = shift;
988 my $file_deleted = shift;
989 my $summary = shift;
990 my $newrevid;
991 my $path = "File:" . $complete_file_name;
992 my %hashFiles = get_allowed_file_extensions();
993 if (!exists($hashFiles{$extension})) {
994 print STDERR "$complete_file_name is not a permitted file on this wiki.\n";
995 print STDERR "Check the configuration of file uploads in your mediawiki.\n";
996 return $newrevid;
997 }
998 # Deleting and uploading a file requires a priviledged user
999 if ($file_deleted) {
1000 mw_connect_maybe();
1001 my $query = {
1002 action => 'delete',
1003 title => $path,
1004 reason => $summary
1005 };
1006 if (!$mediawiki->edit($query)) {
1007 print STDERR "Failed to delete file on remote wiki\n";
1008 print STDERR "Check your permissions on the remote site. Error code:\n";
1009 print STDERR $mediawiki->{error}->{code} . ':' . $mediawiki->{error}->{details};
1010 exit 1;
1011 }
1012 } else {
1013 # Don't let perl try to interpret file content as UTF-8 => use "raw"
1014 my $content = run_git("cat-file blob $new_sha1", "raw");
1015 if ($content ne "") {
1016 mw_connect_maybe();
1017 $mediawiki->{config}->{upload_url} =
1018 "$url/index.php/Special:Upload";
1019 $mediawiki->edit({
1020 action => 'upload',
1021 filename => $complete_file_name,
1022 comment => $summary,
1023 file => [undef,
1024 $complete_file_name,
1025 Content => $content],
1026 ignorewarnings => 1,
1027 }, {
1028 skip_encoding => 1
1029 } ) || die $mediawiki->{error}->{code} . ':'
1030 . $mediawiki->{error}->{details};
1031 my $last_file_page = $mediawiki->get_page({title => $path});
1032 $newrevid = $last_file_page->{revid};
1033 print STDERR "Pushed file: $new_sha1 - $complete_file_name.\n";
1034 } else {
1035 print STDERR "Empty file $complete_file_name not pushed.\n";
1036 }
1037 }
1038 return $newrevid;
1039}
1040
428c995c
JN
1041sub mw_push_file {
1042 my $diff_info = shift;
1043 # $diff_info contains a string in this format:
1044 # 100644 100644 <sha1_of_blob_before_commit> <sha1_of_blob_now> <status>
1045 my @diff_info_split = split(/[ \t]/, $diff_info);
1046
1047 # Filename, including .mw extension
1048 my $complete_file_name = shift;
1049 # Commit message
1050 my $summary = shift;
93f0d338
MM
1051 # MediaWiki revision number. Keep the previous one by default,
1052 # in case there's no edit to perform.
b3d98595
NKT
1053 my $oldrevid = shift;
1054 my $newrevid;
428c995c 1055
dcb1ea62
MM
1056 if ($summary eq EMPTY_MESSAGE) {
1057 $summary = '';
1058 }
1059
428c995c
JN
1060 my $new_sha1 = $diff_info_split[3];
1061 my $old_sha1 = $diff_info_split[2];
1062 my $page_created = ($old_sha1 eq NULL_SHA1);
1063 my $page_deleted = ($new_sha1 eq NULL_SHA1);
1064 $complete_file_name = mediawiki_clean_filename($complete_file_name);
1065
b3d98595
NKT
1066 my ($title, $extension) = $complete_file_name =~ /^(.*)\.([^\.]*)$/;
1067 if (!defined($extension)) {
1068 $extension = "";
1069 }
1070 if ($extension eq "mw") {
428c995c
JN
1071 my $file_content;
1072 if ($page_deleted) {
1073 # Deleting a page usually requires
1074 # special priviledges. A common
1075 # convention is to replace the page
1076 # with this content instead:
1077 $file_content = DELETED_CONTENT;
1078 } else {
1079 $file_content = run_git("cat-file blob $new_sha1");
1080 }
1081
1082 mw_connect_maybe();
1083
1084 my $result = $mediawiki->edit( {
1085 action => 'edit',
1086 summary => $summary,
1087 title => $title,
b3d98595 1088 basetimestamp => $basetimestamps{$oldrevid},
428c995c
JN
1089 text => mediawiki_clean($file_content, $page_created),
1090 }, {
1091 skip_encoding => 1 # Helps with names with accentuated characters
3c1ed90e
MM
1092 });
1093 if (!$result) {
1094 if ($mediawiki->{error}->{code} == 3) {
1095 # edit conflicts, considered as non-fast-forward
1096 print STDERR 'Warning: Error ' .
1097 $mediawiki->{error}->{code} .
1098 ' from mediwiki: ' . $mediawiki->{error}->{details} .
1099 ".\n";
b3d98595 1100 return ($oldrevid, "non-fast-forward");
3c1ed90e
MM
1101 } else {
1102 # Other errors. Shouldn't happen => just die()
1103 die 'Fatal: Error ' .
1104 $mediawiki->{error}->{code} .
1105 ' from mediwiki: ' . $mediawiki->{error}->{details};
1106 }
1107 }
93f0d338 1108 $newrevid = $result->{edit}->{newrevid};
ac86ec0f 1109 print STDERR "Pushed file: $new_sha1 - $title\n";
428c995c 1110 } else {
b3d98595
NKT
1111 $newrevid = mw_upload_file($complete_file_name, $new_sha1,
1112 $extension, $page_deleted,
1113 $summary);
428c995c 1114 }
b3d98595 1115 $newrevid = ($newrevid or $oldrevid);
3c1ed90e 1116 return ($newrevid, "ok");
428c995c
JN
1117}
1118
1119sub mw_push {
1120 # multiple push statements can follow each other
1121 my @refsspecs = (shift, get_more_refs("push"));
428c995c
JN
1122 my $pushed;
1123 for my $refspec (@refsspecs) {
1124 my ($force, $local, $remote) = $refspec =~ /^(\+)?([^:]*):([^:]*)$/
1125 or die("Invalid refspec for push. Expected <src>:<dst> or +<src>:<dst>");
1126 if ($force) {
1127 print STDERR "Warning: forced push not allowed on a MediaWiki.\n";
1128 }
1129 if ($local eq "") {
1130 print STDERR "Cannot delete remote branch on a MediaWiki\n";
1131 print STDOUT "error $remote cannot delete\n";
1132 next;
1133 }
1134 if ($remote ne "refs/heads/master") {
1135 print STDERR "Only push to the branch 'master' is supported on a MediaWiki\n";
1136 print STDOUT "error $remote only master allowed\n";
1137 next;
1138 }
1139 if (mw_push_revision($local, $remote)) {
1140 $pushed = 1;
1141 }
1142 }
1143
1144 # Notify Git that the push is done
1145 print STDOUT "\n";
1146
93f0d338 1147 if ($pushed && $dumb_push) {
428c995c
JN
1148 print STDERR "Just pushed some revisions to MediaWiki.\n";
1149 print STDERR "The pushed revisions now have to be re-imported, and your current branch\n";
1150 print STDERR "needs to be updated with these re-imported commits. You can do this with\n";
1151 print STDERR "\n";
1152 print STDERR " git pull --rebase\n";
1153 print STDERR "\n";
1154 }
1155}
1156
1157sub mw_push_revision {
1158 my $local = shift;
1159 my $remote = shift; # actually, this has to be "refs/heads/master" at this point.
1160 my $last_local_revid = get_last_local_revision();
1161 print STDERR ".\n"; # Finish sentence started by get_last_local_revision()
1162 my $last_remote_revid = get_last_remote_revision();
93f0d338 1163 my $mw_revision = $last_remote_revid;
428c995c
JN
1164
1165 # Get sha1 of commit pointed by local HEAD
1166 my $HEAD_sha1 = run_git("rev-parse $local 2>/dev/null"); chomp($HEAD_sha1);
1167 # Get sha1 of commit pointed by remotes/$remotename/master
1168 my $remoteorigin_sha1 = run_git("rev-parse refs/remotes/$remotename/master 2>/dev/null");
1169 chomp($remoteorigin_sha1);
1170
1171 if ($last_local_revid > 0 &&
1172 $last_local_revid < $last_remote_revid) {
1173 return error_non_fast_forward($remote);
1174 }
1175
1176 if ($HEAD_sha1 eq $remoteorigin_sha1) {
1177 # nothing to push
1178 return 0;
1179 }
1180
1181 # Get every commit in between HEAD and refs/remotes/origin/master,
1182 # including HEAD and refs/remotes/origin/master
1183 my @commit_pairs = ();
1184 if ($last_local_revid > 0) {
1185 my $parsed_sha1 = $remoteorigin_sha1;
1186 # Find a path from last MediaWiki commit to pushed commit
1187 while ($parsed_sha1 ne $HEAD_sha1) {
1188 my @commit_info = grep(/^$parsed_sha1/, split(/\n/, run_git("rev-list --children $local")));
1189 if (!@commit_info) {
1190 return error_non_fast_forward($remote);
1191 }
1192 my @commit_info_split = split(/ |\n/, $commit_info[0]);
1193 # $commit_info_split[1] is the sha1 of the commit to export
1194 # $commit_info_split[0] is the sha1 of its direct child
1195 push(@commit_pairs, \@commit_info_split);
1196 $parsed_sha1 = $commit_info_split[1];
1197 }
1198 } else {
1199 # No remote mediawiki revision. Export the whole
1200 # history (linearized with --first-parent)
1201 print STDERR "Warning: no common ancestor, pushing complete history\n";
1202 my $history = run_git("rev-list --first-parent --children $local");
1203 my @history = split('\n', $history);
1204 @history = @history[1..$#history];
1205 foreach my $line (reverse @history) {
1206 my @commit_info_split = split(/ |\n/, $line);
1207 push(@commit_pairs, \@commit_info_split);
1208 }
1209 }
1210
1211 foreach my $commit_info_split (@commit_pairs) {
1212 my $sha1_child = @{$commit_info_split}[0];
1213 my $sha1_commit = @{$commit_info_split}[1];
1214 my $diff_infos = run_git("diff-tree -r --raw -z $sha1_child $sha1_commit");
1215 # TODO: we could detect rename, and encode them with a #redirect on the wiki.
1216 # TODO: for now, it's just a delete+add
1217 my @diff_info_list = split(/\0/, $diff_infos);
28c24bd7
MM
1218 # Keep the subject line of the commit message as mediawiki comment for the revision
1219 my $commit_msg = run_git("log --no-walk --format=\"%s\" $sha1_commit");
428c995c
JN
1220 chomp($commit_msg);
1221 # Push every blob
1222 while (@diff_info_list) {
3c1ed90e 1223 my $status;
428c995c
JN
1224 # git diff-tree -z gives an output like
1225 # <metadata>\0<filename1>\0
1226 # <metadata>\0<filename2>\0
1227 # and we've split on \0.
1228 my $info = shift(@diff_info_list);
1229 my $file = shift(@diff_info_list);
3c1ed90e
MM
1230 ($mw_revision, $status) = mw_push_file($info, $file, $commit_msg, $mw_revision);
1231 if ($status eq "non-fast-forward") {
1232 # we may already have sent part of the
1233 # commit to MediaWiki, but it's too
1234 # late to cancel it. Stop the push in
1235 # the middle, but still give an
1236 # accurate error message.
1237 return error_non_fast_forward($remote);
1238 }
1239 if ($status ne "ok") {
1240 die("Unknown error from mw_push_file()");
1241 }
93f0d338
MM
1242 }
1243 unless ($dumb_push) {
1244 run_git("notes --ref=$remotename/mediawiki add -m \"mediawiki_revision: $mw_revision\" $sha1_commit");
1245 run_git("update-ref -m \"Git-MediaWiki push\" refs/mediawiki/$remotename/master $sha1_commit $sha1_child");
428c995c
JN
1246 }
1247 }
1248
1249 print STDOUT "ok $remote\n";
1250 return 1;
1251}
b3d98595
NKT
1252
1253sub get_allowed_file_extensions {
1254 mw_connect_maybe();
1255
1256 my $query = {
1257 action => 'query',
1258 meta => 'siteinfo',
1259 siprop => 'fileextensions'
1260 };
1261 my $result = $mediawiki->api($query);
1262 my @file_extensions= map $_->{ext},@{$result->{query}->{fileextensions}};
1263 my %hashFile = map {$_ => 1}@file_extensions;
1264
1265 return %hashFile;
1266}
6a9e55b0 1267
6df7e0df
MM
1268# In memory cache for MediaWiki namespace ids.
1269my %namespace_id;
1270
1271# Namespaces whose id is cached in the configuration file
1272# (to avoid duplicates)
1273my %cached_mw_namespace_id;
1274
6a9e55b0
PV
1275# Return MediaWiki id for a canonical namespace name.
1276# Ex.: "File", "Project".
6a9e55b0
PV
1277sub get_mw_namespace_id {
1278 mw_connect_maybe();
1279 my $name = shift;
1280
1281 if (!exists $namespace_id{$name}) {
1282 # Look at configuration file, if the record for that namespace is
6df7e0df 1283 # already cached. Namespaces are stored in form:
6a9e55b0 1284 # "Name_of_namespace:Id_namespace", ex.: "File:6".
ac4bbb41 1285 my @temp = split(/[\n]/, run_git("config --get-all remote."
6df7e0df 1286 . $remotename .".namespaceCache"));
6a9e55b0
PV
1287 chomp(@temp);
1288 foreach my $ns (@temp) {
6df7e0df
MM
1289 my ($n, $id) = split(/:/, $ns);
1290 $namespace_id{$n} = $id;
1291 $cached_mw_namespace_id{$n} = 1;
6a9e55b0
PV
1292 }
1293 }
1294
1295 if (!exists $namespace_id{$name}) {
6df7e0df 1296 print STDERR "Namespace $name not found in cache, querying the wiki ...\n";
6a9e55b0
PV
1297 # NS not found => get namespace id from MW and store it in
1298 # configuration file.
1299 my $query = {
1300 action => 'query',
1301 meta => 'siteinfo',
1302 siprop => 'namespaces'
1303 };
1304 my $result = $mediawiki->api($query);
1305
1306 while (my ($id, $ns) = each(%{$result->{query}->{namespaces}})) {
6df7e0df
MM
1307 if (defined($ns->{id}) && defined($ns->{canonical})) {
1308 $namespace_id{$ns->{canonical}} = $ns->{id};
1309 if ($ns->{'*'}) {
1310 # alias (e.g. french Fichier: as alias for canonical File:)
1311 $namespace_id{$ns->{'*'}} = $ns->{id};
1312 }
1313 }
6a9e55b0
PV
1314 }
1315 }
1316
6df7e0df
MM
1317 my $id = $namespace_id{$name};
1318
1319 if (defined $id) {
1320 # Store explicitely requested namespaces on disk
1321 if (!exists $cached_mw_namespace_id{$name}) {
1322 run_git("config --add remote.". $remotename
1323 .".namespaceCache \"". $name .":". $id ."\"");
1324 $cached_mw_namespace_id{$name} = 1;
1325 }
1326 return $id;
6a9e55b0
PV
1327 } else {
1328 die "No such namespace $name on MediaWiki.";
1329 }
1330}