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

   1 #! /usr/bin/perl
   2
   3 # Copyright (C) 2011
   4 #     Jérémie Nikaes <jeremie.nikaes@ensimag.imag.fr>
   5 #     Arnaud Lacurie <arnaud.lacurie@ensimag.imag.fr>
   6 #     Claire Fousse <claire.fousse@ensimag.imag.fr>
   7 #     David Amouyal <david.amouyal@ensimag.imag.fr>
   8 #     Matthieu Moy <matthieu.moy@grenoble-inp.fr>
   9 # License: GPL v2 or later
  10
  11 # Gateway between Git and MediaWiki.
  12 #   https://github.com/Bibzball/Git-Mediawiki/wiki
  13 #
  14 # Known limitations:
  15 #
  16 # - Only wiki pages are managed, no support for [[File:...]]
  17 #   attachments.
  18 #
  19 # - Poor performance in the best case: it takes forever to check
  20 #   whether we're up-to-date (on fetch or push) or to fetch a few
  21 #   revisions from a large wiki, because we use exclusively a
  22 #   page-based synchronization. We could switch to a wiki-wide
  23 #   synchronization when the synchronization involves few revisions
  24 #   but the wiki is large.
  25 #
  26 # - Git renames could be turned into MediaWiki renames (see TODO
  27 #   below)
  28 #
  29 # - login/password support requires the user to write the password
  30 #   cleartext in a file (see TODO below).
  31 #
  32 # - No way to import "one page, and all pages included in it"
  33 #
  34 # - Multiple remote MediaWikis have not been very well tested.
  35
  36 use strict;
  37 use MediaWiki::API;
  38 use DateTime::Format::ISO8601;
  39
  40 # By default, use UTF-8 to communicate with Git and the user
  41 binmode STDERR, ":utf8";
  42 binmode STDOUT, ":utf8";
  43
  44 use URI::Escape;
  45 use IPC::Open2;
  46
  47 use warnings;
  48
  49 # Mediawiki filenames can contain forward slashes. This variable decides by which pattern they should be replaced
  50 use constant SLASH_REPLACEMENT => "%2F";
  51
  52 # It's not always possible to delete pages (may require some
  53 # priviledges). Deleted pages are replaced with this content.
  54 use constant DELETED_CONTENT => "[[Category:Deleted]]\n";
  55
  56 # It's not possible to create empty pages. New empty files in Git are
  57 # sent with this content instead.
  58 use constant EMPTY_CONTENT => "<!-- empty page -->\n";
  59
  60 # used to reflect file creation or deletion in diff.
  61 use constant NULL_SHA1 => "0000000000000000000000000000000000000000";
  62
  63 my $remotename = $ARGV[0];
  64 my $url = $ARGV[1];
  65
  66 # Accept both space-separated and multiple keys in config file.
  67 # Spaces should be written as _ anyway because we'll use chomp.
  68 my @tracked_pages = split(/[ \n]/, run_git("config --get-all remote.". $remotename .".pages"));
  69 chomp(@tracked_pages);
  70
  71 # Just like @tracked_pages, but for MediaWiki categories.
  72 my @tracked_categories = split(/[ \n]/, run_git("config --get-all remote.". $remotename .".categories"));
  73 chomp(@tracked_categories);
  74
  75 my $wiki_login = run_git("config --get remote.". $remotename .".mwLogin");
  76 # TODO: ideally, this should be able to read from keyboard, but we're
  77 # inside a remote helper, so our stdin is connect to git, not to a
  78 # terminal.
  79 my $wiki_passwd = run_git("config --get remote.". $remotename .".mwPassword");
  80 my $wiki_domain = run_git("config --get remote.". $remotename .".mwDomain");
  81 chomp($wiki_login);
  82 chomp($wiki_passwd);
  83 chomp($wiki_domain);
  84
  85 # Import only last revisions (both for clone and fetch)
  86 my $shallow_import = run_git("config --get --bool remote.". $remotename .".shallow");
  87 chomp($shallow_import);
  88 $shallow_import = ($shallow_import eq "true");
  89
  90 # Dumb push: don't update notes and mediawiki ref to reflect the last push.
  91 #
  92 # Configurable with mediawiki.dumbPush, or per-remote with
  93 # remote.<remotename>.dumbPush.
  94 #
  95 # This means the user will have to re-import the just-pushed
  96 # revisions. On the other hand, this means that the Git revisions
  97 # corresponding to MediaWiki revisions are all imported from the wiki,
  98 # regardless of whether they were initially created in Git or from the
  99 # web interface, hence all users will get the same history (i.e. if
 100 # the push from Git to MediaWiki loses some information, everybody
 101 # will get the history with information lost). If the import is
 102 # deterministic, this means everybody gets the same sha1 for each
 103 # MediaWiki revision.
 104 my $dumb_push = run_git("config --get --bool remote.$remotename.dumbPush");
 105 unless ($dumb_push) {
 106         $dumb_push = run_git("config --get --bool mediawiki.dumbPush");
 107 }
 108 chomp($dumb_push);
 109 $dumb_push = ($dumb_push eq "true");
 110
 111 my $wiki_name = $url;
 112 $wiki_name =~ s/[^\/]*:\/\///;
 113 # If URL is like http://user:password@example.com/, we clearly don't
 114 # want the password in $wiki_name. While we're there, also remove user
 115 # and '@' sign, to avoid author like MWUser@HTTPUser@host.com
 116 $wiki_name =~ s/^.*@//;
 117
 118 # Commands parser
 119 my $entry;
 120 my @cmd;
 121 while (<STDIN>) {
 122         chomp;
 123         @cmd = split(/ /);
 124         if (defined($cmd[0])) {
 125                 # Line not blank
 126                 if ($cmd[0] eq "capabilities") {
 127                         die("Too many arguments for capabilities") unless (!defined($cmd[1]));
 128                         mw_capabilities();
 129                 } elsif ($cmd[0] eq "list") {
 130                         die("Too many arguments for list") unless (!defined($cmd[2]));
 131                         mw_list($cmd[1]);
 132                 } elsif ($cmd[0] eq "import") {
 133                         die("Invalid arguments for import") unless ($cmd[1] ne "" && !defined($cmd[2]));
 134                         mw_import($cmd[1]);
 135                 } elsif ($cmd[0] eq "option") {
 136                         die("Too many arguments for option") unless ($cmd[1] ne "" && $cmd[2] ne "" && !defined($cmd[3]));
 137                         mw_option($cmd[1],$cmd[2]);
 138                 } elsif ($cmd[0] eq "push") {
 139                         mw_push($cmd[1]);
 140                 } else {
 141                         print STDERR "Unknown command. Aborting...\n";
 142                         last;
 143                 }
 144         } else {
 145                 # blank line: we should terminate
 146                 last;
 147         }
 148
 149         BEGIN { $| = 1 } # flush STDOUT, to make sure the previous
 150                          # command is fully processed.
 151 }
 152
 153 ########################## Functions ##############################
 154
 155 ## credential API management (generic functions)
 156
 157 sub credential_from_url {
 158         my $url = shift;
 159         my $parsed = URI->new($url);
 160         my %credential;
 161
 162         if ($parsed->scheme) {
 163                 $credential{protocol} = $parsed->scheme;
 164         }
 165         if ($parsed->host) {
 166                 $credential{host} = $parsed->host;
 167         }
 168         if ($parsed->path) {
 169                 $credential{path} = $parsed->path;
 170         }
 171         if ($parsed->userinfo) {
 172                 if ($parsed->userinfo =~ /([^:]*):(.*)/) {
 173                         $credential{username} = $1;
 174                         $credential{password} = $2;
 175                 } else {
 176                         $credential{username} = $parsed->userinfo;
 177                 }
 178         }
 179
 180         return %credential;
 181 }
 182
 183 sub credential_read {
 184         my %credential;
 185         my $reader = shift;
 186         my $op = shift;
 187         while (<$reader>) {
 188                 my ($key, $value) = /([^=]*)=(.*)/;
 189                 if (not defined $key) {
 190                         die "ERROR receiving response from git credential $op:\n$_\n";
 191                 }
 192                 $credential{$key} = $value;
 193         }
 194         return %credential;
 195 }
 196
 197 sub credential_write {
 198         my $credential = shift;
 199         my $writer = shift;
 200         while (my ($key, $value) = each(%$credential) ) {
 201                 if ($value) {
 202                         print $writer "$key=$value\n";
 203                 }
 204         }
 205 }
 206
 207 sub credential_run {
 208         my $op = shift;
 209         my $credential = shift;
 210         my $pid = open2(my $reader, my $writer, "git credential $op");
 211         credential_write($credential, $writer);
 212         print $writer "\n";
 213         close($writer);
 214
 215         if ($op eq "fill") {
 216                 %$credential = credential_read($reader, $op);
 217         } else {
 218                 if (<$reader>) {
 219                         die "ERROR while running git credential $op:\n$_";
 220                 }
 221         }
 222         close($reader);
 223         waitpid($pid, 0);
 224         my $child_exit_status = $? >> 8;
 225         if ($child_exit_status != 0) {
 226                 die "'git credential $op' failed with code $child_exit_status.";
 227         }
 228 }
 229
 230 # MediaWiki API instance, created lazily.
 231 my $mediawiki;
 232
 233 sub mw_connect_maybe {
 234         if ($mediawiki) {
 235                 return;
 236         }
 237         $mediawiki = MediaWiki::API->new;
 238         $mediawiki->{config}->{api_url} = "$url/api.php";
 239         if ($wiki_login) {
 240                 my %credential = credential_from_url($url);
 241                 $credential{username} = $wiki_login;
 242                 $credential{password} = $wiki_passwd;
 243                 credential_run("fill", \%credential);
 244                 my $request = {lgname => $credential{username},
 245                                lgpassword => $credential{password},
 246                                lgdomain => $wiki_domain};
 247                 if ($mediawiki->login($request)) {
 248                         credential_run("approve", \%credential);
 249                         print STDERR "Logged in mediawiki user \"$credential{username}\".\n";
 250                 } else {
 251                         print STDERR "Failed to log in mediawiki user \"$credential{username}\" on $url\n";
 252                         print STDERR "  (error " .
 253                                 $mediawiki->{error}->{code} . ': ' .
 254                                 $mediawiki->{error}->{details} . ")\n";
 255                         credential_run("reject", \%credential);
 256                         exit 1;
 257                 }
 258         }
 259 }
 260
 261 sub get_mw_first_pages {
 262         my $some_pages = shift;
 263         my @some_pages = @{$some_pages};
 264
 265         my $pages = shift;
 266
 267         # pattern 'page1|page2|...' required by the API
 268         my $titles = join('|', @some_pages);
 269
 270         my $mw_pages = $mediawiki->api({
 271                 action => 'query',
 272                 titles => $titles,
 273         });
 274         if (!defined($mw_pages)) {
 275                 print STDERR "fatal: could not query the list of wiki pages.\n";
 276                 print STDERR "fatal: '$url' does not appear to be a mediawiki\n";
 277                 print STDERR "fatal: make sure '$url/api.php' is a valid page.\n";
 278                 exit 1;
 279         }
 280         while (my ($id, $page) = each(%{$mw_pages->{query}->{pages}})) {
 281                 if ($id < 0) {
 282                         print STDERR "Warning: page $page->{title} not found on wiki\n";
 283                 } else {
 284                         $pages->{$page->{title}} = $page;
 285                 }
 286         }
 287 }
 288
 289 sub get_mw_pages {
 290         mw_connect_maybe();
 291
 292         my %pages; # hash on page titles to avoid duplicates
 293         my $user_defined;
 294         if (@tracked_pages) {
 295                 $user_defined = 1;
 296                 # The user provided a list of pages titles, but we
 297                 # still need to query the API to get the page IDs.
 298
 299                 my @some_pages = @tracked_pages;
 300                 while (@some_pages) {
 301                         my $last = 50;
 302                         if ($#some_pages < $last) {
 303                                 $last = $#some_pages;
 304                         }
 305                         my @slice = @some_pages[0..$last];
 306                         get_mw_first_pages(\@slice, \%pages);
 307                         @some_pages = @some_pages[51..$#some_pages];
 308                 }
 309         }
 310         if (@tracked_categories) {
 311                 $user_defined = 1;
 312                 foreach my $category (@tracked_categories) {
 313                         if (index($category, ':') < 0) {
 314                                 # Mediawiki requires the Category
 315                                 # prefix, but let's not force the user
 316                                 # to specify it.
 317                                 $category = "Category:" . $category;
 318                         }
 319                         my $mw_pages = $mediawiki->list( {
 320                                 action => 'query',
 321                                 list => 'categorymembers',
 322                                 cmtitle => $category,
 323                                 cmlimit => 'max' } )
 324                             || die $mediawiki->{error}->{code} . ': ' . $mediawiki->{error}->{details};
 325                         foreach my $page (@{$mw_pages}) {
 326                                 $pages{$page->{title}} = $page;
 327                         }
 328                 }
 329         }
 330         if (!$user_defined) {
 331                 # No user-provided list, get the list of pages from
 332                 # the API.
 333                 my $mw_pages = $mediawiki->list({
 334                         action => 'query',
 335                         list => 'allpages',
 336                         aplimit => 500,
 337                 });
 338                 if (!defined($mw_pages)) {
 339                         print STDERR "fatal: could not get the list of wiki pages.\n";
 340                         print STDERR "fatal: '$url' does not appear to be a mediawiki\n";
 341                         print STDERR "fatal: make sure '$url/api.php' is a valid page.\n";
 342                         exit 1;
 343                 }
 344                 foreach my $page (@{$mw_pages}) {
 345                         $pages{$page->{title}} = $page;
 346                 }
 347         }
 348         return values(%pages);
 349 }
 350
 351 sub run_git {
 352         open(my $git, "-|:encoding(UTF-8)", "git " . $_[0]);
 353         my $res = do { local $/; <$git> };
 354         close($git);
 355
 356         return $res;
 357 }
 358
 359
 360 sub get_last_local_revision {
 361         # Get note regarding last mediawiki revision
 362         my $note = run_git("notes --ref=$remotename/mediawiki show refs/mediawiki/$remotename/master 2>/dev/null");
 363         my @note_info = split(/ /, $note);
 364
 365         my $lastrevision_number;
 366         if (!(defined($note_info[0]) && $note_info[0] eq "mediawiki_revision:")) {
 367                 print STDERR "No previous mediawiki revision found";
 368                 $lastrevision_number = 0;
 369         } else {
 370                 # Notes are formatted : mediawiki_revision: #number
 371                 $lastrevision_number = $note_info[1];
 372                 chomp($lastrevision_number);
 373                 print STDERR "Last local mediawiki revision found is $lastrevision_number";
 374         }
 375         return $lastrevision_number;
 376 }
 377
 378 # Remember the timestamp corresponding to a revision id.
 379 my %basetimestamps;
 380
 381 sub get_last_remote_revision {
 382         mw_connect_maybe();
 383
 384         my @pages = get_mw_pages();
 385
 386         my $max_rev_num = 0;
 387
 388         foreach my $page (@pages) {
 389                 my $id = $page->{pageid};
 390
 391                 my $query = {
 392                         action => 'query',
 393                         prop => 'revisions',
 394                         rvprop => 'ids|timestamp',
 395                         pageids => $id,
 396                 };
 397
 398                 my $result = $mediawiki->api($query);
 399
 400                 my $lastrev = pop(@{$result->{query}->{pages}->{$id}->{revisions}});
 401
 402                 $basetimestamps{$lastrev->{revid}} = $lastrev->{timestamp};
 403
 404                 $max_rev_num = ($lastrev->{revid} > $max_rev_num ? $lastrev->{revid} : $max_rev_num);
 405         }
 406
 407         print STDERR "Last remote revision found is $max_rev_num.\n";
 408         return $max_rev_num;
 409 }
 410
 411 # Clean content before sending it to MediaWiki
 412 sub mediawiki_clean {
 413         my $string = shift;
 414         my $page_created = shift;
 415         # Mediawiki does not allow blank space at the end of a page and ends with a single \n.
 416         # This function right trims a string and adds a \n at the end to follow this rule
 417         $string =~ s/\s+$//;
 418         if ($string eq "" && $page_created) {
 419                 # Creating empty pages is forbidden.
 420                 $string = EMPTY_CONTENT;
 421         }
 422         return $string."\n";
 423 }
 424
 425 # Filter applied on MediaWiki data before adding them to Git
 426 sub mediawiki_smudge {
 427         my $string = shift;
 428         if ($string eq EMPTY_CONTENT) {
 429                 $string = "";
 430         }
 431         # This \n is important. This is due to mediawiki's way to handle end of files.
 432         return $string."\n";
 433 }
 434
 435 sub mediawiki_clean_filename {
 436         my $filename = shift;
 437         $filename =~ s/@{[SLASH_REPLACEMENT]}/\//g;
 438         # [, ], |, {, and } are forbidden by MediaWiki, even URL-encoded.
 439         # Do a variant of URL-encoding, i.e. looks like URL-encoding,
 440         # but with _ added to prevent MediaWiki from thinking this is
 441         # an actual special character.
 442         $filename =~ s/[\[\]\{\}\|]/sprintf("_%%_%x", ord($&))/ge;
 443         # If we use the uri escape before
 444         # we should unescape here, before anything
 445
 446         return $filename;
 447 }
 448
 449 sub mediawiki_smudge_filename {
 450         my $filename = shift;
 451         $filename =~ s/\//@{[SLASH_REPLACEMENT]}/g;
 452         $filename =~ s/ /_/g;
 453         # Decode forbidden characters encoded in mediawiki_clean_filename
 454         $filename =~ s/_%_([0-9a-fA-F][0-9a-fA-F])/sprintf("%c", hex($1))/ge;
 455         return $filename;
 456 }
 457
 458 sub literal_data {
 459         my ($content) = @_;
 460         print STDOUT "data ", bytes::length($content), "\n", $content;
 461 }
 462
 463 sub mw_capabilities {
 464         # Revisions are imported to the private namespace
 465         # refs/mediawiki/$remotename/ by the helper and fetched into
 466         # refs/remotes/$remotename later by fetch.
 467         print STDOUT "refspec refs/heads/*:refs/mediawiki/$remotename/*\n";
 468         print STDOUT "import\n";
 469         print STDOUT "list\n";
 470         print STDOUT "push\n";
 471         print STDOUT "\n";
 472 }
 473
 474 sub mw_list {
 475         # MediaWiki do not have branches, we consider one branch arbitrarily
 476         # called master, and HEAD pointing to it.
 477         print STDOUT "? refs/heads/master\n";
 478         print STDOUT "\@refs/heads/master HEAD\n";
 479         print STDOUT "\n";
 480 }
 481
 482 sub mw_option {
 483         print STDERR "remote-helper command 'option $_[0]' not yet implemented\n";
 484         print STDOUT "unsupported\n";
 485 }
 486
 487 sub fetch_mw_revisions_for_page {
 488         my $page = shift;
 489         my $id = shift;
 490         my $fetch_from = shift;
 491         my @page_revs = ();
 492         my $query = {
 493                 action => 'query',
 494                 prop => 'revisions',
 495                 rvprop => 'ids',
 496                 rvdir => 'newer',
 497                 rvstartid => $fetch_from,
 498                 rvlimit => 500,
 499                 pageids => $id,
 500         };
 501
 502         my $revnum = 0;
 503         # Get 500 revisions at a time due to the mediawiki api limit
 504         while (1) {
 505                 my $result = $mediawiki->api($query);
 506
 507                 # Parse each of those 500 revisions
 508                 foreach my $revision (@{$result->{query}->{pages}->{$id}->{revisions}}) {
 509                         my $page_rev_ids;
 510                         $page_rev_ids->{pageid} = $page->{pageid};
 511                         $page_rev_ids->{revid} = $revision->{revid};
 512                         push(@page_revs, $page_rev_ids);
 513                         $revnum++;
 514                 }
 515                 last unless $result->{'query-continue'};
 516                 $query->{rvstartid} = $result->{'query-continue'}->{revisions}->{rvstartid};
 517         }
 518         if ($shallow_import && @page_revs) {
 519                 print STDERR "  Found 1 revision (shallow import).\n";
 520                 @page_revs = sort {$b->{revid} <=> $a->{revid}} (@page_revs);
 521                 return $page_revs[0];
 522         }
 523         print STDERR "  Found ", $revnum, " revision(s).\n";
 524         return @page_revs;
 525 }
 526
 527 sub fetch_mw_revisions {
 528         my $pages = shift; my @pages = @{$pages};
 529         my $fetch_from = shift;
 530
 531         my @revisions = ();
 532         my $n = 1;
 533         foreach my $page (@pages) {
 534                 my $id = $page->{pageid};
 535
 536                 print STDERR "page $n/", scalar(@pages), ": ". $page->{title} ."\n";
 537                 $n++;
 538                 my @page_revs = fetch_mw_revisions_for_page($page, $id, $fetch_from);
 539                 @revisions = (@page_revs, @revisions);
 540         }
 541
 542         return ($n, @revisions);
 543 }
 544
 545 sub import_file_revision {
 546         my $commit = shift;
 547         my %commit = %{$commit};
 548         my $full_import = shift;
 549         my $n = shift;
 550
 551         my $title = $commit{title};
 552         my $comment = $commit{comment};
 553         my $content = $commit{content};
 554         my $author = $commit{author};
 555         my $date = $commit{date};
 556
 557         print STDOUT "commit refs/mediawiki/$remotename/master\n";
 558         print STDOUT "mark :$n\n";
 559         print STDOUT "committer $author <$author\@$wiki_name> ", $date->epoch, " +0000\n";
 560         literal_data($comment);
 561
 562         # If it's not a clone, we need to know where to start from
 563         if (!$full_import && $n == 1) {
 564                 print STDOUT "from refs/mediawiki/$remotename/master^0\n";
 565         }
 566         if ($content ne DELETED_CONTENT) {
 567                 print STDOUT "M 644 inline $title.mw\n";
 568                 literal_data($content);
 569                 print STDOUT "\n\n";
 570         } else {
 571                 print STDOUT "D $title.mw\n";
 572         }
 573
 574         # mediawiki revision number in the git note
 575         if ($full_import && $n == 1) {
 576                 print STDOUT "reset refs/notes/$remotename/mediawiki\n";
 577         }
 578         print STDOUT "commit refs/notes/$remotename/mediawiki\n";
 579         print STDOUT "committer $author <$author\@$wiki_name> ", $date->epoch, " +0000\n";
 580         literal_data("Note added by git-mediawiki during import");
 581         if (!$full_import && $n == 1) {
 582                 print STDOUT "from refs/notes/$remotename/mediawiki^0\n";
 583         }
 584         print STDOUT "N inline :$n\n";
 585         literal_data("mediawiki_revision: " . $commit{mw_revision});
 586         print STDOUT "\n\n";
 587 }
 588
 589 # parse a sequence of
 590 # <cmd> <arg1>
 591 # <cmd> <arg2>
 592 # \n
 593 # (like batch sequence of import and sequence of push statements)
 594 sub get_more_refs {
 595         my $cmd = shift;
 596         my @refs;
 597         while (1) {
 598                 my $line = <STDIN>;
 599                 if ($line =~ m/^$cmd (.*)$/) {
 600                         push(@refs, $1);
 601                 } elsif ($line eq "\n") {
 602                         return @refs;
 603                 } else {
 604                         die("Invalid command in a '$cmd' batch: ". $_);
 605                 }
 606         }
 607 }
 608
 609 sub mw_import {
 610         # multiple import commands can follow each other.
 611         my @refs = (shift, get_more_refs("import"));
 612         foreach my $ref (@refs) {
 613                 mw_import_ref($ref);
 614         }
 615         print STDOUT "done\n";
 616 }
 617
 618 sub mw_import_ref {
 619         my $ref = shift;
 620         # The remote helper will call "import HEAD" and
 621         # "import refs/heads/master".
 622         # Since HEAD is a symbolic ref to master (by convention,
 623         # followed by the output of the command "list" that we gave),
 624         # we don't need to do anything in this case.
 625         if ($ref eq "HEAD") {
 626                 return;
 627         }
 628
 629         mw_connect_maybe();
 630
 631         my @pages = get_mw_pages();
 632
 633         print STDERR "Searching revisions...\n";
 634         my $last_local = get_last_local_revision();
 635         my $fetch_from = $last_local + 1;
 636         if ($fetch_from == 1) {
 637                 print STDERR ", fetching from beginning.\n";
 638         } else {
 639                 print STDERR ", fetching from here.\n";
 640         }
 641         my ($n, @revisions) = fetch_mw_revisions(\@pages, $fetch_from);
 642
 643         # Creation of the fast-import stream
 644         print STDERR "Fetching & writing export data...\n";
 645
 646         $n = 0;
 647         my $last_timestamp = 0; # Placeholer in case $rev->timestamp is undefined
 648
 649         foreach my $pagerevid (sort {$a->{revid} <=> $b->{revid}} @revisions) {
 650                 # fetch the content of the pages
 651                 my $query = {
 652                         action => 'query',
 653                         prop => 'revisions',
 654                         rvprop => 'content|timestamp|comment|user|ids',
 655                         revids => $pagerevid->{revid},
 656                 };
 657
 658                 my $result = $mediawiki->api($query);
 659
 660                 my $rev = pop(@{$result->{query}->{pages}->{$pagerevid->{pageid}}->{revisions}});
 661
 662                 $n++;
 663
 664                 my %commit;
 665                 $commit{author} = $rev->{user} || 'Anonymous';
 666                 $commit{comment} = $rev->{comment} || '*Empty MediaWiki Message*';
 667                 $commit{title} = mediawiki_smudge_filename(
 668                         $result->{query}->{pages}->{$pagerevid->{pageid}}->{title}
 669                     );
 670                 $commit{mw_revision} = $pagerevid->{revid};
 671                 $commit{content} = mediawiki_smudge($rev->{'*'});
 672
 673                 if (!defined($rev->{timestamp})) {
 674                         $last_timestamp++;
 675                 } else {
 676                         $last_timestamp = $rev->{timestamp};
 677                 }
 678                 $commit{date} = DateTime::Format::ISO8601->parse_datetime($last_timestamp);
 679
 680                 print STDERR "$n/", scalar(@revisions), ": Revision #$pagerevid->{revid} of $commit{title}\n";
 681
 682                 import_file_revision(\%commit, ($fetch_from == 1), $n);
 683         }
 684
 685         if ($fetch_from == 1 && $n == 0) {
 686                 print STDERR "You appear to have cloned an empty MediaWiki.\n";
 687                 # Something has to be done remote-helper side. If nothing is done, an error is
 688                 # thrown saying that HEAD is refering to unknown object 0000000000000000000
 689                 # and the clone fails.
 690         }
 691 }
 692
 693 sub error_non_fast_forward {
 694         my $advice = run_git("config --bool advice.pushNonFastForward");
 695         chomp($advice);
 696         if ($advice ne "false") {
 697                 # Native git-push would show this after the summary.
 698                 # We can't ask it to display it cleanly, so print it
 699                 # ourselves before.
 700                 print STDERR "To prevent you from losing history, non-fast-forward updates were rejected\n";
 701                 print STDERR "Merge the remote changes (e.g. 'git pull') before pushing again. See the\n";
 702                 print STDERR "'Note about fast-forwards' section of 'git push --help' for details.\n";
 703         }
 704         print STDOUT "error $_[0] \"non-fast-forward\"\n";
 705         return 0;
 706 }
 707
 708 sub mw_push_file {
 709         my $diff_info = shift;
 710         # $diff_info contains a string in this format:
 711         # 100644 100644 <sha1_of_blob_before_commit> <sha1_of_blob_now> <status>
 712         my @diff_info_split = split(/[ \t]/, $diff_info);
 713
 714         # Filename, including .mw extension
 715         my $complete_file_name = shift;
 716         # Commit message
 717         my $summary = shift;
 718         # MediaWiki revision number. Keep the previous one by default,
 719         # in case there's no edit to perform.
 720         my $newrevid = shift;
 721
 722         my $new_sha1 = $diff_info_split[3];
 723         my $old_sha1 = $diff_info_split[2];
 724         my $page_created = ($old_sha1 eq NULL_SHA1);
 725         my $page_deleted = ($new_sha1 eq NULL_SHA1);
 726         $complete_file_name = mediawiki_clean_filename($complete_file_name);
 727
 728         if (substr($complete_file_name,-3) eq ".mw") {
 729                 my $title = substr($complete_file_name,0,-3);
 730
 731                 my $file_content;
 732                 if ($page_deleted) {
 733                         # Deleting a page usually requires
 734                         # special priviledges. A common
 735                         # convention is to replace the page
 736                         # with this content instead:
 737                         $file_content = DELETED_CONTENT;
 738                 } else {
 739                         $file_content = run_git("cat-file blob $new_sha1");
 740                 }
 741
 742                 mw_connect_maybe();
 743
 744                 my $result = $mediawiki->edit( {
 745                         action => 'edit',
 746                         summary => $summary,
 747                         title => $title,
 748                         basetimestamp => $basetimestamps{$newrevid},
 749                         text => mediawiki_clean($file_content, $page_created),
 750                                   }, {
 751                                           skip_encoding => 1 # Helps with names with accentuated characters
 752                                   });
 753                 if (!$result) {
 754                         if ($mediawiki->{error}->{code} == 3) {
 755                                 # edit conflicts, considered as non-fast-forward
 756                                 print STDERR 'Warning: Error ' .
 757                                     $mediawiki->{error}->{code} .
 758                                     ' from mediwiki: ' . $mediawiki->{error}->{details} .
 759                                     ".\n";
 760                                 return ($newrevid, "non-fast-forward");
 761                         } else {
 762                                 # Other errors. Shouldn't happen => just die()
 763                                 die 'Fatal: Error ' .
 764                                     $mediawiki->{error}->{code} .
 765                                     ' from mediwiki: ' . $mediawiki->{error}->{details};
 766                         }
 767                 }
 768                 $newrevid = $result->{edit}->{newrevid};
 769                 print STDERR "Pushed file: $new_sha1 - $title\n";
 770         } else {
 771                 print STDERR "$complete_file_name not a mediawiki file (Not pushable on this version of git-remote-mediawiki).\n"
 772         }
 773         return ($newrevid, "ok");
 774 }
 775
 776 sub mw_push {
 777         # multiple push statements can follow each other
 778         my @refsspecs = (shift, get_more_refs("push"));
 779         my $pushed;
 780         for my $refspec (@refsspecs) {
 781                 my ($force, $local, $remote) = $refspec =~ /^(\+)?([^:]*):([^:]*)$/
 782                     or die("Invalid refspec for push. Expected <src>:<dst> or +<src>:<dst>");
 783                 if ($force) {
 784                         print STDERR "Warning: forced push not allowed on a MediaWiki.\n";
 785                 }
 786                 if ($local eq "") {
 787                         print STDERR "Cannot delete remote branch on a MediaWiki\n";
 788                         print STDOUT "error $remote cannot delete\n";
 789                         next;
 790                 }
 791                 if ($remote ne "refs/heads/master") {
 792                         print STDERR "Only push to the branch 'master' is supported on a MediaWiki\n";
 793                         print STDOUT "error $remote only master allowed\n";
 794                         next;
 795                 }
 796                 if (mw_push_revision($local, $remote)) {
 797                         $pushed = 1;
 798                 }
 799         }
 800
 801         # Notify Git that the push is done
 802         print STDOUT "\n";
 803
 804         if ($pushed && $dumb_push) {
 805                 print STDERR "Just pushed some revisions to MediaWiki.\n";
 806                 print STDERR "The pushed revisions now have to be re-imported, and your current branch\n";
 807                 print STDERR "needs to be updated with these re-imported commits. You can do this with\n";
 808                 print STDERR "\n";
 809                 print STDERR "  git pull --rebase\n";
 810                 print STDERR "\n";
 811         }
 812 }
 813
 814 sub mw_push_revision {
 815         my $local = shift;
 816         my $remote = shift; # actually, this has to be "refs/heads/master" at this point.
 817         my $last_local_revid = get_last_local_revision();
 818         print STDERR ".\n"; # Finish sentence started by get_last_local_revision()
 819         my $last_remote_revid = get_last_remote_revision();
 820         my $mw_revision = $last_remote_revid;
 821
 822         # Get sha1 of commit pointed by local HEAD
 823         my $HEAD_sha1 = run_git("rev-parse $local 2>/dev/null"); chomp($HEAD_sha1);
 824         # Get sha1 of commit pointed by remotes/$remotename/master
 825         my $remoteorigin_sha1 = run_git("rev-parse refs/remotes/$remotename/master 2>/dev/null");
 826         chomp($remoteorigin_sha1);
 827
 828         if ($last_local_revid > 0 &&
 829             $last_local_revid < $last_remote_revid) {
 830                 return error_non_fast_forward($remote);
 831         }
 832
 833         if ($HEAD_sha1 eq $remoteorigin_sha1) {
 834                 # nothing to push
 835                 return 0;
 836         }
 837
 838         # Get every commit in between HEAD and refs/remotes/origin/master,
 839         # including HEAD and refs/remotes/origin/master
 840         my @commit_pairs = ();
 841         if ($last_local_revid > 0) {
 842                 my $parsed_sha1 = $remoteorigin_sha1;
 843                 # Find a path from last MediaWiki commit to pushed commit
 844                 while ($parsed_sha1 ne $HEAD_sha1) {
 845                         my @commit_info =  grep(/^$parsed_sha1/, split(/\n/, run_git("rev-list --children $local")));
 846                         if (!@commit_info) {
 847                                 return error_non_fast_forward($remote);
 848                         }
 849                         my @commit_info_split = split(/ |\n/, $commit_info[0]);
 850                         # $commit_info_split[1] is the sha1 of the commit to export
 851                         # $commit_info_split[0] is the sha1 of its direct child
 852                         push(@commit_pairs, \@commit_info_split);
 853                         $parsed_sha1 = $commit_info_split[1];
 854                 }
 855         } else {
 856                 # No remote mediawiki revision. Export the whole
 857                 # history (linearized with --first-parent)
 858                 print STDERR "Warning: no common ancestor, pushing complete history\n";
 859                 my $history = run_git("rev-list --first-parent --children $local");
 860                 my @history = split('\n', $history);
 861                 @history = @history[1..$#history];
 862                 foreach my $line (reverse @history) {
 863                         my @commit_info_split = split(/ |\n/, $line);
 864                         push(@commit_pairs, \@commit_info_split);
 865                 }
 866         }
 867
 868         foreach my $commit_info_split (@commit_pairs) {
 869                 my $sha1_child = @{$commit_info_split}[0];
 870                 my $sha1_commit = @{$commit_info_split}[1];
 871                 my $diff_infos = run_git("diff-tree -r --raw -z $sha1_child $sha1_commit");
 872                 # TODO: we could detect rename, and encode them with a #redirect on the wiki.
 873                 # TODO: for now, it's just a delete+add
 874                 my @diff_info_list = split(/\0/, $diff_infos);
 875                 # Keep the subject line of the commit message as mediawiki comment for the revision
 876                 my $commit_msg = run_git("log --no-walk --format=\"%s\" $sha1_commit");
 877                 chomp($commit_msg);
 878                 # Push every blob
 879                 while (@diff_info_list) {
 880                         my $status;
 881                         # git diff-tree -z gives an output like
 882                         # <metadata>\0<filename1>\0
 883                         # <metadata>\0<filename2>\0
 884                         # and we've split on \0.
 885                         my $info = shift(@diff_info_list);
 886                         my $file = shift(@diff_info_list);
 887                         ($mw_revision, $status) = mw_push_file($info, $file, $commit_msg, $mw_revision);
 888                         if ($status eq "non-fast-forward") {
 889                                 # we may already have sent part of the
 890                                 # commit to MediaWiki, but it's too
 891                                 # late to cancel it. Stop the push in
 892                                 # the middle, but still give an
 893                                 # accurate error message.
 894                                 return error_non_fast_forward($remote);
 895                         }
 896                         if ($status ne "ok") {
 897                                 die("Unknown error from mw_push_file()");
 898                         }
 899                 }
 900                 unless ($dumb_push) {
 901                         run_git("notes --ref=$remotename/mediawiki add -m \"mediawiki_revision: $mw_revision\" $sha1_commit");
 902                         run_git("update-ref -m \"Git-MediaWiki push\" refs/mediawiki/$remotename/master $sha1_commit $sha1_child");
 903                 }
 904         }
 905
 906         print STDOUT "ok $remote\n";
 907         return 1;
 908 }