361dbb1b28569a98e919eeb19d6353130e64fdcb
   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
  36use strict;
  37use MediaWiki::API;
  38use DateTime::Format::ISO8601;
  39
  40# By default, use UTF-8 to communicate with Git and the user
  41binmode STDERR, ":utf8";
  42binmode STDOUT, ":utf8";
  43
  44use URI::Escape;
  45use IPC::Open2;
  46
  47use warnings;
  48
  49# Mediawiki filenames can contain forward slashes. This variable decides by which pattern they should be replaced
  50use 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.
  54use 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.
  58use constant EMPTY_CONTENT => "<!-- empty page -->\n";
  59
  60# used to reflect file creation or deletion in diff.
  61use constant NULL_SHA1 => "0000000000000000000000000000000000000000";
  62
  63my $remotename = $ARGV[0];
  64my $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.
  68my @tracked_pages = split(/[ \n]/, run_git("config --get-all remote.". $remotename .".pages"));
  69chomp(@tracked_pages);
  70
  71# Just like @tracked_pages, but for MediaWiki categories.
  72my @tracked_categories = split(/[ \n]/, run_git("config --get-all remote.". $remotename .".categories"));
  73chomp(@tracked_categories);
  74
  75my $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.
  79my $wiki_passwd = run_git("config --get remote.". $remotename .".mwPassword");
  80my $wiki_domain = run_git("config --get remote.". $remotename .".mwDomain");
  81chomp($wiki_login);
  82chomp($wiki_passwd);
  83chomp($wiki_domain);
  84
  85# Import only last revisions (both for clone and fetch)
  86my $shallow_import = run_git("config --get --bool remote.". $remotename .".shallow");
  87chomp($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.
 104my $dumb_push = run_git("config --get --bool remote.$remotename.dumbPush");
 105unless ($dumb_push) {
 106        $dumb_push = run_git("config --get --bool mediawiki.dumbPush");
 107}
 108chomp($dumb_push);
 109$dumb_push = ($dumb_push eq "true");
 110
 111my $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
 119my $entry;
 120my @cmd;
 121while (<STDIN>) {
 122        chomp;
 123        @cmd = split(/ /);
 124        if (defined($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
 157sub 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
 183sub 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
 197sub 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
 207sub 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.
 231my $mediawiki;
 232
 233sub mw_connect_maybe {
 234        if ($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## Functions for listing pages on the remote wiki
 262sub get_mw_tracked_pages {
 263        my $pages = shift;
 264        my @some_pages = @tracked_pages;
 265        while (@some_pages) {
 266                my $last = 50;
 267                if ($#some_pages < $last) {
 268                        $last = $#some_pages;
 269                }
 270                my @slice = @some_pages[0..$last];
 271                get_mw_first_pages(\@slice, $pages);
 272                @some_pages = @some_pages[51..$#some_pages];
 273        }
 274}
 275
 276sub get_mw_tracked_categories {
 277        my $pages = shift;
 278        foreach my $category (@tracked_categories) {
 279                if (index($category, ':') < 0) {
 280                        # Mediawiki requires the Category
 281                        # prefix, but let's not force the user
 282                        # to specify it.
 283                        $category = "Category:" . $category;
 284                }
 285                my $mw_pages = $mediawiki->list( {
 286                        action => 'query',
 287                        list => 'categorymembers',
 288                        cmtitle => $category,
 289                        cmlimit => 'max' } )
 290                        || die $mediawiki->{error}->{code} . ': '
 291                                . $mediawiki->{error}->{details};
 292                foreach my $page (@{$mw_pages}) {
 293                        $pages->{$page->{title}} = $page;
 294                }
 295        }
 296}
 297
 298sub get_mw_all_pages {
 299        my $pages = shift;
 300        # No user-provided list, get the list of pages from the API.
 301        my $mw_pages = $mediawiki->list({
 302                action => 'query',
 303                list => 'allpages',
 304                aplimit => 'max'
 305        });
 306        if (!defined($mw_pages)) {
 307                print STDERR "fatal: could not get the list of wiki pages.\n";
 308                print STDERR "fatal: '$url' does not appear to be a mediawiki\n";
 309                print STDERR "fatal: make sure '$url/api.php' is a valid page.\n";
 310                exit 1;
 311        }
 312        foreach my $page (@{$mw_pages}) {
 313                $pages->{$page->{title}} = $page;
 314        }
 315}
 316
 317# queries the wiki for a set of pages. Meant to be used within a loop
 318# querying the wiki for slices of page list.
 319sub get_mw_first_pages {
 320        my $some_pages = shift;
 321        my @some_pages = @{$some_pages};
 322
 323        my $pages = shift;
 324
 325        # pattern 'page1|page2|...' required by the API
 326        my $titles = join('|', @some_pages);
 327
 328        my $mw_pages = $mediawiki->api({
 329                action => 'query',
 330                titles => $titles,
 331        });
 332        if (!defined($mw_pages)) {
 333                print STDERR "fatal: could not query the list of wiki pages.\n";
 334                print STDERR "fatal: '$url' does not appear to be a mediawiki\n";
 335                print STDERR "fatal: make sure '$url/api.php' is a valid page.\n";
 336                exit 1;
 337        }
 338        while (my ($id, $page) = each(%{$mw_pages->{query}->{pages}})) {
 339                if ($id < 0) {
 340                        print STDERR "Warning: page $page->{title} not found on wiki\n";
 341                } else {
 342                        $pages->{$page->{title}} = $page;
 343                }
 344        }
 345}
 346
 347# Get the list of pages to be fetched according to configuration.
 348sub get_mw_pages {
 349        mw_connect_maybe();
 350
 351        my %pages; # hash on page titles to avoid duplicates
 352        my $user_defined;
 353        if (@tracked_pages) {
 354                $user_defined = 1;
 355                # The user provided a list of pages titles, but we
 356                # still need to query the API to get the page IDs.
 357                get_mw_tracked_pages(\%pages);
 358        }
 359        if (@tracked_categories) {
 360                $user_defined = 1;
 361                get_mw_tracked_categories(\%pages);
 362        }
 363        if (!$user_defined) {
 364                get_mw_all_pages(\%pages);
 365        }
 366        return values(%pages);
 367}
 368
 369# usage: $out = run_git("command args");
 370#        $out = run_git("command args", "raw"); # don't interpret output as UTF-8.
 371sub run_git {
 372        my $args = shift;
 373        my $encoding = (shift || "encoding(UTF-8)");
 374        open(my $git, "-|:$encoding", "git " . $args);
 375        my $res = do { local $/; <$git> };
 376        close($git);
 377
 378        return $res;
 379}
 380
 381
 382sub get_last_local_revision {
 383        # Get note regarding last mediawiki revision
 384        my $note = run_git("notes --ref=$remotename/mediawiki show refs/mediawiki/$remotename/master 2>/dev/null");
 385        my @note_info = split(/ /, $note);
 386
 387        my $lastrevision_number;
 388        if (!(defined($note_info[0]) && $note_info[0] eq "mediawiki_revision:")) {
 389                print STDERR "No previous mediawiki revision found";
 390                $lastrevision_number = 0;
 391        } else {
 392                # Notes are formatted : mediawiki_revision: #number
 393                $lastrevision_number = $note_info[1];
 394                chomp($lastrevision_number);
 395                print STDERR "Last local mediawiki revision found is $lastrevision_number";
 396        }
 397        return $lastrevision_number;
 398}
 399
 400# Remember the timestamp corresponding to a revision id.
 401my %basetimestamps;
 402
 403sub get_last_remote_revision {
 404        mw_connect_maybe();
 405
 406        my @pages = get_mw_pages();
 407
 408        my $max_rev_num = 0;
 409
 410        foreach my $page (@pages) {
 411                my $id = $page->{pageid};
 412
 413                my $query = {
 414                        action => 'query',
 415                        prop => 'revisions',
 416                        rvprop => 'ids|timestamp',
 417                        pageids => $id,
 418                };
 419
 420                my $result = $mediawiki->api($query);
 421
 422                my $lastrev = pop(@{$result->{query}->{pages}->{$id}->{revisions}});
 423
 424                $basetimestamps{$lastrev->{revid}} = $lastrev->{timestamp};
 425
 426                $max_rev_num = ($lastrev->{revid} > $max_rev_num ? $lastrev->{revid} : $max_rev_num);
 427        }
 428
 429        print STDERR "Last remote revision found is $max_rev_num.\n";
 430        return $max_rev_num;
 431}
 432
 433# Clean content before sending it to MediaWiki
 434sub mediawiki_clean {
 435        my $string = shift;
 436        my $page_created = shift;
 437        # Mediawiki does not allow blank space at the end of a page and ends with a single \n.
 438        # This function right trims a string and adds a \n at the end to follow this rule
 439        $string =~ s/\s+$//;
 440        if ($string eq "" && $page_created) {
 441                # Creating empty pages is forbidden.
 442                $string = EMPTY_CONTENT;
 443        }
 444        return $string."\n";
 445}
 446
 447# Filter applied on MediaWiki data before adding them to Git
 448sub mediawiki_smudge {
 449        my $string = shift;
 450        if ($string eq EMPTY_CONTENT) {
 451                $string = "";
 452        }
 453        # This \n is important. This is due to mediawiki's way to handle end of files.
 454        return $string."\n";
 455}
 456
 457sub mediawiki_clean_filename {
 458        my $filename = shift;
 459        $filename =~ s/@{[SLASH_REPLACEMENT]}/\//g;
 460        # [, ], |, {, and } are forbidden by MediaWiki, even URL-encoded.
 461        # Do a variant of URL-encoding, i.e. looks like URL-encoding,
 462        # but with _ added to prevent MediaWiki from thinking this is
 463        # an actual special character.
 464        $filename =~ s/[\[\]\{\}\|]/sprintf("_%%_%x", ord($&))/ge;
 465        # If we use the uri escape before
 466        # we should unescape here, before anything
 467
 468        return $filename;
 469}
 470
 471sub mediawiki_smudge_filename {
 472        my $filename = shift;
 473        $filename =~ s/\//@{[SLASH_REPLACEMENT]}/g;
 474        $filename =~ s/ /_/g;
 475        # Decode forbidden characters encoded in mediawiki_clean_filename
 476        $filename =~ s/_%_([0-9a-fA-F][0-9a-fA-F])/sprintf("%c", hex($1))/ge;
 477        return $filename;
 478}
 479
 480sub literal_data {
 481        my ($content) = @_;
 482        print STDOUT "data ", bytes::length($content), "\n", $content;
 483}
 484
 485sub mw_capabilities {
 486        # Revisions are imported to the private namespace
 487        # refs/mediawiki/$remotename/ by the helper and fetched into
 488        # refs/remotes/$remotename later by fetch.
 489        print STDOUT "refspec refs/heads/*:refs/mediawiki/$remotename/*\n";
 490        print STDOUT "import\n";
 491        print STDOUT "list\n";
 492        print STDOUT "push\n";
 493        print STDOUT "\n";
 494}
 495
 496sub mw_list {
 497        # MediaWiki do not have branches, we consider one branch arbitrarily
 498        # called master, and HEAD pointing to it.
 499        print STDOUT "? refs/heads/master\n";
 500        print STDOUT "\@refs/heads/master HEAD\n";
 501        print STDOUT "\n";
 502}
 503
 504sub mw_option {
 505        print STDERR "remote-helper command 'option $_[0]' not yet implemented\n";
 506        print STDOUT "unsupported\n";
 507}
 508
 509sub fetch_mw_revisions_for_page {
 510        my $page = shift;
 511        my $id = shift;
 512        my $fetch_from = shift;
 513        my @page_revs = ();
 514        my $query = {
 515                action => 'query',
 516                prop => 'revisions',
 517                rvprop => 'ids',
 518                rvdir => 'newer',
 519                rvstartid => $fetch_from,
 520                rvlimit => 500,
 521                pageids => $id,
 522        };
 523
 524        my $revnum = 0;
 525        # Get 500 revisions at a time due to the mediawiki api limit
 526        while (1) {
 527                my $result = $mediawiki->api($query);
 528
 529                # Parse each of those 500 revisions
 530                foreach my $revision (@{$result->{query}->{pages}->{$id}->{revisions}}) {
 531                        my $page_rev_ids;
 532                        $page_rev_ids->{pageid} = $page->{pageid};
 533                        $page_rev_ids->{revid} = $revision->{revid};
 534                        push(@page_revs, $page_rev_ids);
 535                        $revnum++;
 536                }
 537                last unless $result->{'query-continue'};
 538                $query->{rvstartid} = $result->{'query-continue'}->{revisions}->{rvstartid};
 539        }
 540        if ($shallow_import && @page_revs) {
 541                print STDERR "  Found 1 revision (shallow import).\n";
 542                @page_revs = sort {$b->{revid} <=> $a->{revid}} (@page_revs);
 543                return $page_revs[0];
 544        }
 545        print STDERR "  Found ", $revnum, " revision(s).\n";
 546        return @page_revs;
 547}
 548
 549sub fetch_mw_revisions {
 550        my $pages = shift; my @pages = @{$pages};
 551        my $fetch_from = shift;
 552
 553        my @revisions = ();
 554        my $n = 1;
 555        foreach my $page (@pages) {
 556                my $id = $page->{pageid};
 557
 558                print STDERR "page $n/", scalar(@pages), ": ". $page->{title} ."\n";
 559                $n++;
 560                my @page_revs = fetch_mw_revisions_for_page($page, $id, $fetch_from);
 561                @revisions = (@page_revs, @revisions);
 562        }
 563
 564        return ($n, @revisions);
 565}
 566
 567sub import_file_revision {
 568        my $commit = shift;
 569        my %commit = %{$commit};
 570        my $full_import = shift;
 571        my $n = shift;
 572
 573        my $title = $commit{title};
 574        my $comment = $commit{comment};
 575        my $content = $commit{content};
 576        my $author = $commit{author};
 577        my $date = $commit{date};
 578
 579        print STDOUT "commit refs/mediawiki/$remotename/master\n";
 580        print STDOUT "mark :$n\n";
 581        print STDOUT "committer $author <$author\@$wiki_name> ", $date->epoch, " +0000\n";
 582        literal_data($comment);
 583
 584        # If it's not a clone, we need to know where to start from
 585        if (!$full_import && $n == 1) {
 586                print STDOUT "from refs/mediawiki/$remotename/master^0\n";
 587        }
 588        if ($content ne DELETED_CONTENT) {
 589                print STDOUT "M 644 inline $title.mw\n";
 590                literal_data($content);
 591                print STDOUT "\n\n";
 592        } else {
 593                print STDOUT "D $title.mw\n";
 594        }
 595
 596        # mediawiki revision number in the git note
 597        if ($full_import && $n == 1) {
 598                print STDOUT "reset refs/notes/$remotename/mediawiki\n";
 599        }
 600        print STDOUT "commit refs/notes/$remotename/mediawiki\n";
 601        print STDOUT "committer $author <$author\@$wiki_name> ", $date->epoch, " +0000\n";
 602        literal_data("Note added by git-mediawiki during import");
 603        if (!$full_import && $n == 1) {
 604                print STDOUT "from refs/notes/$remotename/mediawiki^0\n";
 605        }
 606        print STDOUT "N inline :$n\n";
 607        literal_data("mediawiki_revision: " . $commit{mw_revision});
 608        print STDOUT "\n\n";
 609}
 610
 611# parse a sequence of
 612# <cmd> <arg1>
 613# <cmd> <arg2>
 614# \n
 615# (like batch sequence of import and sequence of push statements)
 616sub get_more_refs {
 617        my $cmd = shift;
 618        my @refs;
 619        while (1) {
 620                my $line = <STDIN>;
 621                if ($line =~ m/^$cmd (.*)$/) {
 622                        push(@refs, $1);
 623                } elsif ($line eq "\n") {
 624                        return @refs;
 625                } else {
 626                        die("Invalid command in a '$cmd' batch: ". $_);
 627                }
 628        }
 629}
 630
 631sub mw_import {
 632        # multiple import commands can follow each other.
 633        my @refs = (shift, get_more_refs("import"));
 634        foreach my $ref (@refs) {
 635                mw_import_ref($ref);
 636        }
 637        print STDOUT "done\n";
 638}
 639
 640sub mw_import_ref {
 641        my $ref = shift;
 642        # The remote helper will call "import HEAD" and
 643        # "import refs/heads/master".
 644        # Since HEAD is a symbolic ref to master (by convention,
 645        # followed by the output of the command "list" that we gave),
 646        # we don't need to do anything in this case.
 647        if ($ref eq "HEAD") {
 648                return;
 649        }
 650
 651        mw_connect_maybe();
 652
 653        my @pages = get_mw_pages();
 654
 655        print STDERR "Searching revisions...\n";
 656        my $last_local = get_last_local_revision();
 657        my $fetch_from = $last_local + 1;
 658        if ($fetch_from == 1) {
 659                print STDERR ", fetching from beginning.\n";
 660        } else {
 661                print STDERR ", fetching from here.\n";
 662        }
 663        my ($n, @revisions) = fetch_mw_revisions(\@pages, $fetch_from);
 664
 665        # Creation of the fast-import stream
 666        print STDERR "Fetching & writing export data...\n";
 667
 668        $n = 0;
 669        my $last_timestamp = 0; # Placeholer in case $rev->timestamp is undefined
 670
 671        foreach my $pagerevid (sort {$a->{revid} <=> $b->{revid}} @revisions) {
 672                # fetch the content of the pages
 673                my $query = {
 674                        action => 'query',
 675                        prop => 'revisions',
 676                        rvprop => 'content|timestamp|comment|user|ids',
 677                        revids => $pagerevid->{revid},
 678                };
 679
 680                my $result = $mediawiki->api($query);
 681
 682                my $rev = pop(@{$result->{query}->{pages}->{$pagerevid->{pageid}}->{revisions}});
 683
 684                $n++;
 685
 686                my %commit;
 687                $commit{author} = $rev->{user} || 'Anonymous';
 688                $commit{comment} = $rev->{comment} || '*Empty MediaWiki Message*';
 689                $commit{title} = mediawiki_smudge_filename(
 690                        $result->{query}->{pages}->{$pagerevid->{pageid}}->{title}
 691                    );
 692                $commit{mw_revision} = $pagerevid->{revid};
 693                $commit{content} = mediawiki_smudge($rev->{'*'});
 694
 695                if (!defined($rev->{timestamp})) {
 696                        $last_timestamp++;
 697                } else {
 698                        $last_timestamp = $rev->{timestamp};
 699                }
 700                $commit{date} = DateTime::Format::ISO8601->parse_datetime($last_timestamp);
 701
 702                print STDERR "$n/", scalar(@revisions), ": Revision #$pagerevid->{revid} of $commit{title}\n";
 703
 704                import_file_revision(\%commit, ($fetch_from == 1), $n);
 705        }
 706
 707        if ($fetch_from == 1 && $n == 0) {
 708                print STDERR "You appear to have cloned an empty MediaWiki.\n";
 709                # Something has to be done remote-helper side. If nothing is done, an error is
 710                # thrown saying that HEAD is refering to unknown object 0000000000000000000
 711                # and the clone fails.
 712        }
 713}
 714
 715sub error_non_fast_forward {
 716        my $advice = run_git("config --bool advice.pushNonFastForward");
 717        chomp($advice);
 718        if ($advice ne "false") {
 719                # Native git-push would show this after the summary.
 720                # We can't ask it to display it cleanly, so print it
 721                # ourselves before.
 722                print STDERR "To prevent you from losing history, non-fast-forward updates were rejected\n";
 723                print STDERR "Merge the remote changes (e.g. 'git pull') before pushing again. See the\n";
 724                print STDERR "'Note about fast-forwards' section of 'git push --help' for details.\n";
 725        }
 726        print STDOUT "error $_[0] \"non-fast-forward\"\n";
 727        return 0;
 728}
 729
 730sub mw_upload_file {
 731        my $complete_file_name = shift;
 732        my $new_sha1 = shift;
 733        my $extension = shift;
 734        my $file_deleted = shift;
 735        my $summary = shift;
 736        my $newrevid;
 737        my $path = "File:" . $complete_file_name;
 738        my %hashFiles = get_allowed_file_extensions();
 739        if (!exists($hashFiles{$extension})) {
 740                print STDERR "$complete_file_name is not a permitted file on this wiki.\n";
 741                print STDERR "Check the configuration of file uploads in your mediawiki.\n";
 742                return $newrevid;
 743        }
 744        # Deleting and uploading a file requires a priviledged user
 745        if ($file_deleted) {
 746                mw_connect_maybe();
 747                my $query = {
 748                        action => 'delete',
 749                        title => $path,
 750                        reason => $summary
 751                };
 752                if (!$mediawiki->edit($query)) {
 753                        print STDERR "Failed to delete file on remote wiki\n";
 754                        print STDERR "Check your permissions on the remote site. Error code:\n";
 755                        print STDERR $mediawiki->{error}->{code} . ':' . $mediawiki->{error}->{details};
 756                        exit 1;
 757                }
 758        } else {
 759                # Don't let perl try to interpret file content as UTF-8 => use "raw"
 760                my $content = run_git("cat-file blob $new_sha1", "raw");
 761                if ($content ne "") {
 762                        mw_connect_maybe();
 763                        $mediawiki->{config}->{upload_url} =
 764                                "$url/index.php/Special:Upload";
 765                        $mediawiki->edit({
 766                                action => 'upload',
 767                                filename => $complete_file_name,
 768                                comment => $summary,
 769                                file => [undef,
 770                                         $complete_file_name,
 771                                         Content => $content],
 772                                ignorewarnings => 1,
 773                        }, {
 774                                skip_encoding => 1
 775                        } ) || die $mediawiki->{error}->{code} . ':'
 776                                 . $mediawiki->{error}->{details};
 777                        my $last_file_page = $mediawiki->get_page({title => $path});
 778                        $newrevid = $last_file_page->{revid};
 779                        print STDERR "Pushed file: $new_sha1 - $complete_file_name.\n";
 780                } else {
 781                        print STDERR "Empty file $complete_file_name not pushed.\n";
 782                }
 783        }
 784        return $newrevid;
 785}
 786
 787sub mw_push_file {
 788        my $diff_info = shift;
 789        # $diff_info contains a string in this format:
 790        # 100644 100644 <sha1_of_blob_before_commit> <sha1_of_blob_now> <status>
 791        my @diff_info_split = split(/[ \t]/, $diff_info);
 792
 793        # Filename, including .mw extension
 794        my $complete_file_name = shift;
 795        # Commit message
 796        my $summary = shift;
 797        # MediaWiki revision number. Keep the previous one by default,
 798        # in case there's no edit to perform.
 799        my $oldrevid = shift;
 800        my $newrevid;
 801
 802        my $new_sha1 = $diff_info_split[3];
 803        my $old_sha1 = $diff_info_split[2];
 804        my $page_created = ($old_sha1 eq NULL_SHA1);
 805        my $page_deleted = ($new_sha1 eq NULL_SHA1);
 806        $complete_file_name = mediawiki_clean_filename($complete_file_name);
 807
 808        my ($title, $extension) = $complete_file_name =~ /^(.*)\.([^\.]*)$/;
 809        if (!defined($extension)) {
 810                $extension = "";
 811        }
 812        if ($extension eq "mw") {
 813                my $file_content;
 814                if ($page_deleted) {
 815                        # Deleting a page usually requires
 816                        # special priviledges. A common
 817                        # convention is to replace the page
 818                        # with this content instead:
 819                        $file_content = DELETED_CONTENT;
 820                } else {
 821                        $file_content = run_git("cat-file blob $new_sha1");
 822                }
 823
 824                mw_connect_maybe();
 825
 826                my $result = $mediawiki->edit( {
 827                        action => 'edit',
 828                        summary => $summary,
 829                        title => $title,
 830                        basetimestamp => $basetimestamps{$oldrevid},
 831                        text => mediawiki_clean($file_content, $page_created),
 832                                  }, {
 833                                          skip_encoding => 1 # Helps with names with accentuated characters
 834                                  });
 835                if (!$result) {
 836                        if ($mediawiki->{error}->{code} == 3) {
 837                                # edit conflicts, considered as non-fast-forward
 838                                print STDERR 'Warning: Error ' .
 839                                    $mediawiki->{error}->{code} .
 840                                    ' from mediwiki: ' . $mediawiki->{error}->{details} .
 841                                    ".\n";
 842                                return ($oldrevid, "non-fast-forward");
 843                        } else {
 844                                # Other errors. Shouldn't happen => just die()
 845                                die 'Fatal: Error ' .
 846                                    $mediawiki->{error}->{code} .
 847                                    ' from mediwiki: ' . $mediawiki->{error}->{details};
 848                        }
 849                }
 850                $newrevid = $result->{edit}->{newrevid};
 851                print STDERR "Pushed file: $new_sha1 - $title\n";
 852        } else {
 853                $newrevid = mw_upload_file($complete_file_name, $new_sha1,
 854                                           $extension, $page_deleted,
 855                                           $summary);
 856        }
 857        $newrevid = ($newrevid or $oldrevid);
 858        return ($newrevid, "ok");
 859}
 860
 861sub mw_push {
 862        # multiple push statements can follow each other
 863        my @refsspecs = (shift, get_more_refs("push"));
 864        my $pushed;
 865        for my $refspec (@refsspecs) {
 866                my ($force, $local, $remote) = $refspec =~ /^(\+)?([^:]*):([^:]*)$/
 867                    or die("Invalid refspec for push. Expected <src>:<dst> or +<src>:<dst>");
 868                if ($force) {
 869                        print STDERR "Warning: forced push not allowed on a MediaWiki.\n";
 870                }
 871                if ($local eq "") {
 872                        print STDERR "Cannot delete remote branch on a MediaWiki\n";
 873                        print STDOUT "error $remote cannot delete\n";
 874                        next;
 875                }
 876                if ($remote ne "refs/heads/master") {
 877                        print STDERR "Only push to the branch 'master' is supported on a MediaWiki\n";
 878                        print STDOUT "error $remote only master allowed\n";
 879                        next;
 880                }
 881                if (mw_push_revision($local, $remote)) {
 882                        $pushed = 1;
 883                }
 884        }
 885
 886        # Notify Git that the push is done
 887        print STDOUT "\n";
 888
 889        if ($pushed && $dumb_push) {
 890                print STDERR "Just pushed some revisions to MediaWiki.\n";
 891                print STDERR "The pushed revisions now have to be re-imported, and your current branch\n";
 892                print STDERR "needs to be updated with these re-imported commits. You can do this with\n";
 893                print STDERR "\n";
 894                print STDERR "  git pull --rebase\n";
 895                print STDERR "\n";
 896        }
 897}
 898
 899sub mw_push_revision {
 900        my $local = shift;
 901        my $remote = shift; # actually, this has to be "refs/heads/master" at this point.
 902        my $last_local_revid = get_last_local_revision();
 903        print STDERR ".\n"; # Finish sentence started by get_last_local_revision()
 904        my $last_remote_revid = get_last_remote_revision();
 905        my $mw_revision = $last_remote_revid;
 906
 907        # Get sha1 of commit pointed by local HEAD
 908        my $HEAD_sha1 = run_git("rev-parse $local 2>/dev/null"); chomp($HEAD_sha1);
 909        # Get sha1 of commit pointed by remotes/$remotename/master
 910        my $remoteorigin_sha1 = run_git("rev-parse refs/remotes/$remotename/master 2>/dev/null");
 911        chomp($remoteorigin_sha1);
 912
 913        if ($last_local_revid > 0 &&
 914            $last_local_revid < $last_remote_revid) {
 915                return error_non_fast_forward($remote);
 916        }
 917
 918        if ($HEAD_sha1 eq $remoteorigin_sha1) {
 919                # nothing to push
 920                return 0;
 921        }
 922
 923        # Get every commit in between HEAD and refs/remotes/origin/master,
 924        # including HEAD and refs/remotes/origin/master
 925        my @commit_pairs = ();
 926        if ($last_local_revid > 0) {
 927                my $parsed_sha1 = $remoteorigin_sha1;
 928                # Find a path from last MediaWiki commit to pushed commit
 929                while ($parsed_sha1 ne $HEAD_sha1) {
 930                        my @commit_info =  grep(/^$parsed_sha1/, split(/\n/, run_git("rev-list --children $local")));
 931                        if (!@commit_info) {
 932                                return error_non_fast_forward($remote);
 933                        }
 934                        my @commit_info_split = split(/ |\n/, $commit_info[0]);
 935                        # $commit_info_split[1] is the sha1 of the commit to export
 936                        # $commit_info_split[0] is the sha1 of its direct child
 937                        push(@commit_pairs, \@commit_info_split);
 938                        $parsed_sha1 = $commit_info_split[1];
 939                }
 940        } else {
 941                # No remote mediawiki revision. Export the whole
 942                # history (linearized with --first-parent)
 943                print STDERR "Warning: no common ancestor, pushing complete history\n";
 944                my $history = run_git("rev-list --first-parent --children $local");
 945                my @history = split('\n', $history);
 946                @history = @history[1..$#history];
 947                foreach my $line (reverse @history) {
 948                        my @commit_info_split = split(/ |\n/, $line);
 949                        push(@commit_pairs, \@commit_info_split);
 950                }
 951        }
 952
 953        foreach my $commit_info_split (@commit_pairs) {
 954                my $sha1_child = @{$commit_info_split}[0];
 955                my $sha1_commit = @{$commit_info_split}[1];
 956                my $diff_infos = run_git("diff-tree -r --raw -z $sha1_child $sha1_commit");
 957                # TODO: we could detect rename, and encode them with a #redirect on the wiki.
 958                # TODO: for now, it's just a delete+add
 959                my @diff_info_list = split(/\0/, $diff_infos);
 960                # Keep the subject line of the commit message as mediawiki comment for the revision
 961                my $commit_msg = run_git("log --no-walk --format=\"%s\" $sha1_commit");
 962                chomp($commit_msg);
 963                # Push every blob
 964                while (@diff_info_list) {
 965                        my $status;
 966                        # git diff-tree -z gives an output like
 967                        # <metadata>\0<filename1>\0
 968                        # <metadata>\0<filename2>\0
 969                        # and we've split on \0.
 970                        my $info = shift(@diff_info_list);
 971                        my $file = shift(@diff_info_list);
 972                        ($mw_revision, $status) = mw_push_file($info, $file, $commit_msg, $mw_revision);
 973                        if ($status eq "non-fast-forward") {
 974                                # we may already have sent part of the
 975                                # commit to MediaWiki, but it's too
 976                                # late to cancel it. Stop the push in
 977                                # the middle, but still give an
 978                                # accurate error message.
 979                                return error_non_fast_forward($remote);
 980                        }
 981                        if ($status ne "ok") {
 982                                die("Unknown error from mw_push_file()");
 983                        }
 984                }
 985                unless ($dumb_push) {
 986                        run_git("notes --ref=$remotename/mediawiki add -m \"mediawiki_revision: $mw_revision\" $sha1_commit");
 987                        run_git("update-ref -m \"Git-MediaWiki push\" refs/mediawiki/$remotename/master $sha1_commit $sha1_child");
 988                }
 989        }
 990
 991        print STDOUT "ok $remote\n";
 992        return 1;
 993}
 994
 995sub get_allowed_file_extensions {
 996        mw_connect_maybe();
 997
 998        my $query = {
 999                action => 'query',
1000                meta => 'siteinfo',
1001                siprop => 'fileextensions'
1002                };
1003        my $result = $mediawiki->api($query);
1004        my @file_extensions= map $_->{ext},@{$result->{query}->{fileextensions}};
1005        my %hashFile = map {$_ => 1}@file_extensions;
1006
1007        return %hashFile;
1008}