contrib/mw-to-git/git-remote-mediawiki.perl

   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 # Documentation & bugtracker: https://github.com/moy/Git-Mediawiki/
  13
  14 use strict;
  15 use MediaWiki::API;
  16 use Git;
  17 use DateTime::Format::ISO8601;
  18 use warnings;
  19
  20 # By default, use UTF-8 to communicate with Git and the user
  21 binmode STDERR, ':encoding(UTF-8)';
  22 binmode STDOUT, ':encoding(UTF-8)';
  23
  24 use URI::Escape;
  25
  26 # Mediawiki filenames can contain forward slashes. This variable decides by which pattern they should be replaced
  27 use constant SLASH_REPLACEMENT => '%2F';
  28
  29 # It's not always possible to delete pages (may require some
  30 # privileges). Deleted pages are replaced with this content.
  31 use constant DELETED_CONTENT => "[[Category:Deleted]]\n";
  32
  33 # It's not possible to create empty pages. New empty files in Git are
  34 # sent with this content instead.
  35 use constant EMPTY_CONTENT => "<!-- empty page -->\n";
  36
  37 # used to reflect file creation or deletion in diff.
  38 use constant NULL_SHA1 => '0000000000000000000000000000000000000000';
  39
  40 # Used on Git's side to reflect empty edit messages on the wiki
  41 use constant EMPTY_MESSAGE => '*Empty MediaWiki Message*';
  42
  43 use constant EMPTY => q{};
  44
  45 # Number of pages taken into account at once in submodule get_mw_page_list
  46 use constant SLICE_SIZE => 50;
  47
  48 # Number of linked mediafile to get at once in get_linked_mediafiles
  49 # The query is split in small batches because of the MW API limit of
  50 # the number of links to be returned (500 links max).
  51 use constant BATCH_SIZE => 10;
  52
  53 use constant HTTP_CODE_OK => 200;
  54
  55 if (@ARGV != 2) {
  56         exit_error_usage();
  57 }
  58
  59 my $remotename = $ARGV[0];
  60 my $url = $ARGV[1];
  61
  62 # Accept both space-separated and multiple keys in config file.
  63 # Spaces should be written as _ anyway because we'll use chomp.
  64 my @tracked_pages = split(/[ \n]/, run_git("config --get-all remote.${remotename}.pages"));
  65 chomp(@tracked_pages);
  66
  67 # Just like @tracked_pages, but for MediaWiki categories.
  68 my @tracked_categories = split(/[ \n]/, run_git("config --get-all remote.${remotename}.categories"));
  69 chomp(@tracked_categories);
  70
  71 # Import media files on pull
  72 my $import_media = run_git("config --get --bool remote.${remotename}.mediaimport");
  73 chomp($import_media);
  74 $import_media = ($import_media eq 'true');
  75
  76 # Export media files on push
  77 my $export_media = run_git("config --get --bool remote.${remotename}.mediaexport");
  78 chomp($export_media);
  79 $export_media = !($export_media eq 'false');
  80
  81 my $wiki_login = run_git("config --get remote.${remotename}.mwLogin");
  82 # Note: mwPassword is discourraged. Use the credential system instead.
  83 my $wiki_passwd = run_git("config --get remote.${remotename}.mwPassword");
  84 my $wiki_domain = run_git("config --get remote.${remotename}.mwDomain");
  85 chomp($wiki_login);
  86 chomp($wiki_passwd);
  87 chomp($wiki_domain);
  88
  89 # Import only last revisions (both for clone and fetch)
  90 my $shallow_import = run_git("config --get --bool remote.${remotename}.shallow");
  91 chomp($shallow_import);
  92 $shallow_import = ($shallow_import eq 'true');
  93
  94 # Fetch (clone and pull) by revisions instead of by pages. This behavior
  95 # is more efficient when we have a wiki with lots of pages and we fetch
  96 # the revisions quite often so that they concern only few pages.
  97 # Possible values:
  98 # - by_rev: perform one query per new revision on the remote wiki
  99 # - by_page: query each tracked page for new revision
 100 my $fetch_strategy = run_git("config --get remote.${remotename}.fetchStrategy");
 101 if (!$fetch_strategy) {
 102         $fetch_strategy = run_git('config --get mediawiki.fetchStrategy');
 103 }
 104 chomp($fetch_strategy);
 105 if (!$fetch_strategy) {
 106         $fetch_strategy = 'by_page';
 107 }
 108
 109 # Remember the timestamp corresponding to a revision id.
 110 my %basetimestamps;
 111
 112 # Dumb push: don't update notes and mediawiki ref to reflect the last push.
 113 #
 114 # Configurable with mediawiki.dumbPush, or per-remote with
 115 # remote.<remotename>.dumbPush.
 116 #
 117 # This means the user will have to re-import the just-pushed
 118 # revisions. On the other hand, this means that the Git revisions
 119 # corresponding to MediaWiki revisions are all imported from the wiki,
 120 # regardless of whether they were initially created in Git or from the
 121 # web interface, hence all users will get the same history (i.e. if
 122 # the push from Git to MediaWiki loses some information, everybody
 123 # will get the history with information lost). If the import is
 124 # deterministic, this means everybody gets the same sha1 for each
 125 # MediaWiki revision.
 126 my $dumb_push = run_git("config --get --bool remote.${remotename}.dumbPush");
 127 if (!$dumb_push) {
 128         $dumb_push = run_git('config --get --bool mediawiki.dumbPush');
 129 }
 130 chomp($dumb_push);
 131 $dumb_push = ($dumb_push eq 'true');
 132
 133 my $wiki_name = $url;
 134 $wiki_name =~ s{[^/]*://}{};
 135 # If URL is like http://user:password@example.com/, we clearly don't
 136 # want the password in $wiki_name. While we're there, also remove user
 137 # and '@' sign, to avoid author like MWUser@HTTPUser@host.com
 138 $wiki_name =~ s/^.*@//;
 139
 140 # Commands parser
 141 while (<STDIN>) {
 142         chomp;
 143
 144         if (!parse_command($_)) {
 145                 last;
 146         }
 147
 148         BEGIN { $| = 1 } # flush STDOUT, to make sure the previous
 149                          # command is fully processed.
 150 }
 151
 152 ########################## Functions ##############################
 153
 154 ## error handling
 155 sub exit_error_usage {
 156         die "ERROR: git-remote-mediawiki module was not called with a correct number of\n" .
 157             "parameters\n" .
 158             "You may obtain this error because you attempted to run the git-remote-mediawiki\n" .
 159             "module directly.\n" .
 160             "This module can be used the following way:\n" .
 161             "\tgit clone mediawiki://<address of a mediawiki>\n" .
 162             "Then, use git commit, push and pull as with every normal git repository.\n";
 163 }
 164
 165 sub parse_command {
 166         my ($line) = @_;
 167         my @cmd = split(/ /, $line);
 168         if (!defined $cmd[0]) {
 169                 return 0;
 170         }
 171         if ($cmd[0] eq 'capabilities') {
 172                 die("Too many arguments for capabilities\n")
 173                     if (defined($cmd[1]));
 174                 mw_capabilities();
 175         } elsif ($cmd[0] eq 'list') {
 176                 die("Too many arguments for list\n") if (defined($cmd[2]));
 177                 mw_list($cmd[1]);
 178         } elsif ($cmd[0] eq 'import') {
 179                 die("Invalid argument for import\n")
 180                     if ($cmd[1] eq EMPTY);
 181                 die("Too many arguments for import\n")
 182                     if (defined($cmd[2]));
 183                 mw_import($cmd[1]);
 184         } elsif ($cmd[0] eq 'option') {
 185                 die("Invalid arguments for option\n")
 186                     if ($cmd[1] eq EMPTY || $cmd[2] eq EMPTY);
 187                 die("Too many arguments for option\n")
 188                     if (defined($cmd[3]));
 189                 mw_option($cmd[1],$cmd[2]);
 190         } elsif ($cmd[0] eq 'push') {
 191                 mw_push($cmd[1]);
 192         } else {
 193                 print {*STDERR} "Unknown command. Aborting...\n";
 194                 return 0;
 195         }
 196         return 1;
 197 }
 198
 199 # MediaWiki API instance, created lazily.
 200 my $mediawiki;
 201
 202 sub mw_connect_maybe {
 203         if ($mediawiki) {
 204                 return;
 205         }
 206         $mediawiki = MediaWiki::API->new;
 207         $mediawiki->{config}->{api_url} = "${url}/api.php";
 208         if ($wiki_login) {
 209                 my %credential = (
 210                         'url' => $url,
 211                         'username' => $wiki_login,
 212                         'password' => $wiki_passwd
 213                 );
 214                 Git::credential(\%credential);
 215                 my $request = {lgname => $credential{username},
 216                                lgpassword => $credential{password},
 217                                lgdomain => $wiki_domain};
 218                 if ($mediawiki->login($request)) {
 219                         Git::credential(\%credential, 'approve');
 220                         print {*STDERR} qq(Logged in mediawiki user "$credential{username}".\n);
 221                 } else {
 222                         print {*STDERR} qq(Failed to log in mediawiki user "$credential{username}" on ${url}\n);
 223                         print {*STDERR} '  (error ' .
 224                                 $mediawiki->{error}->{code} . ': ' .
 225                                 $mediawiki->{error}->{details} . ")\n";
 226                         Git::credential(\%credential, 'reject');
 227                         exit 1;
 228                 }
 229         }
 230         return;
 231 }
 232
 233 sub fatal_mw_error {
 234         my $action = shift;
 235         print STDERR "fatal: could not $action.\n";
 236         print STDERR "fatal: '$url' does not appear to be a mediawiki\n";
 237         if ($url =~ /^https/) {
 238                 print STDERR "fatal: make sure '$url/api.php' is a valid page\n";
 239                 print STDERR "fatal: and the SSL certificate is correct.\n";
 240         } else {
 241                 print STDERR "fatal: make sure '$url/api.php' is a valid page.\n";
 242         }
 243         print STDERR "fatal: (error " .
 244             $mediawiki->{error}->{code} . ': ' .
 245             $mediawiki->{error}->{details} . ")\n";
 246         exit 1;
 247 }
 248
 249 ## Functions for listing pages on the remote wiki
 250 sub get_mw_tracked_pages {
 251         my $pages = shift;
 252         get_mw_page_list(\@tracked_pages, $pages);
 253         return;
 254 }
 255
 256 sub get_mw_page_list {
 257         my $page_list = shift;
 258         my $pages = shift;
 259         my @some_pages = @{$page_list};
 260         while (@some_pages) {
 261                 my $last_page = SLICE_SIZE;
 262                 if ($#some_pages < $last_page) {
 263                         $last_page = $#some_pages;
 264                 }
 265                 my @slice = @some_pages[0..$last_page];
 266                 get_mw_first_pages(\@slice, $pages);
 267                 @some_pages = @some_pages[(SLICE_SIZE + 1)..$#some_pages];
 268         }
 269         return;
 270 }
 271
 272 sub get_mw_tracked_categories {
 273         my $pages = shift;
 274         foreach my $category (@tracked_categories) {
 275                 if (index($category, ':') < 0) {
 276                         # Mediawiki requires the Category
 277                         # prefix, but let's not force the user
 278                         # to specify it.
 279                         $category = "Category:${category}";
 280                 }
 281                 my $mw_pages = $mediawiki->list( {
 282                         action => 'query',
 283                         list => 'categorymembers',
 284                         cmtitle => $category,
 285                         cmlimit => 'max' } )
 286                         || die $mediawiki->{error}->{code} . ': '
 287                                 . $mediawiki->{error}->{details} . "\n";
 288                 foreach my $page (@{$mw_pages}) {
 289                         $pages->{$page->{title}} = $page;
 290                 }
 291         }
 292         return;
 293 }
 294
 295 sub get_mw_all_pages {
 296         my $pages = shift;
 297         # No user-provided list, get the list of pages from the API.
 298         my $mw_pages = $mediawiki->list({
 299                 action => 'query',
 300                 list => 'allpages',
 301                 aplimit => 'max'
 302         });
 303         if (!defined($mw_pages)) {
 304                 fatal_mw_error("get the list of wiki pages");
 305         }
 306         foreach my $page (@{$mw_pages}) {
 307                 $pages->{$page->{title}} = $page;
 308         }
 309         return;
 310 }
 311
 312 # queries the wiki for a set of pages. Meant to be used within a loop
 313 # querying the wiki for slices of page list.
 314 sub get_mw_first_pages {
 315         my $some_pages = shift;
 316         my @some_pages = @{$some_pages};
 317
 318         my $pages = shift;
 319
 320         # pattern 'page1|page2|...' required by the API
 321         my $titles = join('|', @some_pages);
 322
 323         my $mw_pages = $mediawiki->api({
 324                 action => 'query',
 325                 titles => $titles,
 326         });
 327         if (!defined($mw_pages)) {
 328                 fatal_mw_error("query the list of wiki pages");
 329         }
 330         while (my ($id, $page) = each(%{$mw_pages->{query}->{pages}})) {
 331                 if ($id < 0) {
 332                         print {*STDERR} "Warning: page $page->{title} not found on wiki\n";
 333                 } else {
 334                         $pages->{$page->{title}} = $page;
 335                 }
 336         }
 337         return;
 338 }
 339
 340 # Get the list of pages to be fetched according to configuration.
 341 sub get_mw_pages {
 342         mw_connect_maybe();
 343
 344         print {*STDERR} "Listing pages on remote wiki...\n";
 345
 346         my %pages; # hash on page titles to avoid duplicates
 347         my $user_defined;
 348         if (@tracked_pages) {
 349                 $user_defined = 1;
 350                 # The user provided a list of pages titles, but we
 351                 # still need to query the API to get the page IDs.
 352                 get_mw_tracked_pages(\%pages);
 353         }
 354         if (@tracked_categories) {
 355                 $user_defined = 1;
 356                 get_mw_tracked_categories(\%pages);
 357         }
 358         if (!$user_defined) {
 359                 get_mw_all_pages(\%pages);
 360         }
 361         if ($import_media) {
 362                 print {*STDERR} "Getting media files for selected pages...\n";
 363                 if ($user_defined) {
 364                         get_linked_mediafiles(\%pages);
 365                 } else {
 366                         get_all_mediafiles(\%pages);
 367                 }
 368         }
 369         print {*STDERR} (scalar keys %pages) . " pages found.\n";
 370         return %pages;
 371 }
 372
 373 # usage: $out = run_git("command args");
 374 #        $out = run_git("command args", "raw"); # don't interpret output as UTF-8.
 375 sub run_git {
 376         my $args = shift;
 377         my $encoding = (shift || 'encoding(UTF-8)');
 378         open(my $git, "-|:${encoding}", "git ${args}")
 379             or die "Unable to fork: $!\n";
 380         my $res = do {
 381                 local $/ = undef;
 382                 <$git>
 383         };
 384         close($git);
 385
 386         return $res;
 387 }
 388
 389
 390 sub get_all_mediafiles {
 391         my $pages = shift;
 392         # Attach list of all pages for media files from the API,
 393         # they are in a different namespace, only one namespace
 394         # can be queried at the same moment
 395         my $mw_pages = $mediawiki->list({
 396                 action => 'query',
 397                 list => 'allpages',
 398                 apnamespace => get_mw_namespace_id('File'),
 399                 aplimit => 'max'
 400         });
 401         if (!defined($mw_pages)) {
 402                 print {*STDERR} "fatal: could not get the list of pages for media files.\n";
 403                 print {*STDERR} "fatal: '$url' does not appear to be a mediawiki\n";
 404                 print {*STDERR} "fatal: make sure '$url/api.php' is a valid page.\n";
 405                 exit 1;
 406         }
 407         foreach my $page (@{$mw_pages}) {
 408                 $pages->{$page->{title}} = $page;
 409         }
 410         return;
 411 }
 412
 413 sub get_linked_mediafiles {
 414         my $pages = shift;
 415         my @titles = map { $_->{title} } values(%{$pages});
 416
 417         my $batch = BATCH_SIZE;
 418         while (@titles) {
 419                 if ($#titles < $batch) {
 420                         $batch = $#titles;
 421                 }
 422                 my @slice = @titles[0..$batch];
 423
 424                 # pattern 'page1|page2|...' required by the API
 425                 my $mw_titles = join('|', @slice);
 426
 427                 # Media files could be included or linked from
 428                 # a page, get all related
 429                 my $query = {
 430                         action => 'query',
 431                         prop => 'links|images',
 432                         titles => $mw_titles,
 433                         plnamespace => get_mw_namespace_id('File'),
 434                         pllimit => 'max'
 435                 };
 436                 my $result = $mediawiki->api($query);
 437
 438                 while (my ($id, $page) = each(%{$result->{query}->{pages}})) {
 439                         my @media_titles;
 440                         if (defined($page->{links})) {
 441                                 my @link_titles
 442                                     = map { $_->{title} } @{$page->{links}};
 443                                 push(@media_titles, @link_titles);
 444                         }
 445                         if (defined($page->{images})) {
 446                                 my @image_titles
 447                                     = map { $_->{title} } @{$page->{images}};
 448                                 push(@media_titles, @image_titles);
 449                         }
 450                         if (@media_titles) {
 451                                 get_mw_page_list(\@media_titles, $pages);
 452                         }
 453                 }
 454
 455                 @titles = @titles[($batch+1)..$#titles];
 456         }
 457         return;
 458 }
 459
 460 sub get_mw_mediafile_for_page_revision {
 461         # Name of the file on Wiki, with the prefix.
 462         my $filename = shift;
 463         my $timestamp = shift;
 464         my %mediafile;
 465
 466         # Search if on a media file with given timestamp exists on
 467         # MediaWiki. In that case download the file.
 468         my $query = {
 469                 action => 'query',
 470                 prop => 'imageinfo',
 471                 titles => "File:${filename}",
 472                 iistart => $timestamp,
 473                 iiend => $timestamp,
 474                 iiprop => 'timestamp|archivename|url',
 475                 iilimit => 1
 476         };
 477         my $result = $mediawiki->api($query);
 478
 479         my ($fileid, $file) = each( %{$result->{query}->{pages}} );
 480         # If not defined it means there is no revision of the file for
 481         # given timestamp.
 482         if (defined($file->{imageinfo})) {
 483                 $mediafile{title} = $filename;
 484
 485                 my $fileinfo = pop(@{$file->{imageinfo}});
 486                 $mediafile{timestamp} = $fileinfo->{timestamp};
 487                 # Mediawiki::API's download function doesn't support https URLs
 488                 # and can't download old versions of files.
 489                 print {*STDERR} "\tDownloading file $mediafile{title}, version $mediafile{timestamp}\n";
 490                 $mediafile{content} = download_mw_mediafile($fileinfo->{url});
 491         }
 492         return %mediafile;
 493 }
 494
 495 sub download_mw_mediafile {
 496         my $download_url = shift;
 497
 498         my $response = $mediawiki->{ua}->get($download_url);
 499         if ($response->code == HTTP_CODE_OK) {
 500                 return $response->decoded_content;
 501         } else {
 502                 print {*STDERR} "Error downloading mediafile from :\n";
 503                 print {*STDERR} "URL: ${download_url}\n";
 504                 print {*STDERR} 'Server response: ' . $response->code . q{ } . $response->message . "\n";
 505                 exit 1;
 506         }
 507 }
 508
 509 sub get_last_local_revision {
 510         # Get note regarding last mediawiki revision
 511         my $note = run_git("notes --ref=${remotename}/mediawiki show refs/mediawiki/${remotename}/master 2>/dev/null");
 512         my @note_info = split(/ /, $note);
 513
 514         my $lastrevision_number;
 515         if (!(defined($note_info[0]) && $note_info[0] eq 'mediawiki_revision:')) {
 516                 print {*STDERR} 'No previous mediawiki revision found';
 517                 $lastrevision_number = 0;
 518         } else {
 519                 # Notes are formatted : mediawiki_revision: #number
 520                 $lastrevision_number = $note_info[1];
 521                 chomp($lastrevision_number);
 522                 print {*STDERR} "Last local mediawiki revision found is ${lastrevision_number}";
 523         }
 524         return $lastrevision_number;
 525 }
 526
 527 # Get the last remote revision without taking in account which pages are
 528 # tracked or not. This function makes a single request to the wiki thus
 529 # avoid a loop onto all tracked pages. This is useful for the fetch-by-rev
 530 # option.
 531 sub get_last_global_remote_rev {
 532         mw_connect_maybe();
 533
 534         my $query = {
 535                 action => 'query',
 536                 list => 'recentchanges',
 537                 prop => 'revisions',
 538                 rclimit => '1',
 539                 rcdir => 'older',
 540         };
 541         my $result = $mediawiki->api($query);
 542         return $result->{query}->{recentchanges}[0]->{revid};
 543 }
 544
 545 # Get the last remote revision concerning the tracked pages and the tracked
 546 # categories.
 547 sub get_last_remote_revision {
 548         mw_connect_maybe();
 549
 550         my %pages_hash = get_mw_pages();
 551         my @pages = values(%pages_hash);
 552
 553         my $max_rev_num = 0;
 554
 555         print {*STDERR} "Getting last revision id on tracked pages...\n";
 556
 557         foreach my $page (@pages) {
 558                 my $id = $page->{pageid};
 559
 560                 my $query = {
 561                         action => 'query',
 562                         prop => 'revisions',
 563                         rvprop => 'ids|timestamp',
 564                         pageids => $id,
 565                 };
 566
 567                 my $result = $mediawiki->api($query);
 568
 569                 my $lastrev = pop(@{$result->{query}->{pages}->{$id}->{revisions}});
 570
 571                 $basetimestamps{$lastrev->{revid}} = $lastrev->{timestamp};
 572
 573                 $max_rev_num = ($lastrev->{revid} > $max_rev_num ? $lastrev->{revid} : $max_rev_num);
 574         }
 575
 576         print {*STDERR} "Last remote revision found is $max_rev_num.\n";
 577         return $max_rev_num;
 578 }
 579
 580 # Clean content before sending it to MediaWiki
 581 sub mediawiki_clean {
 582         my $string = shift;
 583         my $page_created = shift;
 584         # Mediawiki does not allow blank space at the end of a page and ends with a single \n.
 585         # This function right trims a string and adds a \n at the end to follow this rule
 586         $string =~ s/\s+$//;
 587         if ($string eq EMPTY && $page_created) {
 588                 # Creating empty pages is forbidden.
 589                 $string = EMPTY_CONTENT;
 590         }
 591         return $string."\n";
 592 }
 593
 594 # Filter applied on MediaWiki data before adding them to Git
 595 sub mediawiki_smudge {
 596         my $string = shift;
 597         if ($string eq EMPTY_CONTENT) {
 598                 $string = EMPTY;
 599         }
 600         # This \n is important. This is due to mediawiki's way to handle end of files.
 601         return "${string}\n";
 602 }
 603
 604 sub mediawiki_clean_filename {
 605         my $filename = shift;
 606         $filename =~ s{@{[SLASH_REPLACEMENT]}}{/}g;
 607         # [, ], |, {, and } are forbidden by MediaWiki, even URL-encoded.
 608         # Do a variant of URL-encoding, i.e. looks like URL-encoding,
 609         # but with _ added to prevent MediaWiki from thinking this is
 610         # an actual special character.
 611         $filename =~ s/[\[\]\{\}\|]/sprintf("_%%_%x", ord($&))/ge;
 612         # If we use the uri escape before
 613         # we should unescape here, before anything
 614
 615         return $filename;
 616 }
 617
 618 sub mediawiki_smudge_filename {
 619         my $filename = shift;
 620         $filename =~ s{/}{@{[SLASH_REPLACEMENT]}}g;
 621         $filename =~ s/ /_/g;
 622         # Decode forbidden characters encoded in mediawiki_clean_filename
 623         $filename =~ s/_%_([0-9a-fA-F][0-9a-fA-F])/sprintf('%c', hex($1))/ge;
 624         return $filename;
 625 }
 626
 627 sub literal_data {
 628         my ($content) = @_;
 629         print {*STDOUT} 'data ', bytes::length($content), "\n", $content;
 630         return;
 631 }
 632
 633 sub literal_data_raw {
 634         # Output possibly binary content.
 635         my ($content) = @_;
 636         # Avoid confusion between size in bytes and in characters
 637         utf8::downgrade($content);
 638         binmode STDOUT, ':raw';
 639         print {*STDOUT} 'data ', bytes::length($content), "\n", $content;
 640         binmode STDOUT, ':encoding(UTF-8)';
 641         return;
 642 }
 643
 644 sub mw_capabilities {
 645         # Revisions are imported to the private namespace
 646         # refs/mediawiki/$remotename/ by the helper and fetched into
 647         # refs/remotes/$remotename later by fetch.
 648         print {*STDOUT} "refspec refs/heads/*:refs/mediawiki/${remotename}/*\n";
 649         print {*STDOUT} "import\n";
 650         print {*STDOUT} "list\n";
 651         print {*STDOUT} "push\n";
 652         print {*STDOUT} "\n";
 653         return;
 654 }
 655
 656 sub mw_list {
 657         # MediaWiki do not have branches, we consider one branch arbitrarily
 658         # called master, and HEAD pointing to it.
 659         print {*STDOUT} "? refs/heads/master\n";
 660         print {*STDOUT} "\@refs/heads/master HEAD\n";
 661         print {*STDOUT} "\n";
 662         return;
 663 }
 664
 665 sub mw_option {
 666         print {*STDERR} "remote-helper command 'option $_[0]' not yet implemented\n";
 667         print {*STDOUT} "unsupported\n";
 668         return;
 669 }
 670
 671 sub fetch_mw_revisions_for_page {
 672         my $page = shift;
 673         my $id = shift;
 674         my $fetch_from = shift;
 675         my @page_revs = ();
 676         my $query = {
 677                 action => 'query',
 678                 prop => 'revisions',
 679                 rvprop => 'ids',
 680                 rvdir => 'newer',
 681                 rvstartid => $fetch_from,
 682                 rvlimit => 500,
 683                 pageids => $id,
 684         };
 685
 686         my $revnum = 0;
 687         # Get 500 revisions at a time due to the mediawiki api limit
 688         while (1) {
 689                 my $result = $mediawiki->api($query);
 690
 691                 # Parse each of those 500 revisions
 692                 foreach my $revision (@{$result->{query}->{pages}->{$id}->{revisions}}) {
 693                         my $page_rev_ids;
 694                         $page_rev_ids->{pageid} = $page->{pageid};
 695                         $page_rev_ids->{revid} = $revision->{revid};
 696                         push(@page_revs, $page_rev_ids);
 697                         $revnum++;
 698                 }
 699                 last if (!$result->{'query-continue'});
 700                 $query->{rvstartid} = $result->{'query-continue'}->{revisions}->{rvstartid};
 701         }
 702         if ($shallow_import && @page_revs) {
 703                 print {*STDERR} "  Found 1 revision (shallow import).\n";
 704                 @page_revs = sort {$b->{revid} <=> $a->{revid}} (@page_revs);
 705                 return $page_revs[0];
 706         }
 707         print {*STDERR} "  Found ${revnum} revision(s).\n";
 708         return @page_revs;
 709 }
 710
 711 sub fetch_mw_revisions {
 712         my $pages = shift; my @pages = @{$pages};
 713         my $fetch_from = shift;
 714
 715         my @revisions = ();
 716         my $n = 1;
 717         foreach my $page (@pages) {
 718                 my $id = $page->{pageid};
 719                 print {*STDERR} "page ${n}/", scalar(@pages), ': ', $page->{title}, "\n";
 720                 $n++;
 721                 my @page_revs = fetch_mw_revisions_for_page($page, $id, $fetch_from);
 722                 @revisions = (@page_revs, @revisions);
 723         }
 724
 725         return ($n, @revisions);
 726 }
 727
 728 sub fe_escape_path {
 729     my $path = shift;
 730     $path =~ s/\\/\\\\/g;
 731     $path =~ s/"/\\"/g;
 732     $path =~ s/\n/\\n/g;
 733     return qq("${path}");
 734 }
 735
 736 sub import_file_revision {
 737         my $commit = shift;
 738         my %commit = %{$commit};
 739         my $full_import = shift;
 740         my $n = shift;
 741         my $mediafile = shift;
 742         my %mediafile;
 743         if ($mediafile) {
 744                 %mediafile = %{$mediafile};
 745         }
 746
 747         my $title = $commit{title};
 748         my $comment = $commit{comment};
 749         my $content = $commit{content};
 750         my $author = $commit{author};
 751         my $date = $commit{date};
 752
 753         print {*STDOUT} "commit refs/mediawiki/${remotename}/master\n";
 754         print {*STDOUT} "mark :${n}\n";
 755         print {*STDOUT} "committer ${author} <${author}\@${wiki_name}> " . $date->epoch . " +0000\n";
 756         literal_data($comment);
 757
 758         # If it's not a clone, we need to know where to start from
 759         if (!$full_import && $n == 1) {
 760                 print {*STDOUT} "from refs/mediawiki/${remotename}/master^0\n";
 761         }
 762         if ($content ne DELETED_CONTENT) {
 763                 print {*STDOUT} 'M 644 inline ' .
 764                     fe_escape_path("${title}.mw") . "\n";
 765                 literal_data($content);
 766                 if (%mediafile) {
 767                         print {*STDOUT} 'M 644 inline '
 768                             . fe_escape_path($mediafile{title}) . "\n";
 769                         literal_data_raw($mediafile{content});
 770                 }
 771                 print {*STDOUT} "\n\n";
 772         } else {
 773                 print {*STDOUT} 'D ' . fe_escape_path("${title}.mw") . "\n";
 774         }
 775
 776         # mediawiki revision number in the git note
 777         if ($full_import && $n == 1) {
 778                 print {*STDOUT} "reset refs/notes/${remotename}/mediawiki\n";
 779         }
 780         print {*STDOUT} "commit refs/notes/${remotename}/mediawiki\n";
 781         print {*STDOUT} "committer ${author} <${author}\@${wiki_name}> " . $date->epoch . " +0000\n";
 782         literal_data('Note added by git-mediawiki during import');
 783         if (!$full_import && $n == 1) {
 784                 print {*STDOUT} "from refs/notes/${remotename}/mediawiki^0\n";
 785         }
 786         print {*STDOUT} "N inline :${n}\n";
 787         literal_data("mediawiki_revision: $commit{mw_revision}");
 788         print {*STDOUT} "\n\n";
 789         return;
 790 }
 791
 792 # parse a sequence of
 793 # <cmd> <arg1>
 794 # <cmd> <arg2>
 795 # \n
 796 # (like batch sequence of import and sequence of push statements)
 797 sub get_more_refs {
 798         my $cmd = shift;
 799         my @refs;
 800         while (1) {
 801                 my $line = <STDIN>;
 802                 if ($line =~ /^$cmd (.*)$/) {
 803                         push(@refs, $1);
 804                 } elsif ($line eq "\n") {
 805                         return @refs;
 806                 } else {
 807                         die("Invalid command in a '$cmd' batch: $_\n");
 808                 }
 809         }
 810         return;
 811 }
 812
 813 sub mw_import {
 814         # multiple import commands can follow each other.
 815         my @refs = (shift, get_more_refs('import'));
 816         foreach my $ref (@refs) {
 817                 mw_import_ref($ref);
 818         }
 819         print {*STDOUT} "done\n";
 820         return;
 821 }
 822
 823 sub mw_import_ref {
 824         my $ref = shift;
 825         # The remote helper will call "import HEAD" and
 826         # "import refs/heads/master".
 827         # Since HEAD is a symbolic ref to master (by convention,
 828         # followed by the output of the command "list" that we gave),
 829         # we don't need to do anything in this case.
 830         if ($ref eq 'HEAD') {
 831                 return;
 832         }
 833
 834         mw_connect_maybe();
 835
 836         print {*STDERR} "Searching revisions...\n";
 837         my $last_local = get_last_local_revision();
 838         my $fetch_from = $last_local + 1;
 839         if ($fetch_from == 1) {
 840                 print {*STDERR} ", fetching from beginning.\n";
 841         } else {
 842                 print {*STDERR} ", fetching from here.\n";
 843         }
 844
 845         my $n = 0;
 846         if ($fetch_strategy eq 'by_rev') {
 847                 print {*STDERR} "Fetching & writing export data by revs...\n";
 848                 $n = mw_import_ref_by_revs($fetch_from);
 849         } elsif ($fetch_strategy eq 'by_page') {
 850                 print {*STDERR} "Fetching & writing export data by pages...\n";
 851                 $n = mw_import_ref_by_pages($fetch_from);
 852         } else {
 853                 print {*STDERR} qq(fatal: invalid fetch strategy "${fetch_strategy}".\n);
 854                 print {*STDERR} "Check your configuration variables remote.${remotename}.fetchStrategy and mediawiki.fetchStrategy\n";
 855                 exit 1;
 856         }
 857
 858         if ($fetch_from == 1 && $n == 0) {
 859                 print {*STDERR} "You appear to have cloned an empty MediaWiki.\n";
 860                 # Something has to be done remote-helper side. If nothing is done, an error is
 861                 # thrown saying that HEAD is referring to unknown object 0000000000000000000
 862                 # and the clone fails.
 863         }
 864         return;
 865 }
 866
 867 sub mw_import_ref_by_pages {
 868
 869         my $fetch_from = shift;
 870         my %pages_hash = get_mw_pages();
 871         my @pages = values(%pages_hash);
 872
 873         my ($n, @revisions) = fetch_mw_revisions(\@pages, $fetch_from);
 874
 875         @revisions = sort {$a->{revid} <=> $b->{revid}} @revisions;
 876         my @revision_ids = map { $_->{revid} } @revisions;
 877
 878         return mw_import_revids($fetch_from, \@revision_ids, \%pages_hash);
 879 }
 880
 881 sub mw_import_ref_by_revs {
 882
 883         my $fetch_from = shift;
 884         my %pages_hash = get_mw_pages();
 885
 886         my $last_remote = get_last_global_remote_rev();
 887         my @revision_ids = $fetch_from..$last_remote;
 888         return mw_import_revids($fetch_from, \@revision_ids, \%pages_hash);
 889 }
 890
 891 # Import revisions given in second argument (array of integers).
 892 # Only pages appearing in the third argument (hash indexed by page titles)
 893 # will be imported.
 894 sub mw_import_revids {
 895         my $fetch_from = shift;
 896         my $revision_ids = shift;
 897         my $pages = shift;
 898
 899         my $n = 0;
 900         my $n_actual = 0;
 901         my $last_timestamp = 0; # Placeholer in case $rev->timestamp is undefined
 902
 903         foreach my $pagerevid (@{$revision_ids}) {
 904                 # Count page even if we skip it, since we display
 905                 # $n/$total and $total includes skipped pages.
 906                 $n++;
 907
 908                 # fetch the content of the pages
 909                 my $query = {
 910                         action => 'query',
 911                         prop => 'revisions',
 912                         rvprop => 'content|timestamp|comment|user|ids',
 913                         revids => $pagerevid,
 914                 };
 915
 916                 my $result = $mediawiki->api($query);
 917
 918                 if (!$result) {
 919                         die "Failed to retrieve modified page for revision $pagerevid\n";
 920                 }
 921
 922                 if (defined($result->{query}->{badrevids}->{$pagerevid})) {
 923                         # The revision id does not exist on the remote wiki.
 924                         next;
 925                 }
 926
 927                 if (!defined($result->{query}->{pages})) {
 928                         die "Invalid revision ${pagerevid}.\n";
 929                 }
 930
 931                 my @result_pages = values(%{$result->{query}->{pages}});
 932                 my $result_page = $result_pages[0];
 933                 my $rev = $result_pages[0]->{revisions}->[0];
 934
 935                 my $page_title = $result_page->{title};
 936
 937                 if (!exists($pages->{$page_title})) {
 938                         print {*STDERR} "${n}/", scalar(@{$revision_ids}),
 939                                 ": Skipping revision #$rev->{revid} of ${page_title}\n";
 940                         next;
 941                 }
 942
 943                 $n_actual++;
 944
 945                 my %commit;
 946                 $commit{author} = $rev->{user} || 'Anonymous';
 947                 $commit{comment} = $rev->{comment} || EMPTY_MESSAGE;
 948                 $commit{title} = mediawiki_smudge_filename($page_title);
 949                 $commit{mw_revision} = $rev->{revid};
 950                 $commit{content} = mediawiki_smudge($rev->{'*'});
 951
 952                 if (!defined($rev->{timestamp})) {
 953                         $last_timestamp++;
 954                 } else {
 955                         $last_timestamp = $rev->{timestamp};
 956                 }
 957                 $commit{date} = DateTime::Format::ISO8601->parse_datetime($last_timestamp);
 958
 959                 # Differentiates classic pages and media files.
 960                 my ($namespace, $filename) = $page_title =~ /^([^:]*):(.*)$/;
 961                 my %mediafile;
 962                 if ($namespace) {
 963                         my $id = get_mw_namespace_id($namespace);
 964                         if ($id && $id == get_mw_namespace_id('File')) {
 965                                 %mediafile = get_mw_mediafile_for_page_revision($filename, $rev->{timestamp});
 966                         }
 967                 }
 968                 # If this is a revision of the media page for new version
 969                 # of a file do one common commit for both file and media page.
 970                 # Else do commit only for that page.
 971                 print {*STDERR} "${n}/", scalar(@{$revision_ids}), ": Revision #$rev->{revid} of $commit{title}\n";
 972                 import_file_revision(\%commit, ($fetch_from == 1), $n_actual, \%mediafile);
 973         }
 974
 975         return $n_actual;
 976 }
 977
 978 sub error_non_fast_forward {
 979         my $advice = run_git('config --bool advice.pushNonFastForward');
 980         chomp($advice);
 981         if ($advice ne 'false') {
 982                 # Native git-push would show this after the summary.
 983                 # We can't ask it to display it cleanly, so print it
 984                 # ourselves before.
 985                 print {*STDERR} "To prevent you from losing history, non-fast-forward updates were rejected\n";
 986                 print {*STDERR} "Merge the remote changes (e.g. 'git pull') before pushing again. See the\n";
 987                 print {*STDERR} "'Note about fast-forwards' section of 'git push --help' for details.\n";
 988         }
 989         print {*STDOUT} qq(error $_[0] "non-fast-forward"\n);
 990         return 0;
 991 }
 992
 993 sub mw_upload_file {
 994         my $complete_file_name = shift;
 995         my $new_sha1 = shift;
 996         my $extension = shift;
 997         my $file_deleted = shift;
 998         my $summary = shift;
 999         my $newrevid;
1000         my $path = "File:${complete_file_name}";
1001         my %hashFiles = get_allowed_file_extensions();
1002         if (!exists($hashFiles{$extension})) {
1003                 print {*STDERR} "${complete_file_name} is not a permitted file on this wiki.\n";
1004                 print {*STDERR} "Check the configuration of file uploads in your mediawiki.\n";
1005                 return $newrevid;
1006         }
1007         # Deleting and uploading a file requires a priviledged user
1008         if ($file_deleted) {
1009                 mw_connect_maybe();
1010                 my $query = {
1011                         action => 'delete',
1012                         title => $path,
1013                         reason => $summary
1014                 };
1015                 if (!$mediawiki->edit($query)) {
1016                         print {*STDERR} "Failed to delete file on remote wiki\n";
1017                         print {*STDERR} "Check your permissions on the remote site. Error code:\n";
1018                         print {*STDERR} $mediawiki->{error}->{code} . ':' . $mediawiki->{error}->{details};
1019                         exit 1;
1020                 }
1021         } else {
1022                 # Don't let perl try to interpret file content as UTF-8 => use "raw"
1023                 my $content = run_git("cat-file blob ${new_sha1}", 'raw');
1024                 if ($content ne EMPTY) {
1025                         mw_connect_maybe();
1026                         $mediawiki->{config}->{upload_url} =
1027                                 "${url}/index.php/Special:Upload";
1028                         $mediawiki->edit({
1029                                 action => 'upload',
1030                                 filename => $complete_file_name,
1031                                 comment => $summary,
1032                                 file => [undef,
1033                                          $complete_file_name,
1034                                          Content => $content],
1035                                 ignorewarnings => 1,
1036                         }, {
1037                                 skip_encoding => 1
1038                         } ) || die $mediawiki->{error}->{code} . ':'
1039                                  . $mediawiki->{error}->{details} . "\n";
1040                         my $last_file_page = $mediawiki->get_page({title => $path});
1041                         $newrevid = $last_file_page->{revid};
1042                         print {*STDERR} "Pushed file: ${new_sha1} - ${complete_file_name}.\n";
1043                 } else {
1044                         print {*STDERR} "Empty file ${complete_file_name} not pushed.\n";
1045                 }
1046         }
1047         return $newrevid;
1048 }
1049
1050 sub mw_push_file {
1051         my $diff_info = shift;
1052         # $diff_info contains a string in this format:
1053         # 100644 100644 <sha1_of_blob_before_commit> <sha1_of_blob_now> <status>
1054         my @diff_info_split = split(/[ \t]/, $diff_info);
1055
1056         # Filename, including .mw extension
1057         my $complete_file_name = shift;
1058         # Commit message
1059         my $summary = shift;
1060         # MediaWiki revision number. Keep the previous one by default,
1061         # in case there's no edit to perform.
1062         my $oldrevid = shift;
1063         my $newrevid;
1064
1065         if ($summary eq EMPTY_MESSAGE) {
1066                 $summary = EMPTY;
1067         }
1068
1069         my $new_sha1 = $diff_info_split[3];
1070         my $old_sha1 = $diff_info_split[2];
1071         my $page_created = ($old_sha1 eq NULL_SHA1);
1072         my $page_deleted = ($new_sha1 eq NULL_SHA1);
1073         $complete_file_name = mediawiki_clean_filename($complete_file_name);
1074
1075         my ($title, $extension) = $complete_file_name =~ /^(.*)\.([^\.]*)$/;
1076         if (!defined($extension)) {
1077                 $extension = EMPTY;
1078         }
1079         if ($extension eq 'mw') {
1080                 my $ns = get_mw_namespace_id_for_page($complete_file_name);
1081                 if ($ns && $ns == get_mw_namespace_id('File') && (!$export_media)) {
1082                         print {*STDERR} "Ignoring media file related page: ${complete_file_name}\n";
1083                         return ($oldrevid, 'ok');
1084                 }
1085                 my $file_content;
1086                 if ($page_deleted) {
1087                         # Deleting a page usually requires
1088                         # special privileges. A common
1089                         # convention is to replace the page
1090                         # with this content instead:
1091                         $file_content = DELETED_CONTENT;
1092                 } else {
1093                         $file_content = run_git("cat-file blob ${new_sha1}");
1094                 }
1095
1096                 mw_connect_maybe();
1097
1098                 my $result = $mediawiki->edit( {
1099                         action => 'edit',
1100                         summary => $summary,
1101                         title => $title,
1102                         basetimestamp => $basetimestamps{$oldrevid},
1103                         text => mediawiki_clean($file_content, $page_created),
1104                                   }, {
1105                                           skip_encoding => 1 # Helps with names with accentuated characters
1106                                   });
1107                 if (!$result) {
1108                         if ($mediawiki->{error}->{code} == 3) {
1109                                 # edit conflicts, considered as non-fast-forward
1110                                 print {*STDERR} 'Warning: Error ' .
1111                                     $mediawiki->{error}->{code} .
1112                                     ' from mediawiki: ' . $mediawiki->{error}->{details} .
1113                                     ".\n";
1114                                 return ($oldrevid, 'non-fast-forward');
1115                         } else {
1116                                 # Other errors. Shouldn't happen => just die()
1117                                 die 'Fatal: Error ' .
1118                                     $mediawiki->{error}->{code} .
1119                                     ' from mediawiki: ' . $mediawiki->{error}->{details} . "\n";
1120                         }
1121                 }
1122                 $newrevid = $result->{edit}->{newrevid};
1123                 print {*STDERR} "Pushed file: ${new_sha1} - ${title}\n";
1124         } elsif ($export_media) {
1125                 $newrevid = mw_upload_file($complete_file_name, $new_sha1,
1126                                            $extension, $page_deleted,
1127                                            $summary);
1128         } else {
1129                 print {*STDERR} "Ignoring media file ${title}\n";
1130         }
1131         $newrevid = ($newrevid or $oldrevid);
1132         return ($newrevid, 'ok');
1133 }
1134
1135 sub mw_push {
1136         # multiple push statements can follow each other
1137         my @refsspecs = (shift, get_more_refs('push'));
1138         my $pushed;
1139         for my $refspec (@refsspecs) {
1140                 my ($force, $local, $remote) = $refspec =~ /^(\+)?([^:]*):([^:]*)$/
1141                     or die("Invalid refspec for push. Expected <src>:<dst> or +<src>:<dst>\n");
1142                 if ($force) {
1143                         print {*STDERR} "Warning: forced push not allowed on a MediaWiki.\n";
1144                 }
1145                 if ($local eq EMPTY) {
1146                         print {*STDERR} "Cannot delete remote branch on a MediaWiki\n";
1147                         print {*STDOUT} "error ${remote} cannot delete\n";
1148                         next;
1149                 }
1150                 if ($remote ne 'refs/heads/master') {
1151                         print {*STDERR} "Only push to the branch 'master' is supported on a MediaWiki\n";
1152                         print {*STDOUT} "error ${remote} only master allowed\n";
1153                         next;
1154                 }
1155                 if (mw_push_revision($local, $remote)) {
1156                         $pushed = 1;
1157                 }
1158         }
1159
1160         # Notify Git that the push is done
1161         print {*STDOUT} "\n";
1162
1163         if ($pushed && $dumb_push) {
1164                 print {*STDERR} "Just pushed some revisions to MediaWiki.\n";
1165                 print {*STDERR} "The pushed revisions now have to be re-imported, and your current branch\n";
1166                 print {*STDERR} "needs to be updated with these re-imported commits. You can do this with\n";
1167                 print {*STDERR} "\n";
1168                 print {*STDERR} "  git pull --rebase\n";
1169                 print {*STDERR} "\n";
1170         }
1171         return;
1172 }
1173
1174 sub mw_push_revision {
1175         my $local = shift;
1176         my $remote = shift; # actually, this has to be "refs/heads/master" at this point.
1177         my $last_local_revid = get_last_local_revision();
1178         print {*STDERR} ".\n"; # Finish sentence started by get_last_local_revision()
1179         my $last_remote_revid = get_last_remote_revision();
1180         my $mw_revision = $last_remote_revid;
1181
1182         # Get sha1 of commit pointed by local HEAD
1183         my $HEAD_sha1 = run_git("rev-parse ${local} 2>/dev/null");
1184         chomp($HEAD_sha1);
1185         # Get sha1 of commit pointed by remotes/$remotename/master
1186         my $remoteorigin_sha1 = run_git("rev-parse refs/remotes/${remotename}/master 2>/dev/null");
1187         chomp($remoteorigin_sha1);
1188
1189         if ($last_local_revid > 0 &&
1190             $last_local_revid < $last_remote_revid) {
1191                 return error_non_fast_forward($remote);
1192         }
1193
1194         if ($HEAD_sha1 eq $remoteorigin_sha1) {
1195                 # nothing to push
1196                 return 0;
1197         }
1198
1199         # Get every commit in between HEAD and refs/remotes/origin/master,
1200         # including HEAD and refs/remotes/origin/master
1201         my @commit_pairs = ();
1202         if ($last_local_revid > 0) {
1203                 my $parsed_sha1 = $remoteorigin_sha1;
1204                 # Find a path from last MediaWiki commit to pushed commit
1205                 print {*STDERR} "Computing path from local to remote ...\n";
1206                 my @local_ancestry = split(/\n/, run_git("rev-list --boundary --parents ${local} ^${parsed_sha1}"));
1207                 my %local_ancestry;
1208                 foreach my $line (@local_ancestry) {
1209                         if (my ($child, $parents) = $line =~ /^-?([a-f0-9]+) ([a-f0-9 ]+)/) {
1210                                 foreach my $parent (split(/ /, $parents)) {
1211                                         $local_ancestry{$parent} = $child;
1212                                 }
1213                         } elsif (!$line =~ /^([a-f0-9]+)/) {
1214                                 die "Unexpected output from git rev-list: ${line}\n";
1215                         }
1216                 }
1217                 while ($parsed_sha1 ne $HEAD_sha1) {
1218                         my $child = $local_ancestry{$parsed_sha1};
1219                         if (!$child) {
1220                                 print {*STDERR} "Cannot find a path in history from remote commit to last commit\n";
1221                                 return error_non_fast_forward($remote);
1222                         }
1223                         push(@commit_pairs, [$parsed_sha1, $child]);
1224                         $parsed_sha1 = $child;
1225                 }
1226         } else {
1227                 # No remote mediawiki revision. Export the whole
1228                 # history (linearized with --first-parent)
1229                 print {*STDERR} "Warning: no common ancestor, pushing complete history\n";
1230                 my $history = run_git("rev-list --first-parent --children ${local}");
1231                 my @history = split(/\n/, $history);
1232                 @history = @history[1..$#history];
1233                 foreach my $line (reverse @history) {
1234                         my @commit_info_split = split(/[ \n]/, $line);
1235                         push(@commit_pairs, \@commit_info_split);
1236                 }
1237         }
1238
1239         foreach my $commit_info_split (@commit_pairs) {
1240                 my $sha1_child = @{$commit_info_split}[0];
1241                 my $sha1_commit = @{$commit_info_split}[1];
1242                 my $diff_infos = run_git("diff-tree -r --raw -z ${sha1_child} ${sha1_commit}");
1243                 # TODO: we could detect rename, and encode them with a #redirect on the wiki.
1244                 # TODO: for now, it's just a delete+add
1245                 my @diff_info_list = split(/\0/, $diff_infos);
1246                 # Keep the subject line of the commit message as mediawiki comment for the revision
1247                 my $commit_msg = run_git(qq(log --no-walk --format="%s" ${sha1_commit}));
1248                 chomp($commit_msg);
1249                 # Push every blob
1250                 while (@diff_info_list) {
1251                         my $status;
1252                         # git diff-tree -z gives an output like
1253                         # <metadata>\0<filename1>\0
1254                         # <metadata>\0<filename2>\0
1255                         # and we've split on \0.
1256                         my $info = shift(@diff_info_list);
1257                         my $file = shift(@diff_info_list);
1258                         ($mw_revision, $status) = mw_push_file($info, $file, $commit_msg, $mw_revision);
1259                         if ($status eq 'non-fast-forward') {
1260                                 # we may already have sent part of the
1261                                 # commit to MediaWiki, but it's too
1262                                 # late to cancel it. Stop the push in
1263                                 # the middle, but still give an
1264                                 # accurate error message.
1265                                 return error_non_fast_forward($remote);
1266                         }
1267                         if ($status ne 'ok') {
1268                                 die("Unknown error from mw_push_file()\n");
1269                         }
1270                 }
1271                 if (!$dumb_push) {
1272                         run_git(qq(notes --ref=${remotename}/mediawiki add -f -m "mediawiki_revision: ${mw_revision}" ${sha1_commit}));
1273                         run_git(qq(update-ref -m "Git-MediaWiki push" refs/mediawiki/${remotename}/master ${sha1_commit} ${sha1_child}));
1274                 }
1275         }
1276
1277         print {*STDOUT} "ok ${remote}\n";
1278         return 1;
1279 }
1280
1281 sub get_allowed_file_extensions {
1282         mw_connect_maybe();
1283
1284         my $query = {
1285                 action => 'query',
1286                 meta => 'siteinfo',
1287                 siprop => 'fileextensions'
1288                 };
1289         my $result = $mediawiki->api($query);
1290         my @file_extensions = map { $_->{ext}} @{$result->{query}->{fileextensions}};
1291         my %hashFile = map { $_ => 1 } @file_extensions;
1292
1293         return %hashFile;
1294 }
1295
1296 # In memory cache for MediaWiki namespace ids.
1297 my %namespace_id;
1298
1299 # Namespaces whose id is cached in the configuration file
1300 # (to avoid duplicates)
1301 my %cached_mw_namespace_id;
1302
1303 # Return MediaWiki id for a canonical namespace name.
1304 # Ex.: "File", "Project".
1305 sub get_mw_namespace_id {
1306         mw_connect_maybe();
1307         my $name = shift;
1308
1309         if (!exists $namespace_id{$name}) {
1310                 # Look at configuration file, if the record for that namespace is
1311                 # already cached. Namespaces are stored in form:
1312                 # "Name_of_namespace:Id_namespace", ex.: "File:6".
1313                 my @temp = split(/\n/,
1314                                  run_git("config --get-all remote.${remotename}.namespaceCache"));
1315                 chomp(@temp);
1316                 foreach my $ns (@temp) {
1317                         my ($n, $id) = split(/:/, $ns);
1318                         if ($id eq 'notANameSpace') {
1319                                 $namespace_id{$n} = {is_namespace => 0};
1320                         } else {
1321                                 $namespace_id{$n} = {is_namespace => 1, id => $id};
1322                         }
1323                         $cached_mw_namespace_id{$n} = 1;
1324                 }
1325         }
1326
1327         if (!exists $namespace_id{$name}) {
1328                 print {*STDERR} "Namespace ${name} not found in cache, querying the wiki ...\n";
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}})) {
1339                         if (defined($ns->{id}) && defined($ns->{canonical})) {
1340                                 $namespace_id{$ns->{canonical}} = {is_namespace => 1, id => $ns->{id}};
1341                                 if ($ns->{'*'}) {
1342                                         # alias (e.g. french Fichier: as alias for canonical File:)
1343                                         $namespace_id{$ns->{'*'}} = {is_namespace => 1, id => $ns->{id}};
1344                                 }
1345                         }
1346                 }
1347         }
1348
1349         my $ns = $namespace_id{$name};
1350         my $id;
1351
1352         if (!defined $ns) {
1353                 print {*STDERR} "No such namespace ${name} on MediaWiki.\n";
1354                 $ns = {is_namespace => 0};
1355                 $namespace_id{$name} = $ns;
1356         }
1357
1358         if ($ns->{is_namespace}) {
1359                 $id = $ns->{id};
1360         }
1361
1362         # Store "notANameSpace" as special value for inexisting namespaces
1363         my $store_id = ($id || 'notANameSpace');
1364
1365         # Store explicitely requested namespaces on disk
1366         if (!exists $cached_mw_namespace_id{$name}) {
1367                 run_git(qq(config --add remote.${remotename}.namespaceCache "${name}:${store_id}"));
1368                 $cached_mw_namespace_id{$name} = 1;
1369         }
1370         return $id;
1371 }
1372
1373 sub get_mw_namespace_id_for_page {
1374         my $namespace = shift;
1375         if ($namespace =~ /^([^:]*):/) {
1376                 return get_mw_namespace_id($namespace);
1377         } else {
1378                 return;
1379         }
1380 }