Git 1.8.3
[git/git.git] / contrib / mw-to-git / git-remote-mediawiki.perl
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.
9a7365cf 12# Documentation & bugtracker: https://github.com/moy/Git-Mediawiki/
428c995c
JN
13
14use strict;
15use MediaWiki::API;
16use DateTime::Format::ISO8601;
428c995c 17
721a533f 18# By default, use UTF-8 to communicate with Git and the user
428c995c 19binmode STDERR, ":utf8";
721a533f 20binmode STDOUT, ":utf8";
428c995c
JN
21
22use URI::Escape;
a7271ad1
MM
23use IPC::Open2;
24
428c995c
JN
25use warnings;
26
27# Mediawiki filenames can contain forward slashes. This variable decides by which pattern they should be replaced
28use constant SLASH_REPLACEMENT => "%2F";
29
30# It's not always possible to delete pages (may require some
2582ab18 31# privileges). Deleted pages are replaced with this content.
428c995c
JN
32use constant DELETED_CONTENT => "[[Category:Deleted]]\n";
33
34# It's not possible to create empty pages. New empty files in Git are
35# sent with this content instead.
36use constant EMPTY_CONTENT => "<!-- empty page -->\n";
37
38# used to reflect file creation or deletion in diff.
39use constant NULL_SHA1 => "0000000000000000000000000000000000000000";
40
dcb1ea62
MM
41# Used on Git's side to reflect empty edit messages on the wiki
42use constant EMPTY_MESSAGE => '*Empty MediaWiki Message*';
43
428c995c
JN
44my $remotename = $ARGV[0];
45my $url = $ARGV[1];
46
47# Accept both space-separated and multiple keys in config file.
48# Spaces should be written as _ anyway because we'll use chomp.
49my @tracked_pages = split(/[ \n]/, run_git("config --get-all remote.". $remotename .".pages"));
50chomp(@tracked_pages);
51
52# Just like @tracked_pages, but for MediaWiki categories.
53my @tracked_categories = split(/[ \n]/, run_git("config --get-all remote.". $remotename .".categories"));
54chomp(@tracked_categories);
55
2045e293 56# Import media files on pull
6a9e55b0
PV
57my $import_media = run_git("config --get --bool remote.". $remotename .".mediaimport");
58chomp($import_media);
59$import_media = ($import_media eq "true");
60
2045e293
MM
61# Export media files on push
62my $export_media = run_git("config --get --bool remote.". $remotename .".mediaexport");
63chomp($export_media);
64$export_media = !($export_media eq "false");
65
428c995c 66my $wiki_login = run_git("config --get remote.". $remotename .".mwLogin");
2da7830d 67# Note: mwPassword is discourraged. Use the credential system instead.
428c995c 68my $wiki_passwd = run_git("config --get remote.". $remotename .".mwPassword");
1d6abac9 69my $wiki_domain = run_git("config --get remote.". $remotename .".mwDomain");
428c995c
JN
70chomp($wiki_login);
71chomp($wiki_passwd);
1d6abac9 72chomp($wiki_domain);
428c995c
JN
73
74# Import only last revisions (both for clone and fetch)
75my $shallow_import = run_git("config --get --bool remote.". $remotename .".shallow");
76chomp($shallow_import);
77$shallow_import = ($shallow_import eq "true");
78
5a29217d
MM
79# Fetch (clone and pull) by revisions instead of by pages. This behavior
80# is more efficient when we have a wiki with lots of pages and we fetch
81# the revisions quite often so that they concern only few pages.
82# Possible values:
83# - by_rev: perform one query per new revision on the remote wiki
84# - by_page: query each tracked page for new revision
85my $fetch_strategy = run_git("config --get remote.$remotename.fetchStrategy");
86unless ($fetch_strategy) {
87 $fetch_strategy = run_git("config --get mediawiki.fetchStrategy");
88}
89chomp($fetch_strategy);
90unless ($fetch_strategy) {
91 $fetch_strategy = "by_page";
92}
93
93f0d338
MM
94# Dumb push: don't update notes and mediawiki ref to reflect the last push.
95#
96# Configurable with mediawiki.dumbPush, or per-remote with
97# remote.<remotename>.dumbPush.
98#
99# This means the user will have to re-import the just-pushed
100# revisions. On the other hand, this means that the Git revisions
101# corresponding to MediaWiki revisions are all imported from the wiki,
102# regardless of whether they were initially created in Git or from the
103# web interface, hence all users will get the same history (i.e. if
104# the push from Git to MediaWiki loses some information, everybody
105# will get the history with information lost). If the import is
106# deterministic, this means everybody gets the same sha1 for each
107# MediaWiki revision.
108my $dumb_push = run_git("config --get --bool remote.$remotename.dumbPush");
109unless ($dumb_push) {
110 $dumb_push = run_git("config --get --bool mediawiki.dumbPush");
111}
112chomp($dumb_push);
113$dumb_push = ($dumb_push eq "true");
114
428c995c
JN
115my $wiki_name = $url;
116$wiki_name =~ s/[^\/]*:\/\///;
9fb79503
MM
117# If URL is like http://user:password@example.com/, we clearly don't
118# want the password in $wiki_name. While we're there, also remove user
119# and '@' sign, to avoid author like MWUser@HTTPUser@host.com
120$wiki_name =~ s/^.*@//;
428c995c
JN
121
122# Commands parser
123my $entry;
124my @cmd;
125while (<STDIN>) {
126 chomp;
127 @cmd = split(/ /);
128 if (defined($cmd[0])) {
129 # Line not blank
130 if ($cmd[0] eq "capabilities") {
131 die("Too many arguments for capabilities") unless (!defined($cmd[1]));
132 mw_capabilities();
133 } elsif ($cmd[0] eq "list") {
134 die("Too many arguments for list") unless (!defined($cmd[2]));
135 mw_list($cmd[1]);
136 } elsif ($cmd[0] eq "import") {
137 die("Invalid arguments for import") unless ($cmd[1] ne "" && !defined($cmd[2]));
138 mw_import($cmd[1]);
139 } elsif ($cmd[0] eq "option") {
140 die("Too many arguments for option") unless ($cmd[1] ne "" && $cmd[2] ne "" && !defined($cmd[3]));
141 mw_option($cmd[1],$cmd[2]);
142 } elsif ($cmd[0] eq "push") {
143 mw_push($cmd[1]);
144 } else {
145 print STDERR "Unknown command. Aborting...\n";
146 last;
147 }
148 } else {
149 # blank line: we should terminate
150 last;
151 }
152
153 BEGIN { $| = 1 } # flush STDOUT, to make sure the previous
154 # command is fully processed.
155}
156
157########################## Functions ##############################
158
a7271ad1
MM
159## credential API management (generic functions)
160
a7271ad1
MM
161sub credential_read {
162 my %credential;
163 my $reader = shift;
164 my $op = shift;
165 while (<$reader>) {
166 my ($key, $value) = /([^=]*)=(.*)/;
167 if (not defined $key) {
168 die "ERROR receiving response from git credential $op:\n$_\n";
169 }
170 $credential{$key} = $value;
171 }
172 return %credential;
173}
174
175sub credential_write {
176 my $credential = shift;
177 my $writer = shift;
77eab053
JK
178 # url overwrites other fields, so it must come first
179 print $writer "url=$credential->{url}\n" if exists $credential->{url};
a7271ad1 180 while (my ($key, $value) = each(%$credential) ) {
77eab053 181 if (length $value && $key ne 'url') {
a7271ad1
MM
182 print $writer "$key=$value\n";
183 }
184 }
185}
186
187sub credential_run {
188 my $op = shift;
189 my $credential = shift;
190 my $pid = open2(my $reader, my $writer, "git credential $op");
191 credential_write($credential, $writer);
192 print $writer "\n";
193 close($writer);
194
195 if ($op eq "fill") {
196 %$credential = credential_read($reader, $op);
197 } else {
198 if (<$reader>) {
199 die "ERROR while running git credential $op:\n$_";
200 }
201 }
202 close($reader);
203 waitpid($pid, 0);
204 my $child_exit_status = $? >> 8;
205 if ($child_exit_status != 0) {
206 die "'git credential $op' failed with code $child_exit_status.";
207 }
208}
209
428c995c
JN
210# MediaWiki API instance, created lazily.
211my $mediawiki;
212
213sub mw_connect_maybe {
214 if ($mediawiki) {
a7271ad1 215 return;
428c995c
JN
216 }
217 $mediawiki = MediaWiki::API->new;
218 $mediawiki->{config}->{api_url} = "$url/api.php";
219 if ($wiki_login) {
77eab053 220 my %credential = (url => $url);
a7271ad1
MM
221 $credential{username} = $wiki_login;
222 $credential{password} = $wiki_passwd;
223 credential_run("fill", \%credential);
224 my $request = {lgname => $credential{username},
225 lgpassword => $credential{password},
226 lgdomain => $wiki_domain};
227 if ($mediawiki->login($request)) {
228 credential_run("approve", \%credential);
229 print STDERR "Logged in mediawiki user \"$credential{username}\".\n";
428c995c 230 } else {
a7271ad1
MM
231 print STDERR "Failed to log in mediawiki user \"$credential{username}\" on $url\n";
232 print STDERR " (error " .
233 $mediawiki->{error}->{code} . ': ' .
234 $mediawiki->{error}->{details} . ")\n";
235 credential_run("reject", \%credential);
236 exit 1;
428c995c
JN
237 }
238 }
239}
240
9cb74f36
MM
241## Functions for listing pages on the remote wiki
242sub get_mw_tracked_pages {
243 my $pages = shift;
6df7e0df
MM
244 get_mw_page_list(\@tracked_pages, $pages);
245}
246
247sub get_mw_page_list {
248 my $page_list = shift;
249 my $pages = shift;
250 my @some_pages = @$page_list;
9cb74f36
MM
251 while (@some_pages) {
252 my $last = 50;
253 if ($#some_pages < $last) {
254 $last = $#some_pages;
255 }
256 my @slice = @some_pages[0..$last];
257 get_mw_first_pages(\@slice, $pages);
258 @some_pages = @some_pages[51..$#some_pages];
259 }
260}
261
262sub get_mw_tracked_categories {
263 my $pages = shift;
264 foreach my $category (@tracked_categories) {
265 if (index($category, ':') < 0) {
266 # Mediawiki requires the Category
267 # prefix, but let's not force the user
268 # to specify it.
269 $category = "Category:" . $category;
270 }
271 my $mw_pages = $mediawiki->list( {
272 action => 'query',
273 list => 'categorymembers',
274 cmtitle => $category,
275 cmlimit => 'max' } )
276 || die $mediawiki->{error}->{code} . ': '
277 . $mediawiki->{error}->{details};
278 foreach my $page (@{$mw_pages}) {
279 $pages->{$page->{title}} = $page;
280 }
281 }
282}
283
284sub get_mw_all_pages {
285 my $pages = shift;
286 # No user-provided list, get the list of pages from the API.
287 my $mw_pages = $mediawiki->list({
288 action => 'query',
289 list => 'allpages',
290 aplimit => 'max'
291 });
292 if (!defined($mw_pages)) {
293 print STDERR "fatal: could not get the list of wiki pages.\n";
294 print STDERR "fatal: '$url' does not appear to be a mediawiki\n";
295 print STDERR "fatal: make sure '$url/api.php' is a valid page.\n";
296 exit 1;
297 }
298 foreach my $page (@{$mw_pages}) {
299 $pages->{$page->{title}} = $page;
300 }
301}
302
303# queries the wiki for a set of pages. Meant to be used within a loop
304# querying the wiki for slices of page list.
428c995c
JN
305sub get_mw_first_pages {
306 my $some_pages = shift;
307 my @some_pages = @{$some_pages};
308
309 my $pages = shift;
310
311 # pattern 'page1|page2|...' required by the API
312 my $titles = join('|', @some_pages);
313
314 my $mw_pages = $mediawiki->api({
315 action => 'query',
316 titles => $titles,
317 });
318 if (!defined($mw_pages)) {
319 print STDERR "fatal: could not query the list of wiki pages.\n";
320 print STDERR "fatal: '$url' does not appear to be a mediawiki\n";
321 print STDERR "fatal: make sure '$url/api.php' is a valid page.\n";
322 exit 1;
323 }
324 while (my ($id, $page) = each(%{$mw_pages->{query}->{pages}})) {
325 if ($id < 0) {
326 print STDERR "Warning: page $page->{title} not found on wiki\n";
327 } else {
328 $pages->{$page->{title}} = $page;
329 }
330 }
331}
332
9cb74f36 333# Get the list of pages to be fetched according to configuration.
428c995c
JN
334sub get_mw_pages {
335 mw_connect_maybe();
336
f690ddfd
MM
337 print STDERR "Listing pages on remote wiki...\n";
338
428c995c
JN
339 my %pages; # hash on page titles to avoid duplicates
340 my $user_defined;
341 if (@tracked_pages) {
342 $user_defined = 1;
343 # The user provided a list of pages titles, but we
344 # still need to query the API to get the page IDs.
9cb74f36 345 get_mw_tracked_pages(\%pages);
428c995c
JN
346 }
347 if (@tracked_categories) {
348 $user_defined = 1;
9cb74f36 349 get_mw_tracked_categories(\%pages);
428c995c
JN
350 }
351 if (!$user_defined) {
9cb74f36 352 get_mw_all_pages(\%pages);
428c995c 353 }
6a9e55b0
PV
354 if ($import_media) {
355 print STDERR "Getting media files for selected pages...\n";
356 if ($user_defined) {
357 get_linked_mediafiles(\%pages);
358 } else {
359 get_all_mediafiles(\%pages);
428c995c
JN
360 }
361 }
f690ddfd 362 print STDERR (scalar keys %pages) . " pages found.\n";
b1ede9a9 363 return %pages;
428c995c
JN
364}
365
b3d98595
NKT
366# usage: $out = run_git("command args");
367# $out = run_git("command args", "raw"); # don't interpret output as UTF-8.
428c995c 368sub run_git {
b3d98595
NKT
369 my $args = shift;
370 my $encoding = (shift || "encoding(UTF-8)");
371 open(my $git, "-|:$encoding", "git " . $args);
428c995c
JN
372 my $res = do { local $/; <$git> };
373 close($git);
374
375 return $res;
376}
377
378
6a9e55b0
PV
379sub get_all_mediafiles {
380 my $pages = shift;
381 # Attach list of all pages for media files from the API,
382 # they are in a different namespace, only one namespace
383 # can be queried at the same moment
384 my $mw_pages = $mediawiki->list({
385 action => 'query',
386 list => 'allpages',
387 apnamespace => get_mw_namespace_id("File"),
388 aplimit => 'max'
389 });
390 if (!defined($mw_pages)) {
391 print STDERR "fatal: could not get the list of pages for media files.\n";
392 print STDERR "fatal: '$url' does not appear to be a mediawiki\n";
393 print STDERR "fatal: make sure '$url/api.php' is a valid page.\n";
394 exit 1;
395 }
396 foreach my $page (@{$mw_pages}) {
397 $pages->{$page->{title}} = $page;
398 }
399}
400
401sub get_linked_mediafiles {
402 my $pages = shift;
403 my @titles = map $_->{title}, values(%{$pages});
404
405 # The query is split in small batches because of the MW API limit of
406 # the number of links to be returned (500 links max).
407 my $batch = 10;
408 while (@titles) {
409 if ($#titles < $batch) {
410 $batch = $#titles;
411 }
412 my @slice = @titles[0..$batch];
413
414 # pattern 'page1|page2|...' required by the API
415 my $mw_titles = join('|', @slice);
416
417 # Media files could be included or linked from
418 # a page, get all related
419 my $query = {
420 action => 'query',
421 prop => 'links|images',
422 titles => $mw_titles,
423 plnamespace => get_mw_namespace_id("File"),
424 pllimit => 'max'
425 };
426 my $result = $mediawiki->api($query);
427
428 while (my ($id, $page) = each(%{$result->{query}->{pages}})) {
6df7e0df 429 my @media_titles;
6a9e55b0
PV
430 if (defined($page->{links})) {
431 my @link_titles = map $_->{title}, @{$page->{links}};
6df7e0df 432 push(@media_titles, @link_titles);
6a9e55b0
PV
433 }
434 if (defined($page->{images})) {
435 my @image_titles = map $_->{title}, @{$page->{images}};
6df7e0df 436 push(@media_titles, @image_titles);
6a9e55b0 437 }
6df7e0df
MM
438 if (@media_titles) {
439 get_mw_page_list(\@media_titles, $pages);
6a9e55b0
PV
440 }
441 }
442
443 @titles = @titles[($batch+1)..$#titles];
444 }
445}
446
447sub get_mw_mediafile_for_page_revision {
448 # Name of the file on Wiki, with the prefix.
6df7e0df 449 my $filename = shift;
6a9e55b0
PV
450 my $timestamp = shift;
451 my %mediafile;
452
6df7e0df
MM
453 # Search if on a media file with given timestamp exists on
454 # MediaWiki. In that case download the file.
6a9e55b0
PV
455 my $query = {
456 action => 'query',
457 prop => 'imageinfo',
6df7e0df 458 titles => "File:" . $filename,
6a9e55b0
PV
459 iistart => $timestamp,
460 iiend => $timestamp,
461 iiprop => 'timestamp|archivename|url',
462 iilimit => 1
463 };
464 my $result = $mediawiki->api($query);
465
6df7e0df 466 my ($fileid, $file) = each( %{$result->{query}->{pages}} );
6a9e55b0
PV
467 # If not defined it means there is no revision of the file for
468 # given timestamp.
469 if (defined($file->{imageinfo})) {
6a9e55b0
PV
470 $mediafile{title} = $filename;
471
472 my $fileinfo = pop(@{$file->{imageinfo}});
473 $mediafile{timestamp} = $fileinfo->{timestamp};
6df7e0df
MM
474 # Mediawiki::API's download function doesn't support https URLs
475 # and can't download old versions of files.
476 print STDERR "\tDownloading file $mediafile{title}, version $mediafile{timestamp}\n";
477 $mediafile{content} = download_mw_mediafile($fileinfo->{url});
6a9e55b0
PV
478 }
479 return %mediafile;
480}
481
6df7e0df 482sub download_mw_mediafile {
6a9e55b0 483 my $url = shift;
6a9e55b0 484
6df7e0df
MM
485 my $response = $mediawiki->{ua}->get($url);
486 if ($response->code == 200) {
487 return $response->decoded_content;
6a9e55b0 488 } else {
6df7e0df
MM
489 print STDERR "Error downloading mediafile from :\n";
490 print STDERR "URL: $url\n";
491 print STDERR "Server response: " . $response->code . " " . $response->message . "\n";
6a9e55b0 492 exit 1;
6a9e55b0
PV
493 }
494}
495
428c995c
JN
496sub get_last_local_revision {
497 # Get note regarding last mediawiki revision
498 my $note = run_git("notes --ref=$remotename/mediawiki show refs/mediawiki/$remotename/master 2>/dev/null");
499 my @note_info = split(/ /, $note);
500
501 my $lastrevision_number;
502 if (!(defined($note_info[0]) && $note_info[0] eq "mediawiki_revision:")) {
503 print STDERR "No previous mediawiki revision found";
504 $lastrevision_number = 0;
505 } else {
506 # Notes are formatted : mediawiki_revision: #number
507 $lastrevision_number = $note_info[1];
508 chomp($lastrevision_number);
509 print STDERR "Last local mediawiki revision found is $lastrevision_number";
510 }
511 return $lastrevision_number;
512}
513
3c1ed90e
MM
514# Remember the timestamp corresponding to a revision id.
515my %basetimestamps;
516
5a29217d
MM
517# Get the last remote revision without taking in account which pages are
518# tracked or not. This function makes a single request to the wiki thus
519# avoid a loop onto all tracked pages. This is useful for the fetch-by-rev
520# option.
521sub get_last_global_remote_rev {
522 mw_connect_maybe();
523
524 my $query = {
525 action => 'query',
526 list => 'recentchanges',
527 prop => 'revisions',
528 rclimit => '1',
529 rcdir => 'older',
530 };
531 my $result = $mediawiki->api($query);
532 return $result->{query}->{recentchanges}[0]->{revid};
533}
534
535# Get the last remote revision concerning the tracked pages and the tracked
536# categories.
428c995c
JN
537sub get_last_remote_revision {
538 mw_connect_maybe();
539
b1ede9a9
MM
540 my %pages_hash = get_mw_pages();
541 my @pages = values(%pages_hash);
428c995c
JN
542
543 my $max_rev_num = 0;
544
a393f488
MM
545 print STDERR "Getting last revision id on tracked pages...\n";
546
428c995c
JN
547 foreach my $page (@pages) {
548 my $id = $page->{pageid};
549
550 my $query = {
551 action => 'query',
552 prop => 'revisions',
3c1ed90e 553 rvprop => 'ids|timestamp',
428c995c
JN
554 pageids => $id,
555 };
556
557 my $result = $mediawiki->api($query);
558
559 my $lastrev = pop(@{$result->{query}->{pages}->{$id}->{revisions}});
560
3c1ed90e
MM
561 $basetimestamps{$lastrev->{revid}} = $lastrev->{timestamp};
562
428c995c
JN
563 $max_rev_num = ($lastrev->{revid} > $max_rev_num ? $lastrev->{revid} : $max_rev_num);
564 }
565
566 print STDERR "Last remote revision found is $max_rev_num.\n";
567 return $max_rev_num;
568}
569
570# Clean content before sending it to MediaWiki
571sub mediawiki_clean {
572 my $string = shift;
573 my $page_created = shift;
574 # Mediawiki does not allow blank space at the end of a page and ends with a single \n.
575 # This function right trims a string and adds a \n at the end to follow this rule
576 $string =~ s/\s+$//;
577 if ($string eq "" && $page_created) {
578 # Creating empty pages is forbidden.
579 $string = EMPTY_CONTENT;
580 }
581 return $string."\n";
582}
583
584# Filter applied on MediaWiki data before adding them to Git
585sub mediawiki_smudge {
586 my $string = shift;
587 if ($string eq EMPTY_CONTENT) {
588 $string = "";
589 }
590 # This \n is important. This is due to mediawiki's way to handle end of files.
591 return $string."\n";
592}
593
594sub mediawiki_clean_filename {
595 my $filename = shift;
596 $filename =~ s/@{[SLASH_REPLACEMENT]}/\//g;
597 # [, ], |, {, and } are forbidden by MediaWiki, even URL-encoded.
598 # Do a variant of URL-encoding, i.e. looks like URL-encoding,
599 # but with _ added to prevent MediaWiki from thinking this is
600 # an actual special character.
601 $filename =~ s/[\[\]\{\}\|]/sprintf("_%%_%x", ord($&))/ge;
602 # If we use the uri escape before
603 # we should unescape here, before anything
604
605 return $filename;
606}
607
608sub mediawiki_smudge_filename {
609 my $filename = shift;
610 $filename =~ s/\//@{[SLASH_REPLACEMENT]}/g;
611 $filename =~ s/ /_/g;
612 # Decode forbidden characters encoded in mediawiki_clean_filename
613 $filename =~ s/_%_([0-9a-fA-F][0-9a-fA-F])/sprintf("%c", hex($1))/ge;
614 return $filename;
615}
616
617sub literal_data {
618 my ($content) = @_;
619 print STDOUT "data ", bytes::length($content), "\n", $content;
620}
621
6a9e55b0
PV
622sub literal_data_raw {
623 # Output possibly binary content.
624 my ($content) = @_;
625 # Avoid confusion between size in bytes and in characters
626 utf8::downgrade($content);
627 binmode STDOUT, ":raw";
628 print STDOUT "data ", bytes::length($content), "\n", $content;
629 binmode STDOUT, ":utf8";
630}
631
428c995c
JN
632sub mw_capabilities {
633 # Revisions are imported to the private namespace
634 # refs/mediawiki/$remotename/ by the helper and fetched into
635 # refs/remotes/$remotename later by fetch.
636 print STDOUT "refspec refs/heads/*:refs/mediawiki/$remotename/*\n";
637 print STDOUT "import\n";
638 print STDOUT "list\n";
639 print STDOUT "push\n";
640 print STDOUT "\n";
641}
642
643sub mw_list {
644 # MediaWiki do not have branches, we consider one branch arbitrarily
645 # called master, and HEAD pointing to it.
646 print STDOUT "? refs/heads/master\n";
647 print STDOUT "\@refs/heads/master HEAD\n";
648 print STDOUT "\n";
649}
650
651sub mw_option {
652 print STDERR "remote-helper command 'option $_[0]' not yet implemented\n";
653 print STDOUT "unsupported\n";
654}
655
656sub fetch_mw_revisions_for_page {
657 my $page = shift;
658 my $id = shift;
659 my $fetch_from = shift;
660 my @page_revs = ();
661 my $query = {
662 action => 'query',
663 prop => 'revisions',
664 rvprop => 'ids',
665 rvdir => 'newer',
666 rvstartid => $fetch_from,
667 rvlimit => 500,
668 pageids => $id,
669 };
670
671 my $revnum = 0;
672 # Get 500 revisions at a time due to the mediawiki api limit
673 while (1) {
674 my $result = $mediawiki->api($query);
675
676 # Parse each of those 500 revisions
677 foreach my $revision (@{$result->{query}->{pages}->{$id}->{revisions}}) {
678 my $page_rev_ids;
679 $page_rev_ids->{pageid} = $page->{pageid};
680 $page_rev_ids->{revid} = $revision->{revid};
681 push(@page_revs, $page_rev_ids);
682 $revnum++;
683 }
684 last unless $result->{'query-continue'};
685 $query->{rvstartid} = $result->{'query-continue'}->{revisions}->{rvstartid};
686 }
687 if ($shallow_import && @page_revs) {
688 print STDERR " Found 1 revision (shallow import).\n";
689 @page_revs = sort {$b->{revid} <=> $a->{revid}} (@page_revs);
690 return $page_revs[0];
691 }
692 print STDERR " Found ", $revnum, " revision(s).\n";
693 return @page_revs;
694}
695
696sub fetch_mw_revisions {
697 my $pages = shift; my @pages = @{$pages};
698 my $fetch_from = shift;
699
700 my @revisions = ();
701 my $n = 1;
702 foreach my $page (@pages) {
703 my $id = $page->{pageid};
704
705 print STDERR "page $n/", scalar(@pages), ": ". $page->{title} ."\n";
706 $n++;
707 my @page_revs = fetch_mw_revisions_for_page($page, $id, $fetch_from);
708 @revisions = (@page_revs, @revisions);
709 }
710
711 return ($n, @revisions);
712}
713
462d97da
MM
714sub fe_escape_path {
715 my $path = shift;
716 $path =~ s/\\/\\\\/g;
717 $path =~ s/"/\\"/g;
718 $path =~ s/\n/\\n/g;
719 return '"' . $path . '"';
720}
721
428c995c
JN
722sub import_file_revision {
723 my $commit = shift;
724 my %commit = %{$commit};
725 my $full_import = shift;
726 my $n = shift;
6a9e55b0
PV
727 my $mediafile = shift;
728 my %mediafile;
729 if ($mediafile) {
730 %mediafile = %{$mediafile};
731 }
428c995c
JN
732
733 my $title = $commit{title};
734 my $comment = $commit{comment};
735 my $content = $commit{content};
736 my $author = $commit{author};
737 my $date = $commit{date};
738
739 print STDOUT "commit refs/mediawiki/$remotename/master\n";
740 print STDOUT "mark :$n\n";
741 print STDOUT "committer $author <$author\@$wiki_name> ", $date->epoch, " +0000\n";
742 literal_data($comment);
743
744 # If it's not a clone, we need to know where to start from
745 if (!$full_import && $n == 1) {
746 print STDOUT "from refs/mediawiki/$remotename/master^0\n";
747 }
748 if ($content ne DELETED_CONTENT) {
462d97da
MM
749 print STDOUT "M 644 inline " .
750 fe_escape_path($title . ".mw") . "\n";
428c995c 751 literal_data($content);
6a9e55b0 752 if (%mediafile) {
462d97da
MM
753 print STDOUT "M 644 inline "
754 . fe_escape_path($mediafile{title}) . "\n";
6a9e55b0
PV
755 literal_data_raw($mediafile{content});
756 }
428c995c
JN
757 print STDOUT "\n\n";
758 } else {
462d97da 759 print STDOUT "D " . fe_escape_path($title . ".mw") . "\n";
428c995c
JN
760 }
761
762 # mediawiki revision number in the git note
763 if ($full_import && $n == 1) {
764 print STDOUT "reset refs/notes/$remotename/mediawiki\n";
765 }
766 print STDOUT "commit refs/notes/$remotename/mediawiki\n";
767 print STDOUT "committer $author <$author\@$wiki_name> ", $date->epoch, " +0000\n";
768 literal_data("Note added by git-mediawiki during import");
769 if (!$full_import && $n == 1) {
770 print STDOUT "from refs/notes/$remotename/mediawiki^0\n";
771 }
772 print STDOUT "N inline :$n\n";
773 literal_data("mediawiki_revision: " . $commit{mw_revision});
774 print STDOUT "\n\n";
775}
776
777# parse a sequence of
778# <cmd> <arg1>
779# <cmd> <arg2>
780# \n
781# (like batch sequence of import and sequence of push statements)
782sub get_more_refs {
783 my $cmd = shift;
784 my @refs;
785 while (1) {
786 my $line = <STDIN>;
787 if ($line =~ m/^$cmd (.*)$/) {
788 push(@refs, $1);
789 } elsif ($line eq "\n") {
790 return @refs;
791 } else {
792 die("Invalid command in a '$cmd' batch: ". $_);
793 }
794 }
795}
796
797sub mw_import {
798 # multiple import commands can follow each other.
799 my @refs = (shift, get_more_refs("import"));
800 foreach my $ref (@refs) {
801 mw_import_ref($ref);
802 }
803 print STDOUT "done\n";
804}
805
806sub mw_import_ref {
807 my $ref = shift;
808 # The remote helper will call "import HEAD" and
809 # "import refs/heads/master".
810 # Since HEAD is a symbolic ref to master (by convention,
811 # followed by the output of the command "list" that we gave),
812 # we don't need to do anything in this case.
813 if ($ref eq "HEAD") {
814 return;
815 }
816
817 mw_connect_maybe();
818
428c995c
JN
819 print STDERR "Searching revisions...\n";
820 my $last_local = get_last_local_revision();
821 my $fetch_from = $last_local + 1;
822 if ($fetch_from == 1) {
823 print STDERR ", fetching from beginning.\n";
824 } else {
825 print STDERR ", fetching from here.\n";
826 }
428c995c 827
5a29217d
MM
828 my $n = 0;
829 if ($fetch_strategy eq "by_rev") {
830 print STDERR "Fetching & writing export data by revs...\n";
831 $n = mw_import_ref_by_revs($fetch_from);
832 } elsif ($fetch_strategy eq "by_page") {
833 print STDERR "Fetching & writing export data by pages...\n";
834 $n = mw_import_ref_by_pages($fetch_from);
835 } else {
836 print STDERR "fatal: invalid fetch strategy \"$fetch_strategy\".\n";
837 print STDERR "Check your configuration variables remote.$remotename.fetchStrategy and mediawiki.fetchStrategy\n";
838 exit 1;
839 }
840
841 if ($fetch_from == 1 && $n == 0) {
842 print STDERR "You appear to have cloned an empty MediaWiki.\n";
843 # Something has to be done remote-helper side. If nothing is done, an error is
2582ab18 844 # thrown saying that HEAD is referring to unknown object 0000000000000000000
5a29217d
MM
845 # and the clone fails.
846 }
847}
848
849sub mw_import_ref_by_pages {
850
851 my $fetch_from = shift;
852 my %pages_hash = get_mw_pages();
853 my @pages = values(%pages_hash);
854
428c995c
JN
855 my ($n, @revisions) = fetch_mw_revisions(\@pages, $fetch_from);
856
4465b6d6
MM
857 @revisions = sort {$a->{revid} <=> $b->{revid}} @revisions;
858 my @revision_ids = map $_->{revid}, @revisions;
859
312fa9aa
MM
860 return mw_import_revids($fetch_from, \@revision_ids, \%pages_hash);
861}
862
5a29217d
MM
863sub mw_import_ref_by_revs {
864
865 my $fetch_from = shift;
866 my %pages_hash = get_mw_pages();
428c995c 867
5a29217d
MM
868 my $last_remote = get_last_global_remote_rev();
869 my @revision_ids = $fetch_from..$last_remote;
870 return mw_import_revids($fetch_from, \@revision_ids, \%pages_hash);
871}
872
873# Import revisions given in second argument (array of integers).
874# Only pages appearing in the third argument (hash indexed by page titles)
875# will be imported.
312fa9aa
MM
876sub mw_import_revids {
877 my $fetch_from = shift;
878 my $revision_ids = shift;
879 my $pages = shift;
880
881 my $n = 0;
5a29217d 882 my $n_actual = 0;
428c995c
JN
883 my $last_timestamp = 0; # Placeholer in case $rev->timestamp is undefined
884
312fa9aa 885 foreach my $pagerevid (@$revision_ids) {
ebd5fe1c
MM
886 # Count page even if we skip it, since we display
887 # $n/$total and $total includes skipped pages.
888 $n++;
889
428c995c
JN
890 # fetch the content of the pages
891 my $query = {
892 action => 'query',
893 prop => 'revisions',
894 rvprop => 'content|timestamp|comment|user|ids',
4465b6d6 895 revids => $pagerevid,
428c995c
JN
896 };
897
898 my $result = $mediawiki->api($query);
899
ff0bfd75
MM
900 if (!$result) {
901 die "Failed to retrieve modified page for revision $pagerevid";
902 }
903
ebd5fe1c
MM
904 if (defined($result->{query}->{badrevids}->{$pagerevid})) {
905 # The revision id does not exist on the remote wiki.
906 next;
907 }
908
ff0bfd75
MM
909 if (!defined($result->{query}->{pages})) {
910 die "Invalid revision $pagerevid.";
911 }
912
4465b6d6
MM
913 my @result_pages = values(%{$result->{query}->{pages}});
914 my $result_page = $result_pages[0];
915 my $rev = $result_pages[0]->{revisions}->[0];
428c995c 916
4465b6d6 917 my $page_title = $result_page->{title};
5a29217d
MM
918
919 if (!exists($pages->{$page_title})) {
920 print STDERR "$n/", scalar(@$revision_ids),
921 ": Skipping revision #$rev->{revid} of $page_title\n";
922 next;
923 }
924
925 $n_actual++;
926
428c995c
JN
927 my %commit;
928 $commit{author} = $rev->{user} || 'Anonymous';
dcb1ea62 929 $commit{comment} = $rev->{comment} || EMPTY_MESSAGE;
6a9e55b0 930 $commit{title} = mediawiki_smudge_filename($page_title);
4465b6d6 931 $commit{mw_revision} = $rev->{revid};
428c995c
JN
932 $commit{content} = mediawiki_smudge($rev->{'*'});
933
934 if (!defined($rev->{timestamp})) {
935 $last_timestamp++;
936 } else {
937 $last_timestamp = $rev->{timestamp};
938 }
939 $commit{date} = DateTime::Format::ISO8601->parse_datetime($last_timestamp);
940
6a9e55b0 941 # Differentiates classic pages and media files.
6df7e0df 942 my ($namespace, $filename) = $page_title =~ /^([^:]*):(.*)$/;
6a9e55b0 943 my %mediafile;
2272400c
MM
944 if ($namespace) {
945 my $id = get_mw_namespace_id($namespace);
946 if ($id && $id == get_mw_namespace_id("File")) {
947 %mediafile = get_mw_mediafile_for_page_revision($filename, $rev->{timestamp});
948 }
6a9e55b0
PV
949 }
950 # If this is a revision of the media page for new version
951 # of a file do one common commit for both file and media page.
952 # Else do commit only for that page.
312fa9aa 953 print STDERR "$n/", scalar(@$revision_ids), ": Revision #$rev->{revid} of $commit{title}\n";
5a29217d 954 import_file_revision(\%commit, ($fetch_from == 1), $n_actual, \%mediafile);
428c995c
JN
955 }
956
5a29217d 957 return $n_actual;
428c995c
JN
958}
959
960sub error_non_fast_forward {
fd47d7b9
MM
961 my $advice = run_git("config --bool advice.pushNonFastForward");
962 chomp($advice);
963 if ($advice ne "false") {
964 # Native git-push would show this after the summary.
965 # We can't ask it to display it cleanly, so print it
966 # ourselves before.
967 print STDERR "To prevent you from losing history, non-fast-forward updates were rejected\n";
968 print STDERR "Merge the remote changes (e.g. 'git pull') before pushing again. See the\n";
969 print STDERR "'Note about fast-forwards' section of 'git push --help' for details.\n";
970 }
428c995c
JN
971 print STDOUT "error $_[0] \"non-fast-forward\"\n";
972 return 0;
973}
974
b3d98595
NKT
975sub mw_upload_file {
976 my $complete_file_name = shift;
977 my $new_sha1 = shift;
978 my $extension = shift;
979 my $file_deleted = shift;
980 my $summary = shift;
981 my $newrevid;
982 my $path = "File:" . $complete_file_name;
983 my %hashFiles = get_allowed_file_extensions();
984 if (!exists($hashFiles{$extension})) {
985 print STDERR "$complete_file_name is not a permitted file on this wiki.\n";
986 print STDERR "Check the configuration of file uploads in your mediawiki.\n";
987 return $newrevid;
988 }
989 # Deleting and uploading a file requires a priviledged user
990 if ($file_deleted) {
991 mw_connect_maybe();
992 my $query = {
993 action => 'delete',
994 title => $path,
995 reason => $summary
996 };
997 if (!$mediawiki->edit($query)) {
998 print STDERR "Failed to delete file on remote wiki\n";
999 print STDERR "Check your permissions on the remote site. Error code:\n";
1000 print STDERR $mediawiki->{error}->{code} . ':' . $mediawiki->{error}->{details};
1001 exit 1;
1002 }
1003 } else {
1004 # Don't let perl try to interpret file content as UTF-8 => use "raw"
1005 my $content = run_git("cat-file blob $new_sha1", "raw");
1006 if ($content ne "") {
1007 mw_connect_maybe();
1008 $mediawiki->{config}->{upload_url} =
1009 "$url/index.php/Special:Upload";
1010 $mediawiki->edit({
1011 action => 'upload',
1012 filename => $complete_file_name,
1013 comment => $summary,
1014 file => [undef,
1015 $complete_file_name,
1016 Content => $content],
1017 ignorewarnings => 1,
1018 }, {
1019 skip_encoding => 1
1020 } ) || die $mediawiki->{error}->{code} . ':'
1021 . $mediawiki->{error}->{details};
1022 my $last_file_page = $mediawiki->get_page({title => $path});
1023 $newrevid = $last_file_page->{revid};
1024 print STDERR "Pushed file: $new_sha1 - $complete_file_name.\n";
1025 } else {
1026 print STDERR "Empty file $complete_file_name not pushed.\n";
1027 }
1028 }
1029 return $newrevid;
1030}
1031
428c995c
JN
1032sub mw_push_file {
1033 my $diff_info = shift;
1034 # $diff_info contains a string in this format:
1035 # 100644 100644 <sha1_of_blob_before_commit> <sha1_of_blob_now> <status>
1036 my @diff_info_split = split(/[ \t]/, $diff_info);
1037
1038 # Filename, including .mw extension
1039 my $complete_file_name = shift;
1040 # Commit message
1041 my $summary = shift;
93f0d338
MM
1042 # MediaWiki revision number. Keep the previous one by default,
1043 # in case there's no edit to perform.
b3d98595
NKT
1044 my $oldrevid = shift;
1045 my $newrevid;
428c995c 1046
dcb1ea62
MM
1047 if ($summary eq EMPTY_MESSAGE) {
1048 $summary = '';
1049 }
1050
428c995c
JN
1051 my $new_sha1 = $diff_info_split[3];
1052 my $old_sha1 = $diff_info_split[2];
1053 my $page_created = ($old_sha1 eq NULL_SHA1);
1054 my $page_deleted = ($new_sha1 eq NULL_SHA1);
1055 $complete_file_name = mediawiki_clean_filename($complete_file_name);
1056
b3d98595
NKT
1057 my ($title, $extension) = $complete_file_name =~ /^(.*)\.([^\.]*)$/;
1058 if (!defined($extension)) {
1059 $extension = "";
1060 }
1061 if ($extension eq "mw") {
2045e293
MM
1062 my $ns = get_mw_namespace_id_for_page($complete_file_name);
1063 if ($ns && $ns == get_mw_namespace_id("File") && (!$export_media)) {
1064 print STDERR "Ignoring media file related page: $complete_file_name\n";
1065 return ($oldrevid, "ok");
1066 }
428c995c
JN
1067 my $file_content;
1068 if ($page_deleted) {
1069 # Deleting a page usually requires
2582ab18 1070 # special privileges. A common
428c995c
JN
1071 # convention is to replace the page
1072 # with this content instead:
1073 $file_content = DELETED_CONTENT;
1074 } else {
1075 $file_content = run_git("cat-file blob $new_sha1");
1076 }
1077
1078 mw_connect_maybe();
1079
1080 my $result = $mediawiki->edit( {
1081 action => 'edit',
1082 summary => $summary,
1083 title => $title,
b3d98595 1084 basetimestamp => $basetimestamps{$oldrevid},
428c995c
JN
1085 text => mediawiki_clean($file_content, $page_created),
1086 }, {
1087 skip_encoding => 1 # Helps with names with accentuated characters
3c1ed90e
MM
1088 });
1089 if (!$result) {
1090 if ($mediawiki->{error}->{code} == 3) {
1091 # edit conflicts, considered as non-fast-forward
1092 print STDERR 'Warning: Error ' .
1093 $mediawiki->{error}->{code} .
1094 ' from mediwiki: ' . $mediawiki->{error}->{details} .
1095 ".\n";
b3d98595 1096 return ($oldrevid, "non-fast-forward");
3c1ed90e
MM
1097 } else {
1098 # Other errors. Shouldn't happen => just die()
1099 die 'Fatal: Error ' .
1100 $mediawiki->{error}->{code} .
1101 ' from mediwiki: ' . $mediawiki->{error}->{details};
1102 }
1103 }
93f0d338 1104 $newrevid = $result->{edit}->{newrevid};
ac86ec0f 1105 print STDERR "Pushed file: $new_sha1 - $title\n";
2045e293 1106 } elsif ($export_media) {
b3d98595
NKT
1107 $newrevid = mw_upload_file($complete_file_name, $new_sha1,
1108 $extension, $page_deleted,
1109 $summary);
2045e293
MM
1110 } else {
1111 print STDERR "Ignoring media file $title\n";
428c995c 1112 }
b3d98595 1113 $newrevid = ($newrevid or $oldrevid);
3c1ed90e 1114 return ($newrevid, "ok");
428c995c
JN
1115}
1116
1117sub mw_push {
1118 # multiple push statements can follow each other
1119 my @refsspecs = (shift, get_more_refs("push"));
428c995c
JN
1120 my $pushed;
1121 for my $refspec (@refsspecs) {
1122 my ($force, $local, $remote) = $refspec =~ /^(\+)?([^:]*):([^:]*)$/
1123 or die("Invalid refspec for push. Expected <src>:<dst> or +<src>:<dst>");
1124 if ($force) {
1125 print STDERR "Warning: forced push not allowed on a MediaWiki.\n";
1126 }
1127 if ($local eq "") {
1128 print STDERR "Cannot delete remote branch on a MediaWiki\n";
1129 print STDOUT "error $remote cannot delete\n";
1130 next;
1131 }
1132 if ($remote ne "refs/heads/master") {
1133 print STDERR "Only push to the branch 'master' is supported on a MediaWiki\n";
1134 print STDOUT "error $remote only master allowed\n";
1135 next;
1136 }
1137 if (mw_push_revision($local, $remote)) {
1138 $pushed = 1;
1139 }
1140 }
1141
1142 # Notify Git that the push is done
1143 print STDOUT "\n";
1144
93f0d338 1145 if ($pushed && $dumb_push) {
428c995c
JN
1146 print STDERR "Just pushed some revisions to MediaWiki.\n";
1147 print STDERR "The pushed revisions now have to be re-imported, and your current branch\n";
1148 print STDERR "needs to be updated with these re-imported commits. You can do this with\n";
1149 print STDERR "\n";
1150 print STDERR " git pull --rebase\n";
1151 print STDERR "\n";
1152 }
1153}
1154
1155sub mw_push_revision {
1156 my $local = shift;
1157 my $remote = shift; # actually, this has to be "refs/heads/master" at this point.
1158 my $last_local_revid = get_last_local_revision();
1159 print STDERR ".\n"; # Finish sentence started by get_last_local_revision()
1160 my $last_remote_revid = get_last_remote_revision();
93f0d338 1161 my $mw_revision = $last_remote_revid;
428c995c
JN
1162
1163 # Get sha1 of commit pointed by local HEAD
1164 my $HEAD_sha1 = run_git("rev-parse $local 2>/dev/null"); chomp($HEAD_sha1);
1165 # Get sha1 of commit pointed by remotes/$remotename/master
1166 my $remoteorigin_sha1 = run_git("rev-parse refs/remotes/$remotename/master 2>/dev/null");
1167 chomp($remoteorigin_sha1);
1168
1169 if ($last_local_revid > 0 &&
1170 $last_local_revid < $last_remote_revid) {
1171 return error_non_fast_forward($remote);
1172 }
1173
1174 if ($HEAD_sha1 eq $remoteorigin_sha1) {
1175 # nothing to push
1176 return 0;
1177 }
1178
1179 # Get every commit in between HEAD and refs/remotes/origin/master,
1180 # including HEAD and refs/remotes/origin/master
1181 my @commit_pairs = ();
1182 if ($last_local_revid > 0) {
1183 my $parsed_sha1 = $remoteorigin_sha1;
1184 # Find a path from last MediaWiki commit to pushed commit
93e92d4d
MM
1185 print STDERR "Computing path from local to remote ...\n";
1186 my @local_ancestry = split(/\n/, run_git("rev-list --boundary --parents $local ^$parsed_sha1"));
1187 my %local_ancestry;
1188 foreach my $line (@local_ancestry) {
1189 if (my ($child, $parents) = $line =~ m/^-?([a-f0-9]+) ([a-f0-9 ]+)/) {
1190 foreach my $parent (split(' ', $parents)) {
1191 $local_ancestry{$parent} = $child;
1192 }
1193 } elsif (!$line =~ m/^([a-f0-9]+)/) {
1194 die "Unexpected output from git rev-list: $line";
1195 }
1196 }
428c995c 1197 while ($parsed_sha1 ne $HEAD_sha1) {
93e92d4d
MM
1198 my $child = $local_ancestry{$parsed_sha1};
1199 if (!$child) {
1200 printf STDERR "Cannot find a path in history from remote commit to last commit\n";
428c995c
JN
1201 return error_non_fast_forward($remote);
1202 }
93e92d4d
MM
1203 push(@commit_pairs, [$parsed_sha1, $child]);
1204 $parsed_sha1 = $child;
428c995c
JN
1205 }
1206 } else {
1207 # No remote mediawiki revision. Export the whole
1208 # history (linearized with --first-parent)
1209 print STDERR "Warning: no common ancestor, pushing complete history\n";
1210 my $history = run_git("rev-list --first-parent --children $local");
1211 my @history = split('\n', $history);
1212 @history = @history[1..$#history];
1213 foreach my $line (reverse @history) {
1214 my @commit_info_split = split(/ |\n/, $line);
1215 push(@commit_pairs, \@commit_info_split);
1216 }
1217 }
1218
1219 foreach my $commit_info_split (@commit_pairs) {
1220 my $sha1_child = @{$commit_info_split}[0];
1221 my $sha1_commit = @{$commit_info_split}[1];
1222 my $diff_infos = run_git("diff-tree -r --raw -z $sha1_child $sha1_commit");
1223 # TODO: we could detect rename, and encode them with a #redirect on the wiki.
1224 # TODO: for now, it's just a delete+add
1225 my @diff_info_list = split(/\0/, $diff_infos);
28c24bd7
MM
1226 # Keep the subject line of the commit message as mediawiki comment for the revision
1227 my $commit_msg = run_git("log --no-walk --format=\"%s\" $sha1_commit");
428c995c
JN
1228 chomp($commit_msg);
1229 # Push every blob
1230 while (@diff_info_list) {
3c1ed90e 1231 my $status;
428c995c
JN
1232 # git diff-tree -z gives an output like
1233 # <metadata>\0<filename1>\0
1234 # <metadata>\0<filename2>\0
1235 # and we've split on \0.
1236 my $info = shift(@diff_info_list);
1237 my $file = shift(@diff_info_list);
3c1ed90e
MM
1238 ($mw_revision, $status) = mw_push_file($info, $file, $commit_msg, $mw_revision);
1239 if ($status eq "non-fast-forward") {
1240 # we may already have sent part of the
1241 # commit to MediaWiki, but it's too
1242 # late to cancel it. Stop the push in
1243 # the middle, but still give an
1244 # accurate error message.
1245 return error_non_fast_forward($remote);
1246 }
1247 if ($status ne "ok") {
1248 die("Unknown error from mw_push_file()");
1249 }
93f0d338
MM
1250 }
1251 unless ($dumb_push) {
5c481745 1252 run_git("notes --ref=$remotename/mediawiki add -f -m \"mediawiki_revision: $mw_revision\" $sha1_commit");
93f0d338 1253 run_git("update-ref -m \"Git-MediaWiki push\" refs/mediawiki/$remotename/master $sha1_commit $sha1_child");
428c995c
JN
1254 }
1255 }
1256
1257 print STDOUT "ok $remote\n";
1258 return 1;
1259}
b3d98595
NKT
1260
1261sub get_allowed_file_extensions {
1262 mw_connect_maybe();
1263
1264 my $query = {
1265 action => 'query',
1266 meta => 'siteinfo',
1267 siprop => 'fileextensions'
1268 };
1269 my $result = $mediawiki->api($query);
1270 my @file_extensions= map $_->{ext},@{$result->{query}->{fileextensions}};
1271 my %hashFile = map {$_ => 1}@file_extensions;
1272
1273 return %hashFile;
1274}
6a9e55b0 1275
6df7e0df
MM
1276# In memory cache for MediaWiki namespace ids.
1277my %namespace_id;
1278
1279# Namespaces whose id is cached in the configuration file
1280# (to avoid duplicates)
1281my %cached_mw_namespace_id;
1282
6a9e55b0
PV
1283# Return MediaWiki id for a canonical namespace name.
1284# Ex.: "File", "Project".
6a9e55b0
PV
1285sub get_mw_namespace_id {
1286 mw_connect_maybe();
1287 my $name = shift;
1288
1289 if (!exists $namespace_id{$name}) {
1290 # Look at configuration file, if the record for that namespace is
6df7e0df 1291 # already cached. Namespaces are stored in form:
6a9e55b0 1292 # "Name_of_namespace:Id_namespace", ex.: "File:6".
ac4bbb41 1293 my @temp = split(/[\n]/, run_git("config --get-all remote."
6df7e0df 1294 . $remotename .".namespaceCache"));
6a9e55b0
PV
1295 chomp(@temp);
1296 foreach my $ns (@temp) {
6df7e0df 1297 my ($n, $id) = split(/:/, $ns);
2272400c
MM
1298 if ($id eq 'notANameSpace') {
1299 $namespace_id{$n} = {is_namespace => 0};
1300 } else {
1301 $namespace_id{$n} = {is_namespace => 1, id => $id};
1302 }
6df7e0df 1303 $cached_mw_namespace_id{$n} = 1;
6a9e55b0
PV
1304 }
1305 }
1306
1307 if (!exists $namespace_id{$name}) {
6df7e0df 1308 print STDERR "Namespace $name not found in cache, querying the wiki ...\n";
6a9e55b0
PV
1309 # NS not found => get namespace id from MW and store it in
1310 # configuration file.
1311 my $query = {
1312 action => 'query',
1313 meta => 'siteinfo',
1314 siprop => 'namespaces'
1315 };
1316 my $result = $mediawiki->api($query);
1317
1318 while (my ($id, $ns) = each(%{$result->{query}->{namespaces}})) {
6df7e0df 1319 if (defined($ns->{id}) && defined($ns->{canonical})) {
2272400c 1320 $namespace_id{$ns->{canonical}} = {is_namespace => 1, id => $ns->{id}};
6df7e0df
MM
1321 if ($ns->{'*'}) {
1322 # alias (e.g. french Fichier: as alias for canonical File:)
2272400c 1323 $namespace_id{$ns->{'*'}} = {is_namespace => 1, id => $ns->{id}};
6df7e0df
MM
1324 }
1325 }
6a9e55b0
PV
1326 }
1327 }
1328
2272400c
MM
1329 my $ns = $namespace_id{$name};
1330 my $id;
6df7e0df 1331
2272400c
MM
1332 unless (defined $ns) {
1333 print STDERR "No such namespace $name on MediaWiki.\n";
1334 $ns = {is_namespace => 0};
1335 $namespace_id{$name} = $ns;
1336 }
1337
1338 if ($ns->{is_namespace}) {
1339 $id = $ns->{id};
1340 }
1341
1342 # Store "notANameSpace" as special value for inexisting namespaces
1343 my $store_id = ($id || 'notANameSpace');
1344
1345 # Store explicitely requested namespaces on disk
1346 if (!exists $cached_mw_namespace_id{$name}) {
1347 run_git("config --add remote.". $remotename
1348 .".namespaceCache \"". $name .":". $store_id ."\"");
1349 $cached_mw_namespace_id{$name} = 1;
6a9e55b0 1350 }
2272400c 1351 return $id;
6a9e55b0 1352}
2045e293
MM
1353
1354sub get_mw_namespace_id_for_page {
1355 if (my ($namespace) = $_[0] =~ /^([^:]*):/) {
1356 return get_mw_namespace_id($namespace);
1357 } else {
1358 return;
1359 }
1360}