git-remote-mediawiki: fix incorrect test usage in test
[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
2045e293 69# Import media files on pull
6a9e55b0
PV
70my $import_media = run_git("config --get --bool remote.". $remotename .".mediaimport");
71chomp($import_media);
72$import_media = ($import_media eq "true");
73
2045e293
MM
74# Export media files on push
75my $export_media = run_git("config --get --bool remote.". $remotename .".mediaexport");
76chomp($export_media);
77$export_media = !($export_media eq "false");
78
428c995c 79my $wiki_login = run_git("config --get remote.". $remotename .".mwLogin");
2da7830d 80# Note: mwPassword is discourraged. Use the credential system instead.
428c995c 81my $wiki_passwd = run_git("config --get remote.". $remotename .".mwPassword");
1d6abac9 82my $wiki_domain = run_git("config --get remote.". $remotename .".mwDomain");
428c995c
JN
83chomp($wiki_login);
84chomp($wiki_passwd);
1d6abac9 85chomp($wiki_domain);
428c995c
JN
86
87# Import only last revisions (both for clone and fetch)
88my $shallow_import = run_git("config --get --bool remote.". $remotename .".shallow");
89chomp($shallow_import);
90$shallow_import = ($shallow_import eq "true");
91
5a29217d
MM
92# Fetch (clone and pull) by revisions instead of by pages. This behavior
93# is more efficient when we have a wiki with lots of pages and we fetch
94# the revisions quite often so that they concern only few pages.
95# Possible values:
96# - by_rev: perform one query per new revision on the remote wiki
97# - by_page: query each tracked page for new revision
98my $fetch_strategy = run_git("config --get remote.$remotename.fetchStrategy");
99unless ($fetch_strategy) {
100 $fetch_strategy = run_git("config --get mediawiki.fetchStrategy");
101}
102chomp($fetch_strategy);
103unless ($fetch_strategy) {
104 $fetch_strategy = "by_page";
105}
106
93f0d338
MM
107# Dumb push: don't update notes and mediawiki ref to reflect the last push.
108#
109# Configurable with mediawiki.dumbPush, or per-remote with
110# remote.<remotename>.dumbPush.
111#
112# This means the user will have to re-import the just-pushed
113# revisions. On the other hand, this means that the Git revisions
114# corresponding to MediaWiki revisions are all imported from the wiki,
115# regardless of whether they were initially created in Git or from the
116# web interface, hence all users will get the same history (i.e. if
117# the push from Git to MediaWiki loses some information, everybody
118# will get the history with information lost). If the import is
119# deterministic, this means everybody gets the same sha1 for each
120# MediaWiki revision.
121my $dumb_push = run_git("config --get --bool remote.$remotename.dumbPush");
122unless ($dumb_push) {
123 $dumb_push = run_git("config --get --bool mediawiki.dumbPush");
124}
125chomp($dumb_push);
126$dumb_push = ($dumb_push eq "true");
127
428c995c
JN
128my $wiki_name = $url;
129$wiki_name =~ s/[^\/]*:\/\///;
9fb79503
MM
130# If URL is like http://user:password@example.com/, we clearly don't
131# want the password in $wiki_name. While we're there, also remove user
132# and '@' sign, to avoid author like MWUser@HTTPUser@host.com
133$wiki_name =~ s/^.*@//;
428c995c
JN
134
135# Commands parser
136my $entry;
137my @cmd;
138while (<STDIN>) {
139 chomp;
140 @cmd = split(/ /);
141 if (defined($cmd[0])) {
142 # Line not blank
143 if ($cmd[0] eq "capabilities") {
144 die("Too many arguments for capabilities") unless (!defined($cmd[1]));
145 mw_capabilities();
146 } elsif ($cmd[0] eq "list") {
147 die("Too many arguments for list") unless (!defined($cmd[2]));
148 mw_list($cmd[1]);
149 } elsif ($cmd[0] eq "import") {
150 die("Invalid arguments for import") unless ($cmd[1] ne "" && !defined($cmd[2]));
151 mw_import($cmd[1]);
152 } elsif ($cmd[0] eq "option") {
153 die("Too many arguments for option") unless ($cmd[1] ne "" && $cmd[2] ne "" && !defined($cmd[3]));
154 mw_option($cmd[1],$cmd[2]);
155 } elsif ($cmd[0] eq "push") {
156 mw_push($cmd[1]);
157 } else {
158 print STDERR "Unknown command. Aborting...\n";
159 last;
160 }
161 } else {
162 # blank line: we should terminate
163 last;
164 }
165
166 BEGIN { $| = 1 } # flush STDOUT, to make sure the previous
167 # command is fully processed.
168}
169
170########################## Functions ##############################
171
a7271ad1
MM
172## credential API management (generic functions)
173
174sub credential_from_url {
175 my $url = shift;
176 my $parsed = URI->new($url);
177 my %credential;
178
179 if ($parsed->scheme) {
180 $credential{protocol} = $parsed->scheme;
181 }
182 if ($parsed->host) {
183 $credential{host} = $parsed->host;
184 }
185 if ($parsed->path) {
186 $credential{path} = $parsed->path;
187 }
188 if ($parsed->userinfo) {
189 if ($parsed->userinfo =~ /([^:]*):(.*)/) {
190 $credential{username} = $1;
191 $credential{password} = $2;
192 } else {
193 $credential{username} = $parsed->userinfo;
194 }
195 }
196
197 return %credential;
198}
199
200sub credential_read {
201 my %credential;
202 my $reader = shift;
203 my $op = shift;
204 while (<$reader>) {
205 my ($key, $value) = /([^=]*)=(.*)/;
206 if (not defined $key) {
207 die "ERROR receiving response from git credential $op:\n$_\n";
208 }
209 $credential{$key} = $value;
210 }
211 return %credential;
212}
213
214sub credential_write {
215 my $credential = shift;
216 my $writer = shift;
217 while (my ($key, $value) = each(%$credential) ) {
218 if ($value) {
219 print $writer "$key=$value\n";
220 }
221 }
222}
223
224sub credential_run {
225 my $op = shift;
226 my $credential = shift;
227 my $pid = open2(my $reader, my $writer, "git credential $op");
228 credential_write($credential, $writer);
229 print $writer "\n";
230 close($writer);
231
232 if ($op eq "fill") {
233 %$credential = credential_read($reader, $op);
234 } else {
235 if (<$reader>) {
236 die "ERROR while running git credential $op:\n$_";
237 }
238 }
239 close($reader);
240 waitpid($pid, 0);
241 my $child_exit_status = $? >> 8;
242 if ($child_exit_status != 0) {
243 die "'git credential $op' failed with code $child_exit_status.";
244 }
245}
246
428c995c
JN
247# MediaWiki API instance, created lazily.
248my $mediawiki;
249
250sub mw_connect_maybe {
251 if ($mediawiki) {
a7271ad1 252 return;
428c995c
JN
253 }
254 $mediawiki = MediaWiki::API->new;
255 $mediawiki->{config}->{api_url} = "$url/api.php";
256 if ($wiki_login) {
a7271ad1
MM
257 my %credential = credential_from_url($url);
258 $credential{username} = $wiki_login;
259 $credential{password} = $wiki_passwd;
260 credential_run("fill", \%credential);
261 my $request = {lgname => $credential{username},
262 lgpassword => $credential{password},
263 lgdomain => $wiki_domain};
264 if ($mediawiki->login($request)) {
265 credential_run("approve", \%credential);
266 print STDERR "Logged in mediawiki user \"$credential{username}\".\n";
428c995c 267 } else {
a7271ad1
MM
268 print STDERR "Failed to log in mediawiki user \"$credential{username}\" on $url\n";
269 print STDERR " (error " .
270 $mediawiki->{error}->{code} . ': ' .
271 $mediawiki->{error}->{details} . ")\n";
272 credential_run("reject", \%credential);
273 exit 1;
428c995c
JN
274 }
275 }
276}
277
9cb74f36
MM
278## Functions for listing pages on the remote wiki
279sub get_mw_tracked_pages {
280 my $pages = shift;
6df7e0df
MM
281 get_mw_page_list(\@tracked_pages, $pages);
282}
283
284sub get_mw_page_list {
285 my $page_list = shift;
286 my $pages = shift;
287 my @some_pages = @$page_list;
9cb74f36
MM
288 while (@some_pages) {
289 my $last = 50;
290 if ($#some_pages < $last) {
291 $last = $#some_pages;
292 }
293 my @slice = @some_pages[0..$last];
294 get_mw_first_pages(\@slice, $pages);
295 @some_pages = @some_pages[51..$#some_pages];
296 }
297}
298
299sub get_mw_tracked_categories {
300 my $pages = shift;
301 foreach my $category (@tracked_categories) {
302 if (index($category, ':') < 0) {
303 # Mediawiki requires the Category
304 # prefix, but let's not force the user
305 # to specify it.
306 $category = "Category:" . $category;
307 }
308 my $mw_pages = $mediawiki->list( {
309 action => 'query',
310 list => 'categorymembers',
311 cmtitle => $category,
312 cmlimit => 'max' } )
313 || die $mediawiki->{error}->{code} . ': '
314 . $mediawiki->{error}->{details};
315 foreach my $page (@{$mw_pages}) {
316 $pages->{$page->{title}} = $page;
317 }
318 }
319}
320
321sub get_mw_all_pages {
322 my $pages = shift;
323 # No user-provided list, get the list of pages from the API.
324 my $mw_pages = $mediawiki->list({
325 action => 'query',
326 list => 'allpages',
327 aplimit => 'max'
328 });
329 if (!defined($mw_pages)) {
330 print STDERR "fatal: could not get the list of wiki pages.\n";
331 print STDERR "fatal: '$url' does not appear to be a mediawiki\n";
332 print STDERR "fatal: make sure '$url/api.php' is a valid page.\n";
333 exit 1;
334 }
335 foreach my $page (@{$mw_pages}) {
336 $pages->{$page->{title}} = $page;
337 }
338}
339
340# queries the wiki for a set of pages. Meant to be used within a loop
341# querying the wiki for slices of page list.
428c995c
JN
342sub get_mw_first_pages {
343 my $some_pages = shift;
344 my @some_pages = @{$some_pages};
345
346 my $pages = shift;
347
348 # pattern 'page1|page2|...' required by the API
349 my $titles = join('|', @some_pages);
350
351 my $mw_pages = $mediawiki->api({
352 action => 'query',
353 titles => $titles,
354 });
355 if (!defined($mw_pages)) {
356 print STDERR "fatal: could not query the list of wiki pages.\n";
357 print STDERR "fatal: '$url' does not appear to be a mediawiki\n";
358 print STDERR "fatal: make sure '$url/api.php' is a valid page.\n";
359 exit 1;
360 }
361 while (my ($id, $page) = each(%{$mw_pages->{query}->{pages}})) {
362 if ($id < 0) {
363 print STDERR "Warning: page $page->{title} not found on wiki\n";
364 } else {
365 $pages->{$page->{title}} = $page;
366 }
367 }
368}
369
9cb74f36 370# Get the list of pages to be fetched according to configuration.
428c995c
JN
371sub get_mw_pages {
372 mw_connect_maybe();
373
f690ddfd
MM
374 print STDERR "Listing pages on remote wiki...\n";
375
428c995c
JN
376 my %pages; # hash on page titles to avoid duplicates
377 my $user_defined;
378 if (@tracked_pages) {
379 $user_defined = 1;
380 # The user provided a list of pages titles, but we
381 # still need to query the API to get the page IDs.
9cb74f36 382 get_mw_tracked_pages(\%pages);
428c995c
JN
383 }
384 if (@tracked_categories) {
385 $user_defined = 1;
9cb74f36 386 get_mw_tracked_categories(\%pages);
428c995c
JN
387 }
388 if (!$user_defined) {
9cb74f36 389 get_mw_all_pages(\%pages);
428c995c 390 }
6a9e55b0
PV
391 if ($import_media) {
392 print STDERR "Getting media files for selected pages...\n";
393 if ($user_defined) {
394 get_linked_mediafiles(\%pages);
395 } else {
396 get_all_mediafiles(\%pages);
428c995c
JN
397 }
398 }
f690ddfd 399 print STDERR (scalar keys %pages) . " pages found.\n";
b1ede9a9 400 return %pages;
428c995c
JN
401}
402
b3d98595
NKT
403# usage: $out = run_git("command args");
404# $out = run_git("command args", "raw"); # don't interpret output as UTF-8.
428c995c 405sub run_git {
b3d98595
NKT
406 my $args = shift;
407 my $encoding = (shift || "encoding(UTF-8)");
408 open(my $git, "-|:$encoding", "git " . $args);
428c995c
JN
409 my $res = do { local $/; <$git> };
410 close($git);
411
412 return $res;
413}
414
415
6a9e55b0
PV
416sub get_all_mediafiles {
417 my $pages = shift;
418 # Attach list of all pages for media files from the API,
419 # they are in a different namespace, only one namespace
420 # can be queried at the same moment
421 my $mw_pages = $mediawiki->list({
422 action => 'query',
423 list => 'allpages',
424 apnamespace => get_mw_namespace_id("File"),
425 aplimit => 'max'
426 });
427 if (!defined($mw_pages)) {
428 print STDERR "fatal: could not get the list of pages for media files.\n";
429 print STDERR "fatal: '$url' does not appear to be a mediawiki\n";
430 print STDERR "fatal: make sure '$url/api.php' is a valid page.\n";
431 exit 1;
432 }
433 foreach my $page (@{$mw_pages}) {
434 $pages->{$page->{title}} = $page;
435 }
436}
437
438sub get_linked_mediafiles {
439 my $pages = shift;
440 my @titles = map $_->{title}, values(%{$pages});
441
442 # The query is split in small batches because of the MW API limit of
443 # the number of links to be returned (500 links max).
444 my $batch = 10;
445 while (@titles) {
446 if ($#titles < $batch) {
447 $batch = $#titles;
448 }
449 my @slice = @titles[0..$batch];
450
451 # pattern 'page1|page2|...' required by the API
452 my $mw_titles = join('|', @slice);
453
454 # Media files could be included or linked from
455 # a page, get all related
456 my $query = {
457 action => 'query',
458 prop => 'links|images',
459 titles => $mw_titles,
460 plnamespace => get_mw_namespace_id("File"),
461 pllimit => 'max'
462 };
463 my $result = $mediawiki->api($query);
464
465 while (my ($id, $page) = each(%{$result->{query}->{pages}})) {
6df7e0df 466 my @media_titles;
6a9e55b0
PV
467 if (defined($page->{links})) {
468 my @link_titles = map $_->{title}, @{$page->{links}};
6df7e0df 469 push(@media_titles, @link_titles);
6a9e55b0
PV
470 }
471 if (defined($page->{images})) {
472 my @image_titles = map $_->{title}, @{$page->{images}};
6df7e0df 473 push(@media_titles, @image_titles);
6a9e55b0 474 }
6df7e0df
MM
475 if (@media_titles) {
476 get_mw_page_list(\@media_titles, $pages);
6a9e55b0
PV
477 }
478 }
479
480 @titles = @titles[($batch+1)..$#titles];
481 }
482}
483
484sub get_mw_mediafile_for_page_revision {
485 # Name of the file on Wiki, with the prefix.
6df7e0df 486 my $filename = shift;
6a9e55b0
PV
487 my $timestamp = shift;
488 my %mediafile;
489
6df7e0df
MM
490 # Search if on a media file with given timestamp exists on
491 # MediaWiki. In that case download the file.
6a9e55b0
PV
492 my $query = {
493 action => 'query',
494 prop => 'imageinfo',
6df7e0df 495 titles => "File:" . $filename,
6a9e55b0
PV
496 iistart => $timestamp,
497 iiend => $timestamp,
498 iiprop => 'timestamp|archivename|url',
499 iilimit => 1
500 };
501 my $result = $mediawiki->api($query);
502
6df7e0df 503 my ($fileid, $file) = each( %{$result->{query}->{pages}} );
6a9e55b0
PV
504 # If not defined it means there is no revision of the file for
505 # given timestamp.
506 if (defined($file->{imageinfo})) {
6a9e55b0
PV
507 $mediafile{title} = $filename;
508
509 my $fileinfo = pop(@{$file->{imageinfo}});
510 $mediafile{timestamp} = $fileinfo->{timestamp};
6df7e0df
MM
511 # Mediawiki::API's download function doesn't support https URLs
512 # and can't download old versions of files.
513 print STDERR "\tDownloading file $mediafile{title}, version $mediafile{timestamp}\n";
514 $mediafile{content} = download_mw_mediafile($fileinfo->{url});
6a9e55b0
PV
515 }
516 return %mediafile;
517}
518
6df7e0df 519sub download_mw_mediafile {
6a9e55b0 520 my $url = shift;
6a9e55b0 521
6df7e0df
MM
522 my $response = $mediawiki->{ua}->get($url);
523 if ($response->code == 200) {
524 return $response->decoded_content;
6a9e55b0 525 } else {
6df7e0df
MM
526 print STDERR "Error downloading mediafile from :\n";
527 print STDERR "URL: $url\n";
528 print STDERR "Server response: " . $response->code . " " . $response->message . "\n";
6a9e55b0 529 exit 1;
6a9e55b0
PV
530 }
531}
532
428c995c
JN
533sub get_last_local_revision {
534 # Get note regarding last mediawiki revision
535 my $note = run_git("notes --ref=$remotename/mediawiki show refs/mediawiki/$remotename/master 2>/dev/null");
536 my @note_info = split(/ /, $note);
537
538 my $lastrevision_number;
539 if (!(defined($note_info[0]) && $note_info[0] eq "mediawiki_revision:")) {
540 print STDERR "No previous mediawiki revision found";
541 $lastrevision_number = 0;
542 } else {
543 # Notes are formatted : mediawiki_revision: #number
544 $lastrevision_number = $note_info[1];
545 chomp($lastrevision_number);
546 print STDERR "Last local mediawiki revision found is $lastrevision_number";
547 }
548 return $lastrevision_number;
549}
550
3c1ed90e
MM
551# Remember the timestamp corresponding to a revision id.
552my %basetimestamps;
553
5a29217d
MM
554# Get the last remote revision without taking in account which pages are
555# tracked or not. This function makes a single request to the wiki thus
556# avoid a loop onto all tracked pages. This is useful for the fetch-by-rev
557# option.
558sub get_last_global_remote_rev {
559 mw_connect_maybe();
560
561 my $query = {
562 action => 'query',
563 list => 'recentchanges',
564 prop => 'revisions',
565 rclimit => '1',
566 rcdir => 'older',
567 };
568 my $result = $mediawiki->api($query);
569 return $result->{query}->{recentchanges}[0]->{revid};
570}
571
572# Get the last remote revision concerning the tracked pages and the tracked
573# categories.
428c995c
JN
574sub get_last_remote_revision {
575 mw_connect_maybe();
576
b1ede9a9
MM
577 my %pages_hash = get_mw_pages();
578 my @pages = values(%pages_hash);
428c995c
JN
579
580 my $max_rev_num = 0;
581
a393f488
MM
582 print STDERR "Getting last revision id on tracked pages...\n";
583
428c995c
JN
584 foreach my $page (@pages) {
585 my $id = $page->{pageid};
586
587 my $query = {
588 action => 'query',
589 prop => 'revisions',
3c1ed90e 590 rvprop => 'ids|timestamp',
428c995c
JN
591 pageids => $id,
592 };
593
594 my $result = $mediawiki->api($query);
595
596 my $lastrev = pop(@{$result->{query}->{pages}->{$id}->{revisions}});
597
3c1ed90e
MM
598 $basetimestamps{$lastrev->{revid}} = $lastrev->{timestamp};
599
428c995c
JN
600 $max_rev_num = ($lastrev->{revid} > $max_rev_num ? $lastrev->{revid} : $max_rev_num);
601 }
602
603 print STDERR "Last remote revision found is $max_rev_num.\n";
604 return $max_rev_num;
605}
606
607# Clean content before sending it to MediaWiki
608sub mediawiki_clean {
609 my $string = shift;
610 my $page_created = shift;
611 # Mediawiki does not allow blank space at the end of a page and ends with a single \n.
612 # This function right trims a string and adds a \n at the end to follow this rule
613 $string =~ s/\s+$//;
614 if ($string eq "" && $page_created) {
615 # Creating empty pages is forbidden.
616 $string = EMPTY_CONTENT;
617 }
618 return $string."\n";
619}
620
621# Filter applied on MediaWiki data before adding them to Git
622sub mediawiki_smudge {
623 my $string = shift;
624 if ($string eq EMPTY_CONTENT) {
625 $string = "";
626 }
627 # This \n is important. This is due to mediawiki's way to handle end of files.
628 return $string."\n";
629}
630
631sub mediawiki_clean_filename {
632 my $filename = shift;
633 $filename =~ s/@{[SLASH_REPLACEMENT]}/\//g;
634 # [, ], |, {, and } are forbidden by MediaWiki, even URL-encoded.
635 # Do a variant of URL-encoding, i.e. looks like URL-encoding,
636 # but with _ added to prevent MediaWiki from thinking this is
637 # an actual special character.
638 $filename =~ s/[\[\]\{\}\|]/sprintf("_%%_%x", ord($&))/ge;
639 # If we use the uri escape before
640 # we should unescape here, before anything
641
642 return $filename;
643}
644
645sub mediawiki_smudge_filename {
646 my $filename = shift;
647 $filename =~ s/\//@{[SLASH_REPLACEMENT]}/g;
648 $filename =~ s/ /_/g;
649 # Decode forbidden characters encoded in mediawiki_clean_filename
650 $filename =~ s/_%_([0-9a-fA-F][0-9a-fA-F])/sprintf("%c", hex($1))/ge;
651 return $filename;
652}
653
654sub literal_data {
655 my ($content) = @_;
656 print STDOUT "data ", bytes::length($content), "\n", $content;
657}
658
6a9e55b0
PV
659sub literal_data_raw {
660 # Output possibly binary content.
661 my ($content) = @_;
662 # Avoid confusion between size in bytes and in characters
663 utf8::downgrade($content);
664 binmode STDOUT, ":raw";
665 print STDOUT "data ", bytes::length($content), "\n", $content;
666 binmode STDOUT, ":utf8";
667}
668
428c995c
JN
669sub mw_capabilities {
670 # Revisions are imported to the private namespace
671 # refs/mediawiki/$remotename/ by the helper and fetched into
672 # refs/remotes/$remotename later by fetch.
673 print STDOUT "refspec refs/heads/*:refs/mediawiki/$remotename/*\n";
674 print STDOUT "import\n";
675 print STDOUT "list\n";
676 print STDOUT "push\n";
677 print STDOUT "\n";
678}
679
680sub mw_list {
681 # MediaWiki do not have branches, we consider one branch arbitrarily
682 # called master, and HEAD pointing to it.
683 print STDOUT "? refs/heads/master\n";
684 print STDOUT "\@refs/heads/master HEAD\n";
685 print STDOUT "\n";
686}
687
688sub mw_option {
689 print STDERR "remote-helper command 'option $_[0]' not yet implemented\n";
690 print STDOUT "unsupported\n";
691}
692
693sub fetch_mw_revisions_for_page {
694 my $page = shift;
695 my $id = shift;
696 my $fetch_from = shift;
697 my @page_revs = ();
698 my $query = {
699 action => 'query',
700 prop => 'revisions',
701 rvprop => 'ids',
702 rvdir => 'newer',
703 rvstartid => $fetch_from,
704 rvlimit => 500,
705 pageids => $id,
706 };
707
708 my $revnum = 0;
709 # Get 500 revisions at a time due to the mediawiki api limit
710 while (1) {
711 my $result = $mediawiki->api($query);
712
713 # Parse each of those 500 revisions
714 foreach my $revision (@{$result->{query}->{pages}->{$id}->{revisions}}) {
715 my $page_rev_ids;
716 $page_rev_ids->{pageid} = $page->{pageid};
717 $page_rev_ids->{revid} = $revision->{revid};
718 push(@page_revs, $page_rev_ids);
719 $revnum++;
720 }
721 last unless $result->{'query-continue'};
722 $query->{rvstartid} = $result->{'query-continue'}->{revisions}->{rvstartid};
723 }
724 if ($shallow_import && @page_revs) {
725 print STDERR " Found 1 revision (shallow import).\n";
726 @page_revs = sort {$b->{revid} <=> $a->{revid}} (@page_revs);
727 return $page_revs[0];
728 }
729 print STDERR " Found ", $revnum, " revision(s).\n";
730 return @page_revs;
731}
732
733sub fetch_mw_revisions {
734 my $pages = shift; my @pages = @{$pages};
735 my $fetch_from = shift;
736
737 my @revisions = ();
738 my $n = 1;
739 foreach my $page (@pages) {
740 my $id = $page->{pageid};
741
742 print STDERR "page $n/", scalar(@pages), ": ". $page->{title} ."\n";
743 $n++;
744 my @page_revs = fetch_mw_revisions_for_page($page, $id, $fetch_from);
745 @revisions = (@page_revs, @revisions);
746 }
747
748 return ($n, @revisions);
749}
750
751sub import_file_revision {
752 my $commit = shift;
753 my %commit = %{$commit};
754 my $full_import = shift;
755 my $n = shift;
6a9e55b0
PV
756 my $mediafile = shift;
757 my %mediafile;
758 if ($mediafile) {
759 %mediafile = %{$mediafile};
760 }
428c995c
JN
761
762 my $title = $commit{title};
763 my $comment = $commit{comment};
764 my $content = $commit{content};
765 my $author = $commit{author};
766 my $date = $commit{date};
767
768 print STDOUT "commit refs/mediawiki/$remotename/master\n";
769 print STDOUT "mark :$n\n";
770 print STDOUT "committer $author <$author\@$wiki_name> ", $date->epoch, " +0000\n";
771 literal_data($comment);
772
773 # If it's not a clone, we need to know where to start from
774 if (!$full_import && $n == 1) {
775 print STDOUT "from refs/mediawiki/$remotename/master^0\n";
776 }
777 if ($content ne DELETED_CONTENT) {
778 print STDOUT "M 644 inline $title.mw\n";
779 literal_data($content);
6a9e55b0
PV
780 if (%mediafile) {
781 print STDOUT "M 644 inline $mediafile{title}\n";
782 literal_data_raw($mediafile{content});
783 }
428c995c
JN
784 print STDOUT "\n\n";
785 } else {
786 print STDOUT "D $title.mw\n";
787 }
788
789 # mediawiki revision number in the git note
790 if ($full_import && $n == 1) {
791 print STDOUT "reset refs/notes/$remotename/mediawiki\n";
792 }
793 print STDOUT "commit refs/notes/$remotename/mediawiki\n";
794 print STDOUT "committer $author <$author\@$wiki_name> ", $date->epoch, " +0000\n";
795 literal_data("Note added by git-mediawiki during import");
796 if (!$full_import && $n == 1) {
797 print STDOUT "from refs/notes/$remotename/mediawiki^0\n";
798 }
799 print STDOUT "N inline :$n\n";
800 literal_data("mediawiki_revision: " . $commit{mw_revision});
801 print STDOUT "\n\n";
802}
803
804# parse a sequence of
805# <cmd> <arg1>
806# <cmd> <arg2>
807# \n
808# (like batch sequence of import and sequence of push statements)
809sub get_more_refs {
810 my $cmd = shift;
811 my @refs;
812 while (1) {
813 my $line = <STDIN>;
814 if ($line =~ m/^$cmd (.*)$/) {
815 push(@refs, $1);
816 } elsif ($line eq "\n") {
817 return @refs;
818 } else {
819 die("Invalid command in a '$cmd' batch: ". $_);
820 }
821 }
822}
823
824sub mw_import {
825 # multiple import commands can follow each other.
826 my @refs = (shift, get_more_refs("import"));
827 foreach my $ref (@refs) {
828 mw_import_ref($ref);
829 }
830 print STDOUT "done\n";
831}
832
833sub mw_import_ref {
834 my $ref = shift;
835 # The remote helper will call "import HEAD" and
836 # "import refs/heads/master".
837 # Since HEAD is a symbolic ref to master (by convention,
838 # followed by the output of the command "list" that we gave),
839 # we don't need to do anything in this case.
840 if ($ref eq "HEAD") {
841 return;
842 }
843
844 mw_connect_maybe();
845
428c995c
JN
846 print STDERR "Searching revisions...\n";
847 my $last_local = get_last_local_revision();
848 my $fetch_from = $last_local + 1;
849 if ($fetch_from == 1) {
850 print STDERR ", fetching from beginning.\n";
851 } else {
852 print STDERR ", fetching from here.\n";
853 }
428c995c 854
5a29217d
MM
855 my $n = 0;
856 if ($fetch_strategy eq "by_rev") {
857 print STDERR "Fetching & writing export data by revs...\n";
858 $n = mw_import_ref_by_revs($fetch_from);
859 } elsif ($fetch_strategy eq "by_page") {
860 print STDERR "Fetching & writing export data by pages...\n";
861 $n = mw_import_ref_by_pages($fetch_from);
862 } else {
863 print STDERR "fatal: invalid fetch strategy \"$fetch_strategy\".\n";
864 print STDERR "Check your configuration variables remote.$remotename.fetchStrategy and mediawiki.fetchStrategy\n";
865 exit 1;
866 }
867
868 if ($fetch_from == 1 && $n == 0) {
869 print STDERR "You appear to have cloned an empty MediaWiki.\n";
870 # Something has to be done remote-helper side. If nothing is done, an error is
871 # thrown saying that HEAD is refering to unknown object 0000000000000000000
872 # and the clone fails.
873 }
874}
875
876sub mw_import_ref_by_pages {
877
878 my $fetch_from = shift;
879 my %pages_hash = get_mw_pages();
880 my @pages = values(%pages_hash);
881
428c995c
JN
882 my ($n, @revisions) = fetch_mw_revisions(\@pages, $fetch_from);
883
4465b6d6
MM
884 @revisions = sort {$a->{revid} <=> $b->{revid}} @revisions;
885 my @revision_ids = map $_->{revid}, @revisions;
886
312fa9aa
MM
887 return mw_import_revids($fetch_from, \@revision_ids, \%pages_hash);
888}
889
5a29217d
MM
890sub mw_import_ref_by_revs {
891
892 my $fetch_from = shift;
893 my %pages_hash = get_mw_pages();
428c995c 894
5a29217d
MM
895 my $last_remote = get_last_global_remote_rev();
896 my @revision_ids = $fetch_from..$last_remote;
897 return mw_import_revids($fetch_from, \@revision_ids, \%pages_hash);
898}
899
900# Import revisions given in second argument (array of integers).
901# Only pages appearing in the third argument (hash indexed by page titles)
902# will be imported.
312fa9aa
MM
903sub mw_import_revids {
904 my $fetch_from = shift;
905 my $revision_ids = shift;
906 my $pages = shift;
907
908 my $n = 0;
5a29217d 909 my $n_actual = 0;
428c995c
JN
910 my $last_timestamp = 0; # Placeholer in case $rev->timestamp is undefined
911
312fa9aa 912 foreach my $pagerevid (@$revision_ids) {
ebd5fe1c
MM
913 # Count page even if we skip it, since we display
914 # $n/$total and $total includes skipped pages.
915 $n++;
916
428c995c
JN
917 # fetch the content of the pages
918 my $query = {
919 action => 'query',
920 prop => 'revisions',
921 rvprop => 'content|timestamp|comment|user|ids',
4465b6d6 922 revids => $pagerevid,
428c995c
JN
923 };
924
925 my $result = $mediawiki->api($query);
926
ff0bfd75
MM
927 if (!$result) {
928 die "Failed to retrieve modified page for revision $pagerevid";
929 }
930
ebd5fe1c
MM
931 if (defined($result->{query}->{badrevids}->{$pagerevid})) {
932 # The revision id does not exist on the remote wiki.
933 next;
934 }
935
ff0bfd75
MM
936 if (!defined($result->{query}->{pages})) {
937 die "Invalid revision $pagerevid.";
938 }
939
4465b6d6
MM
940 my @result_pages = values(%{$result->{query}->{pages}});
941 my $result_page = $result_pages[0];
942 my $rev = $result_pages[0]->{revisions}->[0];
428c995c 943
4465b6d6 944 my $page_title = $result_page->{title};
5a29217d
MM
945
946 if (!exists($pages->{$page_title})) {
947 print STDERR "$n/", scalar(@$revision_ids),
948 ": Skipping revision #$rev->{revid} of $page_title\n";
949 next;
950 }
951
952 $n_actual++;
953
428c995c
JN
954 my %commit;
955 $commit{author} = $rev->{user} || 'Anonymous';
dcb1ea62 956 $commit{comment} = $rev->{comment} || EMPTY_MESSAGE;
6a9e55b0 957 $commit{title} = mediawiki_smudge_filename($page_title);
4465b6d6 958 $commit{mw_revision} = $rev->{revid};
428c995c
JN
959 $commit{content} = mediawiki_smudge($rev->{'*'});
960
961 if (!defined($rev->{timestamp})) {
962 $last_timestamp++;
963 } else {
964 $last_timestamp = $rev->{timestamp};
965 }
966 $commit{date} = DateTime::Format::ISO8601->parse_datetime($last_timestamp);
967
6a9e55b0 968 # Differentiates classic pages and media files.
6df7e0df 969 my ($namespace, $filename) = $page_title =~ /^([^:]*):(.*)$/;
6a9e55b0 970 my %mediafile;
6df7e0df 971 if ($namespace && get_mw_namespace_id($namespace) == get_mw_namespace_id("File")) {
6a9e55b0
PV
972 %mediafile = get_mw_mediafile_for_page_revision($filename, $rev->{timestamp});
973 }
974 # If this is a revision of the media page for new version
975 # of a file do one common commit for both file and media page.
976 # Else do commit only for that page.
312fa9aa 977 print STDERR "$n/", scalar(@$revision_ids), ": Revision #$rev->{revid} of $commit{title}\n";
5a29217d 978 import_file_revision(\%commit, ($fetch_from == 1), $n_actual, \%mediafile);
428c995c
JN
979 }
980
5a29217d 981 return $n_actual;
428c995c
JN
982}
983
984sub error_non_fast_forward {
fd47d7b9
MM
985 my $advice = run_git("config --bool advice.pushNonFastForward");
986 chomp($advice);
987 if ($advice ne "false") {
988 # Native git-push would show this after the summary.
989 # We can't ask it to display it cleanly, so print it
990 # ourselves before.
991 print STDERR "To prevent you from losing history, non-fast-forward updates were rejected\n";
992 print STDERR "Merge the remote changes (e.g. 'git pull') before pushing again. See the\n";
993 print STDERR "'Note about fast-forwards' section of 'git push --help' for details.\n";
994 }
428c995c
JN
995 print STDOUT "error $_[0] \"non-fast-forward\"\n";
996 return 0;
997}
998
b3d98595
NKT
999sub mw_upload_file {
1000 my $complete_file_name = shift;
1001 my $new_sha1 = shift;
1002 my $extension = shift;
1003 my $file_deleted = shift;
1004 my $summary = shift;
1005 my $newrevid;
1006 my $path = "File:" . $complete_file_name;
1007 my %hashFiles = get_allowed_file_extensions();
1008 if (!exists($hashFiles{$extension})) {
1009 print STDERR "$complete_file_name is not a permitted file on this wiki.\n";
1010 print STDERR "Check the configuration of file uploads in your mediawiki.\n";
1011 return $newrevid;
1012 }
1013 # Deleting and uploading a file requires a priviledged user
1014 if ($file_deleted) {
1015 mw_connect_maybe();
1016 my $query = {
1017 action => 'delete',
1018 title => $path,
1019 reason => $summary
1020 };
1021 if (!$mediawiki->edit($query)) {
1022 print STDERR "Failed to delete file on remote wiki\n";
1023 print STDERR "Check your permissions on the remote site. Error code:\n";
1024 print STDERR $mediawiki->{error}->{code} . ':' . $mediawiki->{error}->{details};
1025 exit 1;
1026 }
1027 } else {
1028 # Don't let perl try to interpret file content as UTF-8 => use "raw"
1029 my $content = run_git("cat-file blob $new_sha1", "raw");
1030 if ($content ne "") {
1031 mw_connect_maybe();
1032 $mediawiki->{config}->{upload_url} =
1033 "$url/index.php/Special:Upload";
1034 $mediawiki->edit({
1035 action => 'upload',
1036 filename => $complete_file_name,
1037 comment => $summary,
1038 file => [undef,
1039 $complete_file_name,
1040 Content => $content],
1041 ignorewarnings => 1,
1042 }, {
1043 skip_encoding => 1
1044 } ) || die $mediawiki->{error}->{code} . ':'
1045 . $mediawiki->{error}->{details};
1046 my $last_file_page = $mediawiki->get_page({title => $path});
1047 $newrevid = $last_file_page->{revid};
1048 print STDERR "Pushed file: $new_sha1 - $complete_file_name.\n";
1049 } else {
1050 print STDERR "Empty file $complete_file_name not pushed.\n";
1051 }
1052 }
1053 return $newrevid;
1054}
1055
428c995c
JN
1056sub mw_push_file {
1057 my $diff_info = shift;
1058 # $diff_info contains a string in this format:
1059 # 100644 100644 <sha1_of_blob_before_commit> <sha1_of_blob_now> <status>
1060 my @diff_info_split = split(/[ \t]/, $diff_info);
1061
1062 # Filename, including .mw extension
1063 my $complete_file_name = shift;
1064 # Commit message
1065 my $summary = shift;
93f0d338
MM
1066 # MediaWiki revision number. Keep the previous one by default,
1067 # in case there's no edit to perform.
b3d98595
NKT
1068 my $oldrevid = shift;
1069 my $newrevid;
428c995c 1070
dcb1ea62
MM
1071 if ($summary eq EMPTY_MESSAGE) {
1072 $summary = '';
1073 }
1074
428c995c
JN
1075 my $new_sha1 = $diff_info_split[3];
1076 my $old_sha1 = $diff_info_split[2];
1077 my $page_created = ($old_sha1 eq NULL_SHA1);
1078 my $page_deleted = ($new_sha1 eq NULL_SHA1);
1079 $complete_file_name = mediawiki_clean_filename($complete_file_name);
1080
b3d98595
NKT
1081 my ($title, $extension) = $complete_file_name =~ /^(.*)\.([^\.]*)$/;
1082 if (!defined($extension)) {
1083 $extension = "";
1084 }
1085 if ($extension eq "mw") {
2045e293
MM
1086 my $ns = get_mw_namespace_id_for_page($complete_file_name);
1087 if ($ns && $ns == get_mw_namespace_id("File") && (!$export_media)) {
1088 print STDERR "Ignoring media file related page: $complete_file_name\n";
1089 return ($oldrevid, "ok");
1090 }
428c995c
JN
1091 my $file_content;
1092 if ($page_deleted) {
1093 # Deleting a page usually requires
1094 # special priviledges. A common
1095 # convention is to replace the page
1096 # with this content instead:
1097 $file_content = DELETED_CONTENT;
1098 } else {
1099 $file_content = run_git("cat-file blob $new_sha1");
1100 }
1101
1102 mw_connect_maybe();
1103
1104 my $result = $mediawiki->edit( {
1105 action => 'edit',
1106 summary => $summary,
1107 title => $title,
b3d98595 1108 basetimestamp => $basetimestamps{$oldrevid},
428c995c
JN
1109 text => mediawiki_clean($file_content, $page_created),
1110 }, {
1111 skip_encoding => 1 # Helps with names with accentuated characters
3c1ed90e
MM
1112 });
1113 if (!$result) {
1114 if ($mediawiki->{error}->{code} == 3) {
1115 # edit conflicts, considered as non-fast-forward
1116 print STDERR 'Warning: Error ' .
1117 $mediawiki->{error}->{code} .
1118 ' from mediwiki: ' . $mediawiki->{error}->{details} .
1119 ".\n";
b3d98595 1120 return ($oldrevid, "non-fast-forward");
3c1ed90e
MM
1121 } else {
1122 # Other errors. Shouldn't happen => just die()
1123 die 'Fatal: Error ' .
1124 $mediawiki->{error}->{code} .
1125 ' from mediwiki: ' . $mediawiki->{error}->{details};
1126 }
1127 }
93f0d338 1128 $newrevid = $result->{edit}->{newrevid};
ac86ec0f 1129 print STDERR "Pushed file: $new_sha1 - $title\n";
2045e293 1130 } elsif ($export_media) {
b3d98595
NKT
1131 $newrevid = mw_upload_file($complete_file_name, $new_sha1,
1132 $extension, $page_deleted,
1133 $summary);
2045e293
MM
1134 } else {
1135 print STDERR "Ignoring media file $title\n";
428c995c 1136 }
b3d98595 1137 $newrevid = ($newrevid or $oldrevid);
3c1ed90e 1138 return ($newrevid, "ok");
428c995c
JN
1139}
1140
1141sub mw_push {
1142 # multiple push statements can follow each other
1143 my @refsspecs = (shift, get_more_refs("push"));
428c995c
JN
1144 my $pushed;
1145 for my $refspec (@refsspecs) {
1146 my ($force, $local, $remote) = $refspec =~ /^(\+)?([^:]*):([^:]*)$/
1147 or die("Invalid refspec for push. Expected <src>:<dst> or +<src>:<dst>");
1148 if ($force) {
1149 print STDERR "Warning: forced push not allowed on a MediaWiki.\n";
1150 }
1151 if ($local eq "") {
1152 print STDERR "Cannot delete remote branch on a MediaWiki\n";
1153 print STDOUT "error $remote cannot delete\n";
1154 next;
1155 }
1156 if ($remote ne "refs/heads/master") {
1157 print STDERR "Only push to the branch 'master' is supported on a MediaWiki\n";
1158 print STDOUT "error $remote only master allowed\n";
1159 next;
1160 }
1161 if (mw_push_revision($local, $remote)) {
1162 $pushed = 1;
1163 }
1164 }
1165
1166 # Notify Git that the push is done
1167 print STDOUT "\n";
1168
93f0d338 1169 if ($pushed && $dumb_push) {
428c995c
JN
1170 print STDERR "Just pushed some revisions to MediaWiki.\n";
1171 print STDERR "The pushed revisions now have to be re-imported, and your current branch\n";
1172 print STDERR "needs to be updated with these re-imported commits. You can do this with\n";
1173 print STDERR "\n";
1174 print STDERR " git pull --rebase\n";
1175 print STDERR "\n";
1176 }
1177}
1178
1179sub mw_push_revision {
1180 my $local = shift;
1181 my $remote = shift; # actually, this has to be "refs/heads/master" at this point.
1182 my $last_local_revid = get_last_local_revision();
1183 print STDERR ".\n"; # Finish sentence started by get_last_local_revision()
1184 my $last_remote_revid = get_last_remote_revision();
93f0d338 1185 my $mw_revision = $last_remote_revid;
428c995c
JN
1186
1187 # Get sha1 of commit pointed by local HEAD
1188 my $HEAD_sha1 = run_git("rev-parse $local 2>/dev/null"); chomp($HEAD_sha1);
1189 # Get sha1 of commit pointed by remotes/$remotename/master
1190 my $remoteorigin_sha1 = run_git("rev-parse refs/remotes/$remotename/master 2>/dev/null");
1191 chomp($remoteorigin_sha1);
1192
1193 if ($last_local_revid > 0 &&
1194 $last_local_revid < $last_remote_revid) {
1195 return error_non_fast_forward($remote);
1196 }
1197
1198 if ($HEAD_sha1 eq $remoteorigin_sha1) {
1199 # nothing to push
1200 return 0;
1201 }
1202
1203 # Get every commit in between HEAD and refs/remotes/origin/master,
1204 # including HEAD and refs/remotes/origin/master
1205 my @commit_pairs = ();
1206 if ($last_local_revid > 0) {
1207 my $parsed_sha1 = $remoteorigin_sha1;
1208 # Find a path from last MediaWiki commit to pushed commit
93e92d4d
MM
1209 print STDERR "Computing path from local to remote ...\n";
1210 my @local_ancestry = split(/\n/, run_git("rev-list --boundary --parents $local ^$parsed_sha1"));
1211 my %local_ancestry;
1212 foreach my $line (@local_ancestry) {
1213 if (my ($child, $parents) = $line =~ m/^-?([a-f0-9]+) ([a-f0-9 ]+)/) {
1214 foreach my $parent (split(' ', $parents)) {
1215 $local_ancestry{$parent} = $child;
1216 }
1217 } elsif (!$line =~ m/^([a-f0-9]+)/) {
1218 die "Unexpected output from git rev-list: $line";
1219 }
1220 }
428c995c 1221 while ($parsed_sha1 ne $HEAD_sha1) {
93e92d4d
MM
1222 my $child = $local_ancestry{$parsed_sha1};
1223 if (!$child) {
1224 printf STDERR "Cannot find a path in history from remote commit to last commit\n";
428c995c
JN
1225 return error_non_fast_forward($remote);
1226 }
93e92d4d
MM
1227 push(@commit_pairs, [$parsed_sha1, $child]);
1228 $parsed_sha1 = $child;
428c995c
JN
1229 }
1230 } else {
1231 # No remote mediawiki revision. Export the whole
1232 # history (linearized with --first-parent)
1233 print STDERR "Warning: no common ancestor, pushing complete history\n";
1234 my $history = run_git("rev-list --first-parent --children $local");
1235 my @history = split('\n', $history);
1236 @history = @history[1..$#history];
1237 foreach my $line (reverse @history) {
1238 my @commit_info_split = split(/ |\n/, $line);
1239 push(@commit_pairs, \@commit_info_split);
1240 }
1241 }
1242
1243 foreach my $commit_info_split (@commit_pairs) {
1244 my $sha1_child = @{$commit_info_split}[0];
1245 my $sha1_commit = @{$commit_info_split}[1];
1246 my $diff_infos = run_git("diff-tree -r --raw -z $sha1_child $sha1_commit");
1247 # TODO: we could detect rename, and encode them with a #redirect on the wiki.
1248 # TODO: for now, it's just a delete+add
1249 my @diff_info_list = split(/\0/, $diff_infos);
28c24bd7
MM
1250 # Keep the subject line of the commit message as mediawiki comment for the revision
1251 my $commit_msg = run_git("log --no-walk --format=\"%s\" $sha1_commit");
428c995c
JN
1252 chomp($commit_msg);
1253 # Push every blob
1254 while (@diff_info_list) {
3c1ed90e 1255 my $status;
428c995c
JN
1256 # git diff-tree -z gives an output like
1257 # <metadata>\0<filename1>\0
1258 # <metadata>\0<filename2>\0
1259 # and we've split on \0.
1260 my $info = shift(@diff_info_list);
1261 my $file = shift(@diff_info_list);
3c1ed90e
MM
1262 ($mw_revision, $status) = mw_push_file($info, $file, $commit_msg, $mw_revision);
1263 if ($status eq "non-fast-forward") {
1264 # we may already have sent part of the
1265 # commit to MediaWiki, but it's too
1266 # late to cancel it. Stop the push in
1267 # the middle, but still give an
1268 # accurate error message.
1269 return error_non_fast_forward($remote);
1270 }
1271 if ($status ne "ok") {
1272 die("Unknown error from mw_push_file()");
1273 }
93f0d338
MM
1274 }
1275 unless ($dumb_push) {
5c481745 1276 run_git("notes --ref=$remotename/mediawiki add -f -m \"mediawiki_revision: $mw_revision\" $sha1_commit");
93f0d338 1277 run_git("update-ref -m \"Git-MediaWiki push\" refs/mediawiki/$remotename/master $sha1_commit $sha1_child");
428c995c
JN
1278 }
1279 }
1280
1281 print STDOUT "ok $remote\n";
1282 return 1;
1283}
b3d98595
NKT
1284
1285sub get_allowed_file_extensions {
1286 mw_connect_maybe();
1287
1288 my $query = {
1289 action => 'query',
1290 meta => 'siteinfo',
1291 siprop => 'fileextensions'
1292 };
1293 my $result = $mediawiki->api($query);
1294 my @file_extensions= map $_->{ext},@{$result->{query}->{fileextensions}};
1295 my %hashFile = map {$_ => 1}@file_extensions;
1296
1297 return %hashFile;
1298}
6a9e55b0 1299
6df7e0df
MM
1300# In memory cache for MediaWiki namespace ids.
1301my %namespace_id;
1302
1303# Namespaces whose id is cached in the configuration file
1304# (to avoid duplicates)
1305my %cached_mw_namespace_id;
1306
6a9e55b0
PV
1307# Return MediaWiki id for a canonical namespace name.
1308# Ex.: "File", "Project".
6a9e55b0
PV
1309sub get_mw_namespace_id {
1310 mw_connect_maybe();
1311 my $name = shift;
1312
1313 if (!exists $namespace_id{$name}) {
1314 # Look at configuration file, if the record for that namespace is
6df7e0df 1315 # already cached. Namespaces are stored in form:
6a9e55b0 1316 # "Name_of_namespace:Id_namespace", ex.: "File:6".
ac4bbb41 1317 my @temp = split(/[\n]/, run_git("config --get-all remote."
6df7e0df 1318 . $remotename .".namespaceCache"));
6a9e55b0
PV
1319 chomp(@temp);
1320 foreach my $ns (@temp) {
6df7e0df
MM
1321 my ($n, $id) = split(/:/, $ns);
1322 $namespace_id{$n} = $id;
1323 $cached_mw_namespace_id{$n} = 1;
6a9e55b0
PV
1324 }
1325 }
1326
1327 if (!exists $namespace_id{$name}) {
6df7e0df 1328 print STDERR "Namespace $name not found in cache, querying the wiki ...\n";
6a9e55b0
PV
1329 # NS not found => get namespace id from MW and store it in
1330 # configuration file.
1331 my $query = {
1332 action => 'query',
1333 meta => 'siteinfo',
1334 siprop => 'namespaces'
1335 };
1336 my $result = $mediawiki->api($query);
1337
1338 while (my ($id, $ns) = each(%{$result->{query}->{namespaces}})) {
6df7e0df
MM
1339 if (defined($ns->{id}) && defined($ns->{canonical})) {
1340 $namespace_id{$ns->{canonical}} = $ns->{id};
1341 if ($ns->{'*'}) {
1342 # alias (e.g. french Fichier: as alias for canonical File:)
1343 $namespace_id{$ns->{'*'}} = $ns->{id};
1344 }
1345 }
6a9e55b0
PV
1346 }
1347 }
1348
6df7e0df
MM
1349 my $id = $namespace_id{$name};
1350
1351 if (defined $id) {
1352 # Store explicitely requested namespaces on disk
1353 if (!exists $cached_mw_namespace_id{$name}) {
1354 run_git("config --add remote.". $remotename
1355 .".namespaceCache \"". $name .":". $id ."\"");
1356 $cached_mw_namespace_id{$name} = 1;
1357 }
1358 return $id;
6a9e55b0
PV
1359 } else {
1360 die "No such namespace $name on MediaWiki.";
1361 }
1362}
2045e293
MM
1363
1364sub get_mw_namespace_id_for_page {
1365 if (my ($namespace) = $_[0] =~ /^([^:]*):/) {
1366 return get_mw_namespace_id($namespace);
1367 } else {
1368 return;
1369 }
1370}