Andrew's git / gitweb.git /
e3a79db4258fa75957a3ad52496d80304ed10a1e
   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
  14use strict;
  15use MediaWiki::API;
  16use Git;
  17use DateTime::Format::ISO8601;
  18use warnings;
  19
  20# By default, use UTF-8 to communicate with Git and the user
  21binmode STDERR, ":encoding(UTF-8)";
  22binmode STDOUT, ":encoding(UTF-8)";
  23
  24use URI::Escape;
  25
  26# Mediawiki filenames can contain forward slashes. This variable decides by which pattern they should be replaced
  27use 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.
  31use 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.
  35use constant EMPTY_CONTENT => "<!-- empty page -->\n";
  36
  37# used to reflect file creation or deletion in diff.
  38use constant NULL_SHA1 => "0000000000000000000000000000000000000000";
  39
  40# Used on Git's side to reflect empty edit messages on the wiki
  41use constant EMPTY_MESSAGE => '*Empty MediaWiki Message*';
  42
  43my $remotename = $ARGV[0];
  44my $url = $ARGV[1];
  45
  46# Accept both space-separated and multiple keys in config file.
  47# Spaces should be written as _ anyway because we'll use chomp.
  48my @tracked_pages = split(/[ \n]/, run_git("config --get-all remote.". $remotename .".pages"));
  49chomp(@tracked_pages);
  50
  51# Just like @tracked_pages, but for MediaWiki categories.
  52my @tracked_categories = split(/[ \n]/, run_git("config --get-all remote.". $remotename .".categories"));
  53chomp(@tracked_categories);
  54
  55# Import media files on pull
  56my $import_media = run_git("config --get --bool remote.". $remotename .".mediaimport");
  57chomp($import_media);
  58$import_media = ($import_media eq "true");
  59
  60# Export media files on push
  61my $export_media = run_git("config --get --bool remote.". $remotename .".mediaexport");
  62chomp($export_media);
  63$export_media = !($export_media eq "false");
  64
  65my $wiki_login = run_git("config --get remote.". $remotename .".mwLogin");
  66# Note: mwPassword is discourraged. Use the credential system instead.
  67my $wiki_passwd = run_git("config --get remote.". $remotename .".mwPassword");
  68my $wiki_domain = run_git("config --get remote.". $remotename .".mwDomain");
  69chomp($wiki_login);
  70chomp($wiki_passwd);
  71chomp($wiki_domain);
  72
  73# Import only last revisions (both for clone and fetch)
  74my $shallow_import = run_git("config --get --bool remote.". $remotename .".shallow");
  75chomp($shallow_import);
  76$shallow_import = ($shallow_import eq "true");
  77
  78# Fetch (clone and pull) by revisions instead of by pages. This behavior
  79# is more efficient when we have a wiki with lots of pages and we fetch
  80# the revisions quite often so that they concern only few pages.
  81# Possible values:
  82# - by_rev: perform one query per new revision on the remote wiki
  83# - by_page: query each tracked page for new revision
  84my $fetch_strategy = run_git("config --get remote.$remotename.fetchStrategy");
  85unless ($fetch_strategy) {
  86        $fetch_strategy = run_git("config --get mediawiki.fetchStrategy");
  87}
  88chomp($fetch_strategy);
  89unless ($fetch_strategy) {
  90        $fetch_strategy = "by_page";
  91}
  92
  93# Remember the timestamp corresponding to a revision id.
  94my %basetimestamps;
  95
  96# Dumb push: don't update notes and mediawiki ref to reflect the last push.
  97#
  98# Configurable with mediawiki.dumbPush, or per-remote with
  99# remote.<remotename>.dumbPush.
 100#
 101# This means the user will have to re-import the just-pushed
 102# revisions. On the other hand, this means that the Git revisions
 103# corresponding to MediaWiki revisions are all imported from the wiki,
 104# regardless of whether they were initially created in Git or from the
 105# web interface, hence all users will get the same history (i.e. if
 106# the push from Git to MediaWiki loses some information, everybody
 107# will get the history with information lost). If the import is
 108# deterministic, this means everybody gets the same sha1 for each
 109# MediaWiki revision.
 110my $dumb_push = run_git("config --get --bool remote.$remotename.dumbPush");
 111unless ($dumb_push) {
 112        $dumb_push = run_git("config --get --bool mediawiki.dumbPush");
 113}
 114chomp($dumb_push);
 115$dumb_push = ($dumb_push eq "true");
 116
 117my $wiki_name = $url;
 118$wiki_name =~ s{[^/]*://}{};
 119# If URL is like http://user:password@example.com/, we clearly don't
 120# want the password in $wiki_name. While we're there, also remove user
 121# and '@' sign, to avoid author like MWUser@HTTPUser@host.com
 122$wiki_name =~ s/^.*@//;
 123
 124# Commands parser
 125my @cmd;
 126while (<STDIN>) {
 127        chomp;
 128        @cmd = split(/ /);
 129        if (defined($cmd[0])) {
 130                # Line not blank
 131                if ($cmd[0] eq "capabilities") {
 132                        die("Too many arguments for capabilities\n") if (defined($cmd[1]));
 133                        mw_capabilities();
 134                } elsif ($cmd[0] eq "list") {
 135                        die("Too many arguments for list\n") if (defined($cmd[2]));
 136                        mw_list($cmd[1]);
 137                } elsif ($cmd[0] eq "import") {
 138                        die("Invalid arguments for import\n") if ($cmd[1] eq "" || defined($cmd[2]));
 139                        mw_import($cmd[1]);
 140                } elsif ($cmd[0] eq "option") {
 141                        die("Too many arguments for option\n") if ($cmd[1] eq "" || $cmd[2] eq "" || defined($cmd[3]));
 142                        mw_option($cmd[1],$cmd[2]);
 143                } elsif ($cmd[0] eq "push") {
 144                        mw_push($cmd[1]);
 145                } else {
 146                        print STDERR "Unknown command. Aborting...\n";
 147                        last;
 148                }
 149        } else {
 150                # blank line: we should terminate
 151                last;
 152        }
 153
 154        BEGIN { $| = 1 } # flush STDOUT, to make sure the previous
 155                         # command is fully processed.
 156}
 157
 158########################## Functions ##############################
 159
 160# MediaWiki API instance, created lazily.
 161my $mediawiki;
 162
 163sub mw_connect_maybe {
 164        if ($mediawiki) {
 165                return;
 166        }
 167        $mediawiki = MediaWiki::API->new;
 168        $mediawiki->{config}->{api_url} = "$url/api.php";
 169        if ($wiki_login) {
 170                my %credential = (
 171                        'url' => $url,
 172                        'username' => $wiki_login,
 173                        'password' => $wiki_passwd
 174                );
 175                Git::credential(\%credential);
 176                my $request = {lgname => $credential{username},
 177                               lgpassword => $credential{password},
 178                               lgdomain => $wiki_domain};
 179                if ($mediawiki->login($request)) {
 180                        Git::credential(\%credential, 'approve');
 181                        print STDERR "Logged in mediawiki user \"$credential{username}\".\n";
 182                } else {
 183                        print STDERR "Failed to log in mediawiki user \"$credential{username}\" on $url\n";
 184                        print STDERR "  (error " .
 185                                $mediawiki->{error}->{code} . ': ' .
 186                                $mediawiki->{error}->{details} . ")\n";
 187                        Git::credential(\%credential, 'reject');
 188                        exit 1;
 189                }
 190        }
 191        return;
 192}
 193
 194sub fatal_mw_error {
 195        my $action = shift;
 196        print STDERR "fatal: could not $action.\n";
 197        print STDERR "fatal: '$url' does not appear to be a mediawiki\n";
 198        if ($url =~ /^https/) {
 199                print STDERR "fatal: make sure '$url/api.php' is a valid page\n";
 200                print STDERR "fatal: and the SSL certificate is correct.\n";
 201        } else {
 202                print STDERR "fatal: make sure '$url/api.php' is a valid page.\n";
 203        }
 204        print STDERR "fatal: (error " .
 205            $mediawiki->{error}->{code} . ': ' .
 206            $mediawiki->{error}->{details} . ")\n";
 207        exit 1;
 208}
 209
 210## Functions for listing pages on the remote wiki
 211sub get_mw_tracked_pages {
 212        my $pages = shift;
 213        get_mw_page_list(\@tracked_pages, $pages);
 214        return;
 215}
 216
 217sub get_mw_page_list {
 218        my $page_list = shift;
 219        my $pages = shift;
 220        my @some_pages = @$page_list;
 221        while (@some_pages) {
 222                my $last_page = 50;
 223                if ($#some_pages < $last_page) {
 224                        $last_page = $#some_pages;
 225                }
 226                my @slice = @some_pages[0..$last_page];
 227                get_mw_first_pages(\@slice, $pages);
 228                @some_pages = @some_pages[51..$#some_pages];
 229        }
 230        return;
 231}
 232
 233sub get_mw_tracked_categories {
 234        my $pages = shift;
 235        foreach my $category (@tracked_categories) {
 236                if (index($category, ':') < 0) {
 237                        # Mediawiki requires the Category
 238                        # prefix, but let's not force the user
 239                        # to specify it.
 240                        $category = "Category:" . $category;
 241                }
 242                my $mw_pages = $mediawiki->list( {
 243                        action => 'query',
 244                        list => 'categorymembers',
 245                        cmtitle => $category,
 246                        cmlimit => 'max' } )
 247                        || die $mediawiki->{error}->{code} . ': '
 248                                . $mediawiki->{error}->{details} . "\n";
 249                foreach my $page (@{$mw_pages}) {
 250                        $pages->{$page->{title}} = $page;
 251                }
 252        }
 253        return;
 254}
 255
 256sub get_mw_all_pages {
 257        my $pages = shift;
 258        # No user-provided list, get the list of pages from the API.
 259        my $mw_pages = $mediawiki->list({
 260                action => 'query',
 261                list => 'allpages',
 262                aplimit => 'max'
 263        });
 264        if (!defined($mw_pages)) {
 265                fatal_mw_error("get the list of wiki pages");
 266        }
 267        foreach my $page (@{$mw_pages}) {
 268                $pages->{$page->{title}} = $page;
 269        }
 270        return;
 271}
 272
 273# queries the wiki for a set of pages. Meant to be used within a loop
 274# querying the wiki for slices of page list.
 275sub get_mw_first_pages {
 276        my $some_pages = shift;
 277        my @some_pages = @{$some_pages};
 278
 279        my $pages = shift;
 280
 281        # pattern 'page1|page2|...' required by the API
 282        my $titles = join('|', @some_pages);
 283
 284        my $mw_pages = $mediawiki->api({
 285                action => 'query',
 286                titles => $titles,
 287        });
 288        if (!defined($mw_pages)) {
 289                fatal_mw_error("query the list of wiki pages");
 290        }
 291        while (my ($id, $page) = each(%{$mw_pages->{query}->{pages}})) {
 292                if ($id < 0) {
 293                        print STDERR "Warning: page $page->{title} not found on wiki\n";
 294                } else {
 295                        $pages->{$page->{title}} = $page;
 296                }
 297        }
 298        return;
 299}
 300
 301# Get the list of pages to be fetched according to configuration.
 302sub get_mw_pages {
 303        mw_connect_maybe();
 304
 305        print STDERR "Listing pages on remote wiki...\n";
 306
 307        my %pages; # hash on page titles to avoid duplicates
 308        my $user_defined;
 309        if (@tracked_pages) {
 310                $user_defined = 1;
 311                # The user provided a list of pages titles, but we
 312                # still need to query the API to get the page IDs.
 313                get_mw_tracked_pages(\%pages);
 314        }
 315        if (@tracked_categories) {
 316                $user_defined = 1;
 317                get_mw_tracked_categories(\%pages);
 318        }
 319        if (!$user_defined) {
 320                get_mw_all_pages(\%pages);
 321        }
 322        if ($import_media) {
 323                print STDERR "Getting media files for selected pages...\n";
 324                if ($user_defined) {
 325                        get_linked_mediafiles(\%pages);
 326                } else {
 327                        get_all_mediafiles(\%pages);
 328                }
 329        }
 330        print STDERR (scalar keys %pages) . " pages found.\n";
 331        return %pages;
 332}
 333
 334# usage: $out = run_git("command args");
 335#        $out = run_git("command args", "raw"); # don't interpret output as UTF-8.
 336sub run_git {
 337        my $args = shift;
 338        my $encoding = (shift || "encoding(UTF-8)");
 339        open(my $git, "-|:$encoding", "git " . $args)
 340            or die "Unable to open: $!\n";
 341        my $res = do {
 342                local $/ = undef;
 343                <$git>
 344        };
 345        close($git);
 346
 347        return $res;
 348}
 349
 350
 351sub get_all_mediafiles {
 352        my $pages = shift;
 353        # Attach list of all pages for media files from the API,
 354        # they are in a different namespace, only one namespace
 355        # can be queried at the same moment
 356        my $mw_pages = $mediawiki->list({
 357                action => 'query',
 358                list => 'allpages',
 359                apnamespace => get_mw_namespace_id("File"),
 360                aplimit => 'max'
 361        });
 362        if (!defined($mw_pages)) {
 363                print STDERR "fatal: could not get the list of pages for media files.\n";
 364                print STDERR "fatal: '$url' does not appear to be a mediawiki\n";
 365                print STDERR "fatal: make sure '$url/api.php' is a valid page.\n";
 366                exit 1;
 367        }
 368        foreach my $page (@{$mw_pages}) {
 369                $pages->{$page->{title}} = $page;
 370        }
 371        return;
 372}
 373
 374sub get_linked_mediafiles {
 375        my $pages = shift;
 376        my @titles = map { $_->{title} } values(%{$pages});
 377
 378        # The query is split in small batches because of the MW API limit of
 379        # the number of links to be returned (500 links max).
 380        my $batch = 10;
 381        while (@titles) {
 382                if ($#titles < $batch) {
 383                        $batch = $#titles;
 384                }
 385                my @slice = @titles[0..$batch];
 386
 387                # pattern 'page1|page2|...' required by the API
 388                my $mw_titles = join('|', @slice);
 389
 390                # Media files could be included or linked from
 391                # a page, get all related
 392                my $query = {
 393                        action => 'query',
 394                        prop => 'links|images',
 395                        titles => $mw_titles,
 396                        plnamespace => get_mw_namespace_id("File"),
 397                        pllimit => 'max'
 398                };
 399                my $result = $mediawiki->api($query);
 400
 401                while (my ($id, $page) = each(%{$result->{query}->{pages}})) {
 402                        my @media_titles;
 403                        if (defined($page->{links})) {
 404                                my @link_titles
 405                                    = map { $_->{title} } @{$page->{links}};
 406                                push(@media_titles, @link_titles);
 407                        }
 408                        if (defined($page->{images})) {
 409                                my @image_titles
 410                                    = map { $_->{title} } @{$page->{images}};
 411                                push(@media_titles, @image_titles);
 412                        }
 413                        if (@media_titles) {
 414                                get_mw_page_list(\@media_titles, $pages);
 415                        }
 416                }
 417
 418                @titles = @titles[($batch+1)..$#titles];
 419        }
 420        return;
 421}
 422
 423sub get_mw_mediafile_for_page_revision {
 424        # Name of the file on Wiki, with the prefix.
 425        my $filename = shift;
 426        my $timestamp = shift;
 427        my %mediafile;
 428
 429        # Search if on a media file with given timestamp exists on
 430        # MediaWiki. In that case download the file.
 431        my $query = {
 432                action => 'query',
 433                prop => 'imageinfo',
 434                titles => "File:" . $filename,
 435                iistart => $timestamp,
 436                iiend => $timestamp,
 437                iiprop => 'timestamp|archivename|url',
 438                iilimit => 1
 439        };
 440        my $result = $mediawiki->api($query);
 441
 442        my ($fileid, $file) = each( %{$result->{query}->{pages}} );
 443        # If not defined it means there is no revision of the file for
 444        # given timestamp.
 445        if (defined($file->{imageinfo})) {
 446                $mediafile{title} = $filename;
 447
 448                my $fileinfo = pop(@{$file->{imageinfo}});
 449                $mediafile{timestamp} = $fileinfo->{timestamp};
 450                # Mediawiki::API's download function doesn't support https URLs
 451                # and can't download old versions of files.
 452                print STDERR "\tDownloading file $mediafile{title}, version $mediafile{timestamp}\n";
 453                $mediafile{content} = download_mw_mediafile($fileinfo->{url});
 454        }
 455        return %mediafile;
 456}
 457
 458sub download_mw_mediafile {
 459        my $download_url = shift;
 460
 461        my $response = $mediawiki->{ua}->get($download_url);
 462        if ($response->code == 200) {
 463                return $response->decoded_content;
 464        } else {
 465                print STDERR "Error downloading mediafile from :\n";
 466                print STDERR "URL: $download_url\n";
 467                print STDERR "Server response: " . $response->code . " " . $response->message . "\n";
 468                exit 1;
 469        }
 470}
 471
 472sub get_last_local_revision {
 473        # Get note regarding last mediawiki revision
 474        my $note = run_git("notes --ref=$remotename/mediawiki show refs/mediawiki/$remotename/master 2>/dev/null");
 475        my @note_info = split(/ /, $note);
 476
 477        my $lastrevision_number;
 478        if (!(defined($note_info[0]) && $note_info[0] eq "mediawiki_revision:")) {
 479                print STDERR "No previous mediawiki revision found";
 480                $lastrevision_number = 0;
 481        } else {
 482                # Notes are formatted : mediawiki_revision: #number
 483                $lastrevision_number = $note_info[1];
 484                chomp($lastrevision_number);
 485                print STDERR "Last local mediawiki revision found is $lastrevision_number";
 486        }
 487        return $lastrevision_number;
 488}
 489
 490# Get the last remote revision without taking in account which pages are
 491# tracked or not. This function makes a single request to the wiki thus
 492# avoid a loop onto all tracked pages. This is useful for the fetch-by-rev
 493# option.
 494sub get_last_global_remote_rev {
 495        mw_connect_maybe();
 496
 497        my $query = {
 498                action => 'query',
 499                list => 'recentchanges',
 500                prop => 'revisions',
 501                rclimit => '1',
 502                rcdir => 'older',
 503        };
 504        my $result = $mediawiki->api($query);
 505        return $result->{query}->{recentchanges}[0]->{revid};
 506}
 507
 508# Get the last remote revision concerning the tracked pages and the tracked
 509# categories.
 510sub get_last_remote_revision {
 511        mw_connect_maybe();
 512
 513        my %pages_hash = get_mw_pages();
 514        my @pages = values(%pages_hash);
 515
 516        my $max_rev_num = 0;
 517
 518        print STDERR "Getting last revision id on tracked pages...\n";
 519
 520        foreach my $page (@pages) {
 521                my $id = $page->{pageid};
 522
 523                my $query = {
 524                        action => 'query',
 525                        prop => 'revisions',
 526                        rvprop => 'ids|timestamp',
 527                        pageids => $id,
 528                };
 529
 530                my $result = $mediawiki->api($query);
 531
 532                my $lastrev = pop(@{$result->{query}->{pages}->{$id}->{revisions}});
 533
 534                $basetimestamps{$lastrev->{revid}} = $lastrev->{timestamp};
 535
 536                $max_rev_num = ($lastrev->{revid} > $max_rev_num ? $lastrev->{revid} : $max_rev_num);
 537        }
 538
 539        print STDERR "Last remote revision found is $max_rev_num.\n";
 540        return $max_rev_num;
 541}
 542
 543# Clean content before sending it to MediaWiki
 544sub mediawiki_clean {
 545        my $string = shift;
 546        my $page_created = shift;
 547        # Mediawiki does not allow blank space at the end of a page and ends with a single \n.
 548        # This function right trims a string and adds a \n at the end to follow this rule
 549        $string =~ s/\s+$//;
 550        if ($string eq "" && $page_created) {
 551                # Creating empty pages is forbidden.
 552                $string = EMPTY_CONTENT;
 553        }
 554        return $string."\n";
 555}
 556
 557# Filter applied on MediaWiki data before adding them to Git
 558sub mediawiki_smudge {
 559        my $string = shift;
 560        if ($string eq EMPTY_CONTENT) {
 561                $string = "";
 562        }
 563        # This \n is important. This is due to mediawiki's way to handle end of files.
 564        return $string."\n";
 565}
 566
 567sub mediawiki_clean_filename {
 568        my $filename = shift;
 569        $filename =~ s{@{[SLASH_REPLACEMENT]}}{/}g;
 570        # [, ], |, {, and } are forbidden by MediaWiki, even URL-encoded.
 571        # Do a variant of URL-encoding, i.e. looks like URL-encoding,
 572        # but with _ added to prevent MediaWiki from thinking this is
 573        # an actual special character.
 574        $filename =~ s/[\[\]\{\}\|]/sprintf("_%%_%x", ord($&))/ge;
 575        # If we use the uri escape before
 576        # we should unescape here, before anything
 577
 578        return $filename;
 579}
 580
 581sub mediawiki_smudge_filename {
 582        my $filename = shift;
 583        $filename =~ s{/}{@{[SLASH_REPLACEMENT]}}g;
 584        $filename =~ s/ /_/g;
 585        # Decode forbidden characters encoded in mediawiki_clean_filename
 586        $filename =~ s/_%_([0-9a-fA-F][0-9a-fA-F])/sprintf("%c", hex($1))/ge;
 587        return $filename;
 588}
 589
 590sub literal_data {
 591        my ($content) = @_;
 592        print STDOUT "data ", bytes::length($content), "\n", $content;
 593        return;
 594}
 595
 596sub literal_data_raw {
 597        # Output possibly binary content.
 598        my ($content) = @_;
 599        # Avoid confusion between size in bytes and in characters
 600        utf8::downgrade($content);
 601        binmode STDOUT, ":raw";
 602        print STDOUT "data ", bytes::length($content), "\n", $content;
 603        binmode STDOUT, ":encoding(UTF-8)";
 604        return;
 605}
 606
 607sub mw_capabilities {
 608        # Revisions are imported to the private namespace
 609        # refs/mediawiki/$remotename/ by the helper and fetched into
 610        # refs/remotes/$remotename later by fetch.
 611        print STDOUT "refspec refs/heads/*:refs/mediawiki/$remotename/*\n";
 612        print STDOUT "import\n";
 613        print STDOUT "list\n";
 614        print STDOUT "push\n";
 615        print STDOUT "\n";
 616        return;
 617}
 618
 619sub mw_list {
 620        # MediaWiki do not have branches, we consider one branch arbitrarily
 621        # called master, and HEAD pointing to it.
 622        print STDOUT "? refs/heads/master\n";
 623        print STDOUT "\@refs/heads/master HEAD\n";
 624        print STDOUT "\n";
 625        return;
 626}
 627
 628sub mw_option {
 629        print STDERR "remote-helper command 'option $_[0]' not yet implemented\n";
 630        print STDOUT "unsupported\n";
 631        return;
 632}
 633
 634sub fetch_mw_revisions_for_page {
 635        my $page = shift;
 636        my $id = shift;
 637        my $fetch_from = shift;
 638        my @page_revs = ();
 639        my $query = {
 640                action => 'query',
 641                prop => 'revisions',
 642                rvprop => 'ids',
 643                rvdir => 'newer',
 644                rvstartid => $fetch_from,
 645                rvlimit => 500,
 646                pageids => $id,
 647        };
 648
 649        my $revnum = 0;
 650        # Get 500 revisions at a time due to the mediawiki api limit
 651        while (1) {
 652                my $result = $mediawiki->api($query);
 653
 654                # Parse each of those 500 revisions
 655                foreach my $revision (@{$result->{query}->{pages}->{$id}->{revisions}}) {
 656                        my $page_rev_ids;
 657                        $page_rev_ids->{pageid} = $page->{pageid};
 658                        $page_rev_ids->{revid} = $revision->{revid};
 659                        push(@page_revs, $page_rev_ids);
 660                        $revnum++;
 661                }
 662                last unless $result->{'query-continue'};
 663                $query->{rvstartid} = $result->{'query-continue'}->{revisions}->{rvstartid};
 664        }
 665        if ($shallow_import && @page_revs) {
 666                print STDERR "  Found 1 revision (shallow import).\n";
 667                @page_revs = sort {$b->{revid} <=> $a->{revid}} (@page_revs);
 668                return $page_revs[0];
 669        }
 670        print STDERR "  Found ", $revnum, " revision(s).\n";
 671        return @page_revs;
 672}
 673
 674sub fetch_mw_revisions {
 675        my $pages = shift; my @pages = @{$pages};
 676        my $fetch_from = shift;
 677
 678        my @revisions = ();
 679        my $n = 1;
 680        foreach my $page (@pages) {
 681                my $id = $page->{pageid};
 682
 683                print STDERR "page $n/", scalar(@pages), ": ". $page->{title} ."\n";
 684                $n++;
 685                my @page_revs = fetch_mw_revisions_for_page($page, $id, $fetch_from);
 686                @revisions = (@page_revs, @revisions);
 687        }
 688
 689        return ($n, @revisions);
 690}
 691
 692sub fe_escape_path {
 693    my $path = shift;
 694    $path =~ s/\\/\\\\/g;
 695    $path =~ s/"/\\"/g;
 696    $path =~ s/\n/\\n/g;
 697    return '"' . $path . '"';
 698}
 699
 700sub import_file_revision {
 701        my $commit = shift;
 702        my %commit = %{$commit};
 703        my $full_import = shift;
 704        my $n = shift;
 705        my $mediafile = shift;
 706        my %mediafile;
 707        if ($mediafile) {
 708                %mediafile = %{$mediafile};
 709        }
 710
 711        my $title = $commit{title};
 712        my $comment = $commit{comment};
 713        my $content = $commit{content};
 714        my $author = $commit{author};
 715        my $date = $commit{date};
 716
 717        print STDOUT "commit refs/mediawiki/$remotename/master\n";
 718        print STDOUT "mark :$n\n";
 719        print STDOUT "committer $author <$author\@$wiki_name> ", $date->epoch, " +0000\n";
 720        literal_data($comment);
 721
 722        # If it's not a clone, we need to know where to start from
 723        if (!$full_import && $n == 1) {
 724                print STDOUT "from refs/mediawiki/$remotename/master^0\n";
 725        }
 726        if ($content ne DELETED_CONTENT) {
 727                print STDOUT "M 644 inline " .
 728                    fe_escape_path($title . ".mw") . "\n";
 729                literal_data($content);
 730                if (%mediafile) {
 731                        print STDOUT "M 644 inline "
 732                            . fe_escape_path($mediafile{title}) . "\n";
 733                        literal_data_raw($mediafile{content});
 734                }
 735                print STDOUT "\n\n";
 736        } else {
 737                print STDOUT "D " . fe_escape_path($title . ".mw") . "\n";
 738        }
 739
 740        # mediawiki revision number in the git note
 741        if ($full_import && $n == 1) {
 742                print STDOUT "reset refs/notes/$remotename/mediawiki\n";
 743        }
 744        print STDOUT "commit refs/notes/$remotename/mediawiki\n";
 745        print STDOUT "committer $author <$author\@$wiki_name> ", $date->epoch, " +0000\n";
 746        literal_data("Note added by git-mediawiki during import");
 747        if (!$full_import && $n == 1) {
 748                print STDOUT "from refs/notes/$remotename/mediawiki^0\n";
 749        }
 750        print STDOUT "N inline :$n\n";
 751        literal_data("mediawiki_revision: " . $commit{mw_revision});
 752        print STDOUT "\n\n";
 753        return;
 754}
 755
 756# parse a sequence of
 757# <cmd> <arg1>
 758# <cmd> <arg2>
 759# \n
 760# (like batch sequence of import and sequence of push statements)
 761sub get_more_refs {
 762        my $cmd = shift;
 763        my @refs;
 764        while (1) {
 765                my $line = <STDIN>;
 766                if ($line =~ /^$cmd (.*)$/) {
 767                        push(@refs, $1);
 768                } elsif ($line eq "\n") {
 769                        return @refs;
 770                } else {
 771                        die("Invalid command in a '$cmd' batch: $_\n");
 772                }
 773        }
 774        return;
 775}
 776
 777sub mw_import {
 778        # multiple import commands can follow each other.
 779        my @refs = (shift, get_more_refs("import"));
 780        foreach my $ref (@refs) {
 781                mw_import_ref($ref);
 782        }
 783        print STDOUT "done\n";
 784        return;
 785}
 786
 787sub mw_import_ref {
 788        my $ref = shift;
 789        # The remote helper will call "import HEAD" and
 790        # "import refs/heads/master".
 791        # Since HEAD is a symbolic ref to master (by convention,
 792        # followed by the output of the command "list" that we gave),
 793        # we don't need to do anything in this case.
 794        if ($ref eq "HEAD") {
 795                return;
 796        }
 797
 798        mw_connect_maybe();
 799
 800        print STDERR "Searching revisions...\n";
 801        my $last_local = get_last_local_revision();
 802        my $fetch_from = $last_local + 1;
 803        if ($fetch_from == 1) {
 804                print STDERR ", fetching from beginning.\n";
 805        } else {
 806                print STDERR ", fetching from here.\n";
 807        }
 808
 809        my $n = 0;
 810        if ($fetch_strategy eq "by_rev") {
 811                print STDERR "Fetching & writing export data by revs...\n";
 812                $n = mw_import_ref_by_revs($fetch_from);
 813        } elsif ($fetch_strategy eq "by_page") {
 814                print STDERR "Fetching & writing export data by pages...\n";
 815                $n = mw_import_ref_by_pages($fetch_from);
 816        } else {
 817                print STDERR "fatal: invalid fetch strategy \"$fetch_strategy\".\n";
 818                print STDERR "Check your configuration variables remote.$remotename.fetchStrategy and mediawiki.fetchStrategy\n";
 819                exit 1;
 820        }
 821
 822        if ($fetch_from == 1 && $n == 0) {
 823                print STDERR "You appear to have cloned an empty MediaWiki.\n";
 824                # Something has to be done remote-helper side. If nothing is done, an error is
 825                # thrown saying that HEAD is referring to unknown object 0000000000000000000
 826                # and the clone fails.
 827        }
 828        return;
 829}
 830
 831sub mw_import_ref_by_pages {
 832
 833        my $fetch_from = shift;
 834        my %pages_hash = get_mw_pages();
 835        my @pages = values(%pages_hash);
 836
 837        my ($n, @revisions) = fetch_mw_revisions(\@pages, $fetch_from);
 838
 839        @revisions = sort {$a->{revid} <=> $b->{revid}} @revisions;
 840        my @revision_ids = map { $_->{revid} } @revisions;
 841
 842        return mw_import_revids($fetch_from, \@revision_ids, \%pages_hash);
 843}
 844
 845sub mw_import_ref_by_revs {
 846
 847        my $fetch_from = shift;
 848        my %pages_hash = get_mw_pages();
 849
 850        my $last_remote = get_last_global_remote_rev();
 851        my @revision_ids = $fetch_from..$last_remote;
 852        return mw_import_revids($fetch_from, \@revision_ids, \%pages_hash);
 853}
 854
 855# Import revisions given in second argument (array of integers).
 856# Only pages appearing in the third argument (hash indexed by page titles)
 857# will be imported.
 858sub mw_import_revids {
 859        my $fetch_from = shift;
 860        my $revision_ids = shift;
 861        my $pages = shift;
 862
 863        my $n = 0;
 864        my $n_actual = 0;
 865        my $last_timestamp = 0; # Placeholer in case $rev->timestamp is undefined
 866
 867        foreach my $pagerevid (@$revision_ids) {
 868                # Count page even if we skip it, since we display
 869                # $n/$total and $total includes skipped pages.
 870                $n++;
 871
 872                # fetch the content of the pages
 873                my $query = {
 874                        action => 'query',
 875                        prop => 'revisions',
 876                        rvprop => 'content|timestamp|comment|user|ids',
 877                        revids => $pagerevid,
 878                };
 879
 880                my $result = $mediawiki->api($query);
 881
 882                if (!$result) {
 883                        die "Failed to retrieve modified page for revision $pagerevid\n";
 884                }
 885
 886                if (defined($result->{query}->{badrevids}->{$pagerevid})) {
 887                        # The revision id does not exist on the remote wiki.
 888                        next;
 889                }
 890
 891                if (!defined($result->{query}->{pages})) {
 892                        die "Invalid revision $pagerevid.\n";
 893                }
 894
 895                my @result_pages = values(%{$result->{query}->{pages}});
 896                my $result_page = $result_pages[0];
 897                my $rev = $result_pages[0]->{revisions}->[0];
 898
 899                my $page_title = $result_page->{title};
 900
 901                if (!exists($pages->{$page_title})) {
 902                        print STDERR "$n/", scalar(@$revision_ids),
 903                                ": Skipping revision #$rev->{revid} of $page_title\n";
 904                        next;
 905                }
 906
 907                $n_actual++;
 908
 909                my %commit;
 910                $commit{author} = $rev->{user} || 'Anonymous';
 911                $commit{comment} = $rev->{comment} || EMPTY_MESSAGE;
 912                $commit{title} = mediawiki_smudge_filename($page_title);
 913                $commit{mw_revision} = $rev->{revid};
 914                $commit{content} = mediawiki_smudge($rev->{'*'});
 915
 916                if (!defined($rev->{timestamp})) {
 917                        $last_timestamp++;
 918                } else {
 919                        $last_timestamp = $rev->{timestamp};
 920                }
 921                $commit{date} = DateTime::Format::ISO8601->parse_datetime($last_timestamp);
 922
 923                # Differentiates classic pages and media files.
 924                my ($namespace, $filename) = $page_title =~ /^([^:]*):(.*)$/;
 925                my %mediafile;
 926                if ($namespace) {
 927                        my $id = get_mw_namespace_id($namespace);
 928                        if ($id && $id == get_mw_namespace_id("File")) {
 929                                %mediafile = get_mw_mediafile_for_page_revision($filename, $rev->{timestamp});
 930                        }
 931                }
 932                # If this is a revision of the media page for new version
 933                # of a file do one common commit for both file and media page.
 934                # Else do commit only for that page.
 935                print STDERR "$n/", scalar(@$revision_ids), ": Revision #$rev->{revid} of $commit{title}\n";
 936                import_file_revision(\%commit, ($fetch_from == 1), $n_actual, \%mediafile);
 937        }
 938
 939        return $n_actual;
 940}
 941
 942sub error_non_fast_forward {
 943        my $advice = run_git("config --bool advice.pushNonFastForward");
 944        chomp($advice);
 945        if ($advice ne "false") {
 946                # Native git-push would show this after the summary.
 947                # We can't ask it to display it cleanly, so print it
 948                # ourselves before.
 949                print STDERR "To prevent you from losing history, non-fast-forward updates were rejected\n";
 950                print STDERR "Merge the remote changes (e.g. 'git pull') before pushing again. See the\n";
 951                print STDERR "'Note about fast-forwards' section of 'git push --help' for details.\n";
 952        }
 953        print STDOUT "error $_[0] \"non-fast-forward\"\n";
 954        return 0;
 955}
 956
 957sub mw_upload_file {
 958        my $complete_file_name = shift;
 959        my $new_sha1 = shift;
 960        my $extension = shift;
 961        my $file_deleted = shift;
 962        my $summary = shift;
 963        my $newrevid;
 964        my $path = "File:" . $complete_file_name;
 965        my %hashFiles = get_allowed_file_extensions();
 966        if (!exists($hashFiles{$extension})) {
 967                print STDERR "$complete_file_name is not a permitted file on this wiki.\n";
 968                print STDERR "Check the configuration of file uploads in your mediawiki.\n";
 969                return $newrevid;
 970        }
 971        # Deleting and uploading a file requires a priviledged user
 972        if ($file_deleted) {
 973                mw_connect_maybe();
 974                my $query = {
 975                        action => 'delete',
 976                        title => $path,
 977                        reason => $summary
 978                };
 979                if (!$mediawiki->edit($query)) {
 980                        print STDERR "Failed to delete file on remote wiki\n";
 981                        print STDERR "Check your permissions on the remote site. Error code:\n";
 982                        print STDERR $mediawiki->{error}->{code} . ':' . $mediawiki->{error}->{details};
 983                        exit 1;
 984                }
 985        } else {
 986                # Don't let perl try to interpret file content as UTF-8 => use "raw"
 987                my $content = run_git("cat-file blob $new_sha1", "raw");
 988                if ($content ne "") {
 989                        mw_connect_maybe();
 990                        $mediawiki->{config}->{upload_url} =
 991                                "$url/index.php/Special:Upload";
 992                        $mediawiki->edit({
 993                                action => 'upload',
 994                                filename => $complete_file_name,
 995                                comment => $summary,
 996                                file => [undef,
 997                                         $complete_file_name,
 998                                         Content => $content],
 999                                ignorewarnings => 1,
1000                        }, {
1001                                skip_encoding => 1
1002                        } ) || die $mediawiki->{error}->{code} . ':'
1003                                 . $mediawiki->{error}->{details} . "\n";
1004                        my $last_file_page = $mediawiki->get_page({title => $path});
1005                        $newrevid = $last_file_page->{revid};
1006                        print STDERR "Pushed file: $new_sha1 - $complete_file_name.\n";
1007                } else {
1008                        print STDERR "Empty file $complete_file_name not pushed.\n";
1009                }
1010        }
1011        return $newrevid;
1012}
1013
1014sub mw_push_file {
1015        my $diff_info = shift;
1016        # $diff_info contains a string in this format:
1017        # 100644 100644 <sha1_of_blob_before_commit> <sha1_of_blob_now> <status>
1018        my @diff_info_split = split(/[ \t]/, $diff_info);
1019
1020        # Filename, including .mw extension
1021        my $complete_file_name = shift;
1022        # Commit message
1023        my $summary = shift;
1024        # MediaWiki revision number. Keep the previous one by default,
1025        # in case there's no edit to perform.
1026        my $oldrevid = shift;
1027        my $newrevid;
1028
1029        if ($summary eq EMPTY_MESSAGE) {
1030                $summary = '';
1031        }
1032
1033        my $new_sha1 = $diff_info_split[3];
1034        my $old_sha1 = $diff_info_split[2];
1035        my $page_created = ($old_sha1 eq NULL_SHA1);
1036        my $page_deleted = ($new_sha1 eq NULL_SHA1);
1037        $complete_file_name = mediawiki_clean_filename($complete_file_name);
1038
1039        my ($title, $extension) = $complete_file_name =~ /^(.*)\.([^\.]*)$/;
1040        if (!defined($extension)) {
1041                $extension = "";
1042        }
1043        if ($extension eq "mw") {
1044                my $ns = get_mw_namespace_id_for_page($complete_file_name);
1045                if ($ns && $ns == get_mw_namespace_id("File") && (!$export_media)) {
1046                        print STDERR "Ignoring media file related page: $complete_file_name\n";
1047                        return ($oldrevid, "ok");
1048                }
1049                my $file_content;
1050                if ($page_deleted) {
1051                        # Deleting a page usually requires
1052                        # special privileges. A common
1053                        # convention is to replace the page
1054                        # with this content instead:
1055                        $file_content = DELETED_CONTENT;
1056                } else {
1057                        $file_content = run_git("cat-file blob $new_sha1");
1058                }
1059
1060                mw_connect_maybe();
1061
1062                my $result = $mediawiki->edit( {
1063                        action => 'edit',
1064                        summary => $summary,
1065                        title => $title,
1066                        basetimestamp => $basetimestamps{$oldrevid},
1067                        text => mediawiki_clean($file_content, $page_created),
1068                                  }, {
1069                                          skip_encoding => 1 # Helps with names with accentuated characters
1070                                  });
1071                if (!$result) {
1072                        if ($mediawiki->{error}->{code} == 3) {
1073                                # edit conflicts, considered as non-fast-forward
1074                                print STDERR 'Warning: Error ' .
1075                                    $mediawiki->{error}->{code} .
1076                                    ' from mediwiki: ' . $mediawiki->{error}->{details} .
1077                                    ".\n";
1078                                return ($oldrevid, "non-fast-forward");
1079                        } else {
1080                                # Other errors. Shouldn't happen => just die()
1081                                die 'Fatal: Error ' .
1082                                    $mediawiki->{error}->{code} .
1083                                    ' from mediwiki: ' . $mediawiki->{error}->{details} . "\n";
1084                        }
1085                }
1086                $newrevid = $result->{edit}->{newrevid};
1087                print STDERR "Pushed file: $new_sha1 - $title\n";
1088        } elsif ($export_media) {
1089                $newrevid = mw_upload_file($complete_file_name, $new_sha1,
1090                                           $extension, $page_deleted,
1091                                           $summary);
1092        } else {
1093                print STDERR "Ignoring media file $title\n";
1094        }
1095        $newrevid = ($newrevid or $oldrevid);
1096        return ($newrevid, "ok");
1097}
1098
1099sub mw_push {
1100        # multiple push statements can follow each other
1101        my @refsspecs = (shift, get_more_refs("push"));
1102        my $pushed;
1103        for my $refspec (@refsspecs) {
1104                my ($force, $local, $remote) = $refspec =~ /^(\+)?([^:]*):([^:]*)$/
1105                    or die("Invalid refspec for push. Expected <src>:<dst> or +<src>:<dst>\n");
1106                if ($force) {
1107                        print STDERR "Warning: forced push not allowed on a MediaWiki.\n";
1108                }
1109                if ($local eq "") {
1110                        print STDERR "Cannot delete remote branch on a MediaWiki\n";
1111                        print STDOUT "error $remote cannot delete\n";
1112                        next;
1113                }
1114                if ($remote ne "refs/heads/master") {
1115                        print STDERR "Only push to the branch 'master' is supported on a MediaWiki\n";
1116                        print STDOUT "error $remote only master allowed\n";
1117                        next;
1118                }
1119                if (mw_push_revision($local, $remote)) {
1120                        $pushed = 1;
1121                }
1122        }
1123
1124        # Notify Git that the push is done
1125        print STDOUT "\n";
1126
1127        if ($pushed && $dumb_push) {
1128                print STDERR "Just pushed some revisions to MediaWiki.\n";
1129                print STDERR "The pushed revisions now have to be re-imported, and your current branch\n";
1130                print STDERR "needs to be updated with these re-imported commits. You can do this with\n";
1131                print STDERR "\n";
1132                print STDERR "  git pull --rebase\n";
1133                print STDERR "\n";
1134        }
1135        return;
1136}
1137
1138sub mw_push_revision {
1139        my $local = shift;
1140        my $remote = shift; # actually, this has to be "refs/heads/master" at this point.
1141        my $last_local_revid = get_last_local_revision();
1142        print STDERR ".\n"; # Finish sentence started by get_last_local_revision()
1143        my $last_remote_revid = get_last_remote_revision();
1144        my $mw_revision = $last_remote_revid;
1145
1146        # Get sha1 of commit pointed by local HEAD
1147        my $HEAD_sha1 = run_git("rev-parse $local 2>/dev/null"); chomp($HEAD_sha1);
1148        # Get sha1 of commit pointed by remotes/$remotename/master
1149        my $remoteorigin_sha1 = run_git("rev-parse refs/remotes/$remotename/master 2>/dev/null");
1150        chomp($remoteorigin_sha1);
1151
1152        if ($last_local_revid > 0 &&
1153            $last_local_revid < $last_remote_revid) {
1154                return error_non_fast_forward($remote);
1155        }
1156
1157        if ($HEAD_sha1 eq $remoteorigin_sha1) {
1158                # nothing to push
1159                return 0;
1160        }
1161
1162        # Get every commit in between HEAD and refs/remotes/origin/master,
1163        # including HEAD and refs/remotes/origin/master
1164        my @commit_pairs = ();
1165        if ($last_local_revid > 0) {
1166                my $parsed_sha1 = $remoteorigin_sha1;
1167                # Find a path from last MediaWiki commit to pushed commit
1168                print STDERR "Computing path from local to remote ...\n";
1169                my @local_ancestry = split(/\n/, run_git("rev-list --boundary --parents $local ^$parsed_sha1"));
1170                my %local_ancestry;
1171                foreach my $line (@local_ancestry) {
1172                        if (my ($child, $parents) = $line =~ /^-?([a-f0-9]+) ([a-f0-9 ]+)/) {
1173                                foreach my $parent (split(/ /, $parents)) {
1174                                        $local_ancestry{$parent} = $child;
1175                                }
1176                        } elsif (!$line =~ /^([a-f0-9]+)/) {
1177                                die "Unexpected output from git rev-list: $line\n";
1178                        }
1179                }
1180                while ($parsed_sha1 ne $HEAD_sha1) {
1181                        my $child = $local_ancestry{$parsed_sha1};
1182                        if (!$child) {
1183                                printf STDERR "Cannot find a path in history from remote commit to last commit\n";
1184                                return error_non_fast_forward($remote);
1185                        }
1186                        push(@commit_pairs, [$parsed_sha1, $child]);
1187                        $parsed_sha1 = $child;
1188                }
1189        } else {
1190                # No remote mediawiki revision. Export the whole
1191                # history (linearized with --first-parent)
1192                print STDERR "Warning: no common ancestor, pushing complete history\n";
1193                my $history = run_git("rev-list --first-parent --children $local");
1194                my @history = split(/\n/, $history);
1195                @history = @history[1..$#history];
1196                foreach my $line (reverse @history) {
1197                        my @commit_info_split = split(/[ \n]/, $line);
1198                        push(@commit_pairs, \@commit_info_split);
1199                }
1200        }
1201
1202        foreach my $commit_info_split (@commit_pairs) {
1203                my $sha1_child = @{$commit_info_split}[0];
1204                my $sha1_commit = @{$commit_info_split}[1];
1205                my $diff_infos = run_git("diff-tree -r --raw -z $sha1_child $sha1_commit");
1206                # TODO: we could detect rename, and encode them with a #redirect on the wiki.
1207                # TODO: for now, it's just a delete+add
1208                my @diff_info_list = split(/\0/, $diff_infos);
1209                # Keep the subject line of the commit message as mediawiki comment for the revision
1210                my $commit_msg = run_git("log --no-walk --format=\"%s\" $sha1_commit");
1211                chomp($commit_msg);
1212                # Push every blob
1213                while (@diff_info_list) {
1214                        my $status;
1215                        # git diff-tree -z gives an output like
1216                        # <metadata>\0<filename1>\0
1217                        # <metadata>\0<filename2>\0
1218                        # and we've split on \0.
1219                        my $info = shift(@diff_info_list);
1220                        my $file = shift(@diff_info_list);
1221                        ($mw_revision, $status) = mw_push_file($info, $file, $commit_msg, $mw_revision);
1222                        if ($status eq "non-fast-forward") {
1223                                # we may already have sent part of the
1224                                # commit to MediaWiki, but it's too
1225                                # late to cancel it. Stop the push in
1226                                # the middle, but still give an
1227                                # accurate error message.
1228                                return error_non_fast_forward($remote);
1229                        }
1230                        if ($status ne "ok") {
1231                                die("Unknown error from mw_push_file()\n");
1232                        }
1233                }
1234                unless ($dumb_push) {
1235                        run_git("notes --ref=$remotename/mediawiki add -f -m \"mediawiki_revision: $mw_revision\" $sha1_commit");
1236                        run_git("update-ref -m \"Git-MediaWiki push\" refs/mediawiki/$remotename/master $sha1_commit $sha1_child");
1237                }
1238        }
1239
1240        print STDOUT "ok $remote\n";
1241        return 1;
1242}
1243
1244sub get_allowed_file_extensions {
1245        mw_connect_maybe();
1246
1247        my $query = {
1248                action => 'query',
1249                meta => 'siteinfo',
1250                siprop => 'fileextensions'
1251                };
1252        my $result = $mediawiki->api($query);
1253        my @file_extensions = map { $_->{ext}} @{$result->{query}->{fileextensions}};
1254        my %hashFile = map { $_ => 1 } @file_extensions;
1255
1256        return %hashFile;
1257}
1258
1259# In memory cache for MediaWiki namespace ids.
1260my %namespace_id;
1261
1262# Namespaces whose id is cached in the configuration file
1263# (to avoid duplicates)
1264my %cached_mw_namespace_id;
1265
1266# Return MediaWiki id for a canonical namespace name.
1267# Ex.: "File", "Project".
1268sub get_mw_namespace_id {
1269        mw_connect_maybe();
1270        my $name = shift;
1271
1272        if (!exists $namespace_id{$name}) {
1273                # Look at configuration file, if the record for that namespace is
1274                # already cached. Namespaces are stored in form:
1275                # "Name_of_namespace:Id_namespace", ex.: "File:6".
1276                my @temp = split(/\n/, run_git("config --get-all remote."
1277                                                . $remotename .".namespaceCache"));
1278                chomp(@temp);
1279                foreach my $ns (@temp) {
1280                        my ($n, $id) = split(/:/, $ns);
1281                        if ($id eq 'notANameSpace') {
1282                                $namespace_id{$n} = {is_namespace => 0};
1283                        } else {
1284                                $namespace_id{$n} = {is_namespace => 1, id => $id};
1285                        }
1286                        $cached_mw_namespace_id{$n} = 1;
1287                }
1288        }
1289
1290        if (!exists $namespace_id{$name}) {
1291                print STDERR "Namespace $name not found in cache, querying the wiki ...\n";
1292                # NS not found => get namespace id from MW and store it in
1293                # configuration file.
1294                my $query = {
1295                        action => 'query',
1296                        meta => 'siteinfo',
1297                        siprop => 'namespaces'
1298                };
1299                my $result = $mediawiki->api($query);
1300
1301                while (my ($id, $ns) = each(%{$result->{query}->{namespaces}})) {
1302                        if (defined($ns->{id}) && defined($ns->{canonical})) {
1303                                $namespace_id{$ns->{canonical}} = {is_namespace => 1, id => $ns->{id}};
1304                                if ($ns->{'*'}) {
1305                                        # alias (e.g. french Fichier: as alias for canonical File:)
1306                                        $namespace_id{$ns->{'*'}} = {is_namespace => 1, id => $ns->{id}};
1307                                }
1308                        }
1309                }
1310        }
1311
1312        my $ns = $namespace_id{$name};
1313        my $id;
1314
1315        unless (defined $ns) {
1316                print STDERR "No such namespace $name on MediaWiki.\n";
1317                $ns = {is_namespace => 0};
1318                $namespace_id{$name} = $ns;
1319        }
1320
1321        if ($ns->{is_namespace}) {
1322                $id = $ns->{id};
1323        }
1324
1325        # Store "notANameSpace" as special value for inexisting namespaces
1326        my $store_id = ($id || 'notANameSpace');
1327
1328        # Store explicitely requested namespaces on disk
1329        if (!exists $cached_mw_namespace_id{$name}) {
1330                run_git("config --add remote.". $remotename
1331                        .".namespaceCache \"". $name .":". $store_id ."\"");
1332                $cached_mw_namespace_id{$name} = 1;
1333        }
1334        return $id;
1335}
1336
1337sub get_mw_namespace_id_for_page {
1338        my $namespace = shift;
1339        if ($namespace =~ /^([^:]*):/) {
1340                return get_mw_namespace_id($namespace);
1341        } else {
1342                return;
1343        }
1344}