git-remote-mediawiki: send "File:" attachments to a remote wiki
[git/git.git] / contrib / mw-to-git / git-remote-mediawiki
CommitLineData
428c995c
JN
1#! /usr/bin/perl
2
3# Copyright (C) 2011
4# Jérémie Nikaes <jeremie.nikaes@ensimag.imag.fr>
5# Arnaud Lacurie <arnaud.lacurie@ensimag.imag.fr>
6# Claire Fousse <claire.fousse@ensimag.imag.fr>
7# David Amouyal <david.amouyal@ensimag.imag.fr>
8# Matthieu Moy <matthieu.moy@grenoble-inp.fr>
9# License: GPL v2 or later
10
11# Gateway between Git and MediaWiki.
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;
428c995c 39
721a533f 40# By default, use UTF-8 to communicate with Git and the user
428c995c 41binmode STDERR, ":utf8";
721a533f 42binmode STDOUT, ":utf8";
428c995c
JN
43
44use URI::Escape;
a7271ad1
MM
45use IPC::Open2;
46
428c995c
JN
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");
1d6abac9 80my $wiki_domain = run_git("config --get remote.". $remotename .".mwDomain");
428c995c
JN
81chomp($wiki_login);
82chomp($wiki_passwd);
1d6abac9 83chomp($wiki_domain);
428c995c
JN
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
93f0d338
MM
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
428c995c
JN
111my $wiki_name = $url;
112$wiki_name =~ s/[^\/]*:\/\///;
9fb79503
MM
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/^.*@//;
428c995c
JN
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
a7271ad1
MM
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
428c995c
JN
230# MediaWiki API instance, created lazily.
231my $mediawiki;
232
233sub mw_connect_maybe {
234 if ($mediawiki) {
a7271ad1 235 return;
428c995c
JN
236 }
237 $mediawiki = MediaWiki::API->new;
238 $mediawiki->{config}->{api_url} = "$url/api.php";
239 if ($wiki_login) {
a7271ad1
MM
240 my %credential = credential_from_url($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";
428c995c 250 } else {
a7271ad1
MM
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;
428c995c
JN
257 }
258 }
259}
260
261sub 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
289sub 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
b3d98595
NKT
351# usage: $out = run_git("command args");
352# $out = run_git("command args", "raw"); # don't interpret output as UTF-8.
428c995c 353sub run_git {
b3d98595
NKT
354 my $args = shift;
355 my $encoding = (shift || "encoding(UTF-8)");
356 open(my $git, "-|:$encoding", "git " . $args);
428c995c
JN
357 my $res = do { local $/; <$git> };
358 close($git);
359
360 return $res;
361}
362
363
364sub get_last_local_revision {
365 # Get note regarding last mediawiki revision
366 my $note = run_git("notes --ref=$remotename/mediawiki show refs/mediawiki/$remotename/master 2>/dev/null");
367 my @note_info = split(/ /, $note);
368
369 my $lastrevision_number;
370 if (!(defined($note_info[0]) && $note_info[0] eq "mediawiki_revision:")) {
371 print STDERR "No previous mediawiki revision found";
372 $lastrevision_number = 0;
373 } else {
374 # Notes are formatted : mediawiki_revision: #number
375 $lastrevision_number = $note_info[1];
376 chomp($lastrevision_number);
377 print STDERR "Last local mediawiki revision found is $lastrevision_number";
378 }
379 return $lastrevision_number;
380}
381
3c1ed90e
MM
382# Remember the timestamp corresponding to a revision id.
383my %basetimestamps;
384
428c995c
JN
385sub get_last_remote_revision {
386 mw_connect_maybe();
387
388 my @pages = get_mw_pages();
389
390 my $max_rev_num = 0;
391
392 foreach my $page (@pages) {
393 my $id = $page->{pageid};
394
395 my $query = {
396 action => 'query',
397 prop => 'revisions',
3c1ed90e 398 rvprop => 'ids|timestamp',
428c995c
JN
399 pageids => $id,
400 };
401
402 my $result = $mediawiki->api($query);
403
404 my $lastrev = pop(@{$result->{query}->{pages}->{$id}->{revisions}});
405
3c1ed90e
MM
406 $basetimestamps{$lastrev->{revid}} = $lastrev->{timestamp};
407
428c995c
JN
408 $max_rev_num = ($lastrev->{revid} > $max_rev_num ? $lastrev->{revid} : $max_rev_num);
409 }
410
411 print STDERR "Last remote revision found is $max_rev_num.\n";
412 return $max_rev_num;
413}
414
415# Clean content before sending it to MediaWiki
416sub mediawiki_clean {
417 my $string = shift;
418 my $page_created = shift;
419 # Mediawiki does not allow blank space at the end of a page and ends with a single \n.
420 # This function right trims a string and adds a \n at the end to follow this rule
421 $string =~ s/\s+$//;
422 if ($string eq "" && $page_created) {
423 # Creating empty pages is forbidden.
424 $string = EMPTY_CONTENT;
425 }
426 return $string."\n";
427}
428
429# Filter applied on MediaWiki data before adding them to Git
430sub mediawiki_smudge {
431 my $string = shift;
432 if ($string eq EMPTY_CONTENT) {
433 $string = "";
434 }
435 # This \n is important. This is due to mediawiki's way to handle end of files.
436 return $string."\n";
437}
438
439sub mediawiki_clean_filename {
440 my $filename = shift;
441 $filename =~ s/@{[SLASH_REPLACEMENT]}/\//g;
442 # [, ], |, {, and } are forbidden by MediaWiki, even URL-encoded.
443 # Do a variant of URL-encoding, i.e. looks like URL-encoding,
444 # but with _ added to prevent MediaWiki from thinking this is
445 # an actual special character.
446 $filename =~ s/[\[\]\{\}\|]/sprintf("_%%_%x", ord($&))/ge;
447 # If we use the uri escape before
448 # we should unescape here, before anything
449
450 return $filename;
451}
452
453sub mediawiki_smudge_filename {
454 my $filename = shift;
455 $filename =~ s/\//@{[SLASH_REPLACEMENT]}/g;
456 $filename =~ s/ /_/g;
457 # Decode forbidden characters encoded in mediawiki_clean_filename
458 $filename =~ s/_%_([0-9a-fA-F][0-9a-fA-F])/sprintf("%c", hex($1))/ge;
459 return $filename;
460}
461
462sub literal_data {
463 my ($content) = @_;
464 print STDOUT "data ", bytes::length($content), "\n", $content;
465}
466
467sub mw_capabilities {
468 # Revisions are imported to the private namespace
469 # refs/mediawiki/$remotename/ by the helper and fetched into
470 # refs/remotes/$remotename later by fetch.
471 print STDOUT "refspec refs/heads/*:refs/mediawiki/$remotename/*\n";
472 print STDOUT "import\n";
473 print STDOUT "list\n";
474 print STDOUT "push\n";
475 print STDOUT "\n";
476}
477
478sub mw_list {
479 # MediaWiki do not have branches, we consider one branch arbitrarily
480 # called master, and HEAD pointing to it.
481 print STDOUT "? refs/heads/master\n";
482 print STDOUT "\@refs/heads/master HEAD\n";
483 print STDOUT "\n";
484}
485
486sub mw_option {
487 print STDERR "remote-helper command 'option $_[0]' not yet implemented\n";
488 print STDOUT "unsupported\n";
489}
490
491sub fetch_mw_revisions_for_page {
492 my $page = shift;
493 my $id = shift;
494 my $fetch_from = shift;
495 my @page_revs = ();
496 my $query = {
497 action => 'query',
498 prop => 'revisions',
499 rvprop => 'ids',
500 rvdir => 'newer',
501 rvstartid => $fetch_from,
502 rvlimit => 500,
503 pageids => $id,
504 };
505
506 my $revnum = 0;
507 # Get 500 revisions at a time due to the mediawiki api limit
508 while (1) {
509 my $result = $mediawiki->api($query);
510
511 # Parse each of those 500 revisions
512 foreach my $revision (@{$result->{query}->{pages}->{$id}->{revisions}}) {
513 my $page_rev_ids;
514 $page_rev_ids->{pageid} = $page->{pageid};
515 $page_rev_ids->{revid} = $revision->{revid};
516 push(@page_revs, $page_rev_ids);
517 $revnum++;
518 }
519 last unless $result->{'query-continue'};
520 $query->{rvstartid} = $result->{'query-continue'}->{revisions}->{rvstartid};
521 }
522 if ($shallow_import && @page_revs) {
523 print STDERR " Found 1 revision (shallow import).\n";
524 @page_revs = sort {$b->{revid} <=> $a->{revid}} (@page_revs);
525 return $page_revs[0];
526 }
527 print STDERR " Found ", $revnum, " revision(s).\n";
528 return @page_revs;
529}
530
531sub fetch_mw_revisions {
532 my $pages = shift; my @pages = @{$pages};
533 my $fetch_from = shift;
534
535 my @revisions = ();
536 my $n = 1;
537 foreach my $page (@pages) {
538 my $id = $page->{pageid};
539
540 print STDERR "page $n/", scalar(@pages), ": ". $page->{title} ."\n";
541 $n++;
542 my @page_revs = fetch_mw_revisions_for_page($page, $id, $fetch_from);
543 @revisions = (@page_revs, @revisions);
544 }
545
546 return ($n, @revisions);
547}
548
549sub import_file_revision {
550 my $commit = shift;
551 my %commit = %{$commit};
552 my $full_import = shift;
553 my $n = shift;
554
555 my $title = $commit{title};
556 my $comment = $commit{comment};
557 my $content = $commit{content};
558 my $author = $commit{author};
559 my $date = $commit{date};
560
561 print STDOUT "commit refs/mediawiki/$remotename/master\n";
562 print STDOUT "mark :$n\n";
563 print STDOUT "committer $author <$author\@$wiki_name> ", $date->epoch, " +0000\n";
564 literal_data($comment);
565
566 # If it's not a clone, we need to know where to start from
567 if (!$full_import && $n == 1) {
568 print STDOUT "from refs/mediawiki/$remotename/master^0\n";
569 }
570 if ($content ne DELETED_CONTENT) {
571 print STDOUT "M 644 inline $title.mw\n";
572 literal_data($content);
573 print STDOUT "\n\n";
574 } else {
575 print STDOUT "D $title.mw\n";
576 }
577
578 # mediawiki revision number in the git note
579 if ($full_import && $n == 1) {
580 print STDOUT "reset refs/notes/$remotename/mediawiki\n";
581 }
582 print STDOUT "commit refs/notes/$remotename/mediawiki\n";
583 print STDOUT "committer $author <$author\@$wiki_name> ", $date->epoch, " +0000\n";
584 literal_data("Note added by git-mediawiki during import");
585 if (!$full_import && $n == 1) {
586 print STDOUT "from refs/notes/$remotename/mediawiki^0\n";
587 }
588 print STDOUT "N inline :$n\n";
589 literal_data("mediawiki_revision: " . $commit{mw_revision});
590 print STDOUT "\n\n";
591}
592
593# parse a sequence of
594# <cmd> <arg1>
595# <cmd> <arg2>
596# \n
597# (like batch sequence of import and sequence of push statements)
598sub get_more_refs {
599 my $cmd = shift;
600 my @refs;
601 while (1) {
602 my $line = <STDIN>;
603 if ($line =~ m/^$cmd (.*)$/) {
604 push(@refs, $1);
605 } elsif ($line eq "\n") {
606 return @refs;
607 } else {
608 die("Invalid command in a '$cmd' batch: ". $_);
609 }
610 }
611}
612
613sub mw_import {
614 # multiple import commands can follow each other.
615 my @refs = (shift, get_more_refs("import"));
616 foreach my $ref (@refs) {
617 mw_import_ref($ref);
618 }
619 print STDOUT "done\n";
620}
621
622sub mw_import_ref {
623 my $ref = shift;
624 # The remote helper will call "import HEAD" and
625 # "import refs/heads/master".
626 # Since HEAD is a symbolic ref to master (by convention,
627 # followed by the output of the command "list" that we gave),
628 # we don't need to do anything in this case.
629 if ($ref eq "HEAD") {
630 return;
631 }
632
633 mw_connect_maybe();
634
635 my @pages = get_mw_pages();
636
637 print STDERR "Searching revisions...\n";
638 my $last_local = get_last_local_revision();
639 my $fetch_from = $last_local + 1;
640 if ($fetch_from == 1) {
641 print STDERR ", fetching from beginning.\n";
642 } else {
643 print STDERR ", fetching from here.\n";
644 }
645 my ($n, @revisions) = fetch_mw_revisions(\@pages, $fetch_from);
646
647 # Creation of the fast-import stream
648 print STDERR "Fetching & writing export data...\n";
649
650 $n = 0;
651 my $last_timestamp = 0; # Placeholer in case $rev->timestamp is undefined
652
653 foreach my $pagerevid (sort {$a->{revid} <=> $b->{revid}} @revisions) {
654 # fetch the content of the pages
655 my $query = {
656 action => 'query',
657 prop => 'revisions',
658 rvprop => 'content|timestamp|comment|user|ids',
659 revids => $pagerevid->{revid},
660 };
661
662 my $result = $mediawiki->api($query);
663
664 my $rev = pop(@{$result->{query}->{pages}->{$pagerevid->{pageid}}->{revisions}});
665
666 $n++;
667
668 my %commit;
669 $commit{author} = $rev->{user} || 'Anonymous';
670 $commit{comment} = $rev->{comment} || '*Empty MediaWiki Message*';
671 $commit{title} = mediawiki_smudge_filename(
672 $result->{query}->{pages}->{$pagerevid->{pageid}}->{title}
673 );
674 $commit{mw_revision} = $pagerevid->{revid};
675 $commit{content} = mediawiki_smudge($rev->{'*'});
676
677 if (!defined($rev->{timestamp})) {
678 $last_timestamp++;
679 } else {
680 $last_timestamp = $rev->{timestamp};
681 }
682 $commit{date} = DateTime::Format::ISO8601->parse_datetime($last_timestamp);
683
684 print STDERR "$n/", scalar(@revisions), ": Revision #$pagerevid->{revid} of $commit{title}\n";
685
686 import_file_revision(\%commit, ($fetch_from == 1), $n);
687 }
688
689 if ($fetch_from == 1 && $n == 0) {
690 print STDERR "You appear to have cloned an empty MediaWiki.\n";
691 # Something has to be done remote-helper side. If nothing is done, an error is
692 # thrown saying that HEAD is refering to unknown object 0000000000000000000
693 # and the clone fails.
694 }
695}
696
697sub error_non_fast_forward {
fd47d7b9
MM
698 my $advice = run_git("config --bool advice.pushNonFastForward");
699 chomp($advice);
700 if ($advice ne "false") {
701 # Native git-push would show this after the summary.
702 # We can't ask it to display it cleanly, so print it
703 # ourselves before.
704 print STDERR "To prevent you from losing history, non-fast-forward updates were rejected\n";
705 print STDERR "Merge the remote changes (e.g. 'git pull') before pushing again. See the\n";
706 print STDERR "'Note about fast-forwards' section of 'git push --help' for details.\n";
707 }
428c995c
JN
708 print STDOUT "error $_[0] \"non-fast-forward\"\n";
709 return 0;
710}
711
b3d98595
NKT
712sub mw_upload_file {
713 my $complete_file_name = shift;
714 my $new_sha1 = shift;
715 my $extension = shift;
716 my $file_deleted = shift;
717 my $summary = shift;
718 my $newrevid;
719 my $path = "File:" . $complete_file_name;
720 my %hashFiles = get_allowed_file_extensions();
721 if (!exists($hashFiles{$extension})) {
722 print STDERR "$complete_file_name is not a permitted file on this wiki.\n";
723 print STDERR "Check the configuration of file uploads in your mediawiki.\n";
724 return $newrevid;
725 }
726 # Deleting and uploading a file requires a priviledged user
727 if ($file_deleted) {
728 mw_connect_maybe();
729 my $query = {
730 action => 'delete',
731 title => $path,
732 reason => $summary
733 };
734 if (!$mediawiki->edit($query)) {
735 print STDERR "Failed to delete file on remote wiki\n";
736 print STDERR "Check your permissions on the remote site. Error code:\n";
737 print STDERR $mediawiki->{error}->{code} . ':' . $mediawiki->{error}->{details};
738 exit 1;
739 }
740 } else {
741 # Don't let perl try to interpret file content as UTF-8 => use "raw"
742 my $content = run_git("cat-file blob $new_sha1", "raw");
743 if ($content ne "") {
744 mw_connect_maybe();
745 $mediawiki->{config}->{upload_url} =
746 "$url/index.php/Special:Upload";
747 $mediawiki->edit({
748 action => 'upload',
749 filename => $complete_file_name,
750 comment => $summary,
751 file => [undef,
752 $complete_file_name,
753 Content => $content],
754 ignorewarnings => 1,
755 }, {
756 skip_encoding => 1
757 } ) || die $mediawiki->{error}->{code} . ':'
758 . $mediawiki->{error}->{details};
759 my $last_file_page = $mediawiki->get_page({title => $path});
760 $newrevid = $last_file_page->{revid};
761 print STDERR "Pushed file: $new_sha1 - $complete_file_name.\n";
762 } else {
763 print STDERR "Empty file $complete_file_name not pushed.\n";
764 }
765 }
766 return $newrevid;
767}
768
428c995c
JN
769sub mw_push_file {
770 my $diff_info = shift;
771 # $diff_info contains a string in this format:
772 # 100644 100644 <sha1_of_blob_before_commit> <sha1_of_blob_now> <status>
773 my @diff_info_split = split(/[ \t]/, $diff_info);
774
775 # Filename, including .mw extension
776 my $complete_file_name = shift;
777 # Commit message
778 my $summary = shift;
93f0d338
MM
779 # MediaWiki revision number. Keep the previous one by default,
780 # in case there's no edit to perform.
b3d98595
NKT
781 my $oldrevid = shift;
782 my $newrevid;
428c995c
JN
783
784 my $new_sha1 = $diff_info_split[3];
785 my $old_sha1 = $diff_info_split[2];
786 my $page_created = ($old_sha1 eq NULL_SHA1);
787 my $page_deleted = ($new_sha1 eq NULL_SHA1);
788 $complete_file_name = mediawiki_clean_filename($complete_file_name);
789
b3d98595
NKT
790 my ($title, $extension) = $complete_file_name =~ /^(.*)\.([^\.]*)$/;
791 if (!defined($extension)) {
792 $extension = "";
793 }
794 if ($extension eq "mw") {
428c995c
JN
795 my $file_content;
796 if ($page_deleted) {
797 # Deleting a page usually requires
798 # special priviledges. A common
799 # convention is to replace the page
800 # with this content instead:
801 $file_content = DELETED_CONTENT;
802 } else {
803 $file_content = run_git("cat-file blob $new_sha1");
804 }
805
806 mw_connect_maybe();
807
808 my $result = $mediawiki->edit( {
809 action => 'edit',
810 summary => $summary,
811 title => $title,
b3d98595 812 basetimestamp => $basetimestamps{$oldrevid},
428c995c
JN
813 text => mediawiki_clean($file_content, $page_created),
814 }, {
815 skip_encoding => 1 # Helps with names with accentuated characters
3c1ed90e
MM
816 });
817 if (!$result) {
818 if ($mediawiki->{error}->{code} == 3) {
819 # edit conflicts, considered as non-fast-forward
820 print STDERR 'Warning: Error ' .
821 $mediawiki->{error}->{code} .
822 ' from mediwiki: ' . $mediawiki->{error}->{details} .
823 ".\n";
b3d98595 824 return ($oldrevid, "non-fast-forward");
3c1ed90e
MM
825 } else {
826 # Other errors. Shouldn't happen => just die()
827 die 'Fatal: Error ' .
828 $mediawiki->{error}->{code} .
829 ' from mediwiki: ' . $mediawiki->{error}->{details};
830 }
831 }
93f0d338 832 $newrevid = $result->{edit}->{newrevid};
ac86ec0f 833 print STDERR "Pushed file: $new_sha1 - $title\n";
428c995c 834 } else {
b3d98595
NKT
835 $newrevid = mw_upload_file($complete_file_name, $new_sha1,
836 $extension, $page_deleted,
837 $summary);
428c995c 838 }
b3d98595 839 $newrevid = ($newrevid or $oldrevid);
3c1ed90e 840 return ($newrevid, "ok");
428c995c
JN
841}
842
843sub mw_push {
844 # multiple push statements can follow each other
845 my @refsspecs = (shift, get_more_refs("push"));
428c995c
JN
846 my $pushed;
847 for my $refspec (@refsspecs) {
848 my ($force, $local, $remote) = $refspec =~ /^(\+)?([^:]*):([^:]*)$/
849 or die("Invalid refspec for push. Expected <src>:<dst> or +<src>:<dst>");
850 if ($force) {
851 print STDERR "Warning: forced push not allowed on a MediaWiki.\n";
852 }
853 if ($local eq "") {
854 print STDERR "Cannot delete remote branch on a MediaWiki\n";
855 print STDOUT "error $remote cannot delete\n";
856 next;
857 }
858 if ($remote ne "refs/heads/master") {
859 print STDERR "Only push to the branch 'master' is supported on a MediaWiki\n";
860 print STDOUT "error $remote only master allowed\n";
861 next;
862 }
863 if (mw_push_revision($local, $remote)) {
864 $pushed = 1;
865 }
866 }
867
868 # Notify Git that the push is done
869 print STDOUT "\n";
870
93f0d338 871 if ($pushed && $dumb_push) {
428c995c
JN
872 print STDERR "Just pushed some revisions to MediaWiki.\n";
873 print STDERR "The pushed revisions now have to be re-imported, and your current branch\n";
874 print STDERR "needs to be updated with these re-imported commits. You can do this with\n";
875 print STDERR "\n";
876 print STDERR " git pull --rebase\n";
877 print STDERR "\n";
878 }
879}
880
881sub mw_push_revision {
882 my $local = shift;
883 my $remote = shift; # actually, this has to be "refs/heads/master" at this point.
884 my $last_local_revid = get_last_local_revision();
885 print STDERR ".\n"; # Finish sentence started by get_last_local_revision()
886 my $last_remote_revid = get_last_remote_revision();
93f0d338 887 my $mw_revision = $last_remote_revid;
428c995c
JN
888
889 # Get sha1 of commit pointed by local HEAD
890 my $HEAD_sha1 = run_git("rev-parse $local 2>/dev/null"); chomp($HEAD_sha1);
891 # Get sha1 of commit pointed by remotes/$remotename/master
892 my $remoteorigin_sha1 = run_git("rev-parse refs/remotes/$remotename/master 2>/dev/null");
893 chomp($remoteorigin_sha1);
894
895 if ($last_local_revid > 0 &&
896 $last_local_revid < $last_remote_revid) {
897 return error_non_fast_forward($remote);
898 }
899
900 if ($HEAD_sha1 eq $remoteorigin_sha1) {
901 # nothing to push
902 return 0;
903 }
904
905 # Get every commit in between HEAD and refs/remotes/origin/master,
906 # including HEAD and refs/remotes/origin/master
907 my @commit_pairs = ();
908 if ($last_local_revid > 0) {
909 my $parsed_sha1 = $remoteorigin_sha1;
910 # Find a path from last MediaWiki commit to pushed commit
911 while ($parsed_sha1 ne $HEAD_sha1) {
912 my @commit_info = grep(/^$parsed_sha1/, split(/\n/, run_git("rev-list --children $local")));
913 if (!@commit_info) {
914 return error_non_fast_forward($remote);
915 }
916 my @commit_info_split = split(/ |\n/, $commit_info[0]);
917 # $commit_info_split[1] is the sha1 of the commit to export
918 # $commit_info_split[0] is the sha1 of its direct child
919 push(@commit_pairs, \@commit_info_split);
920 $parsed_sha1 = $commit_info_split[1];
921 }
922 } else {
923 # No remote mediawiki revision. Export the whole
924 # history (linearized with --first-parent)
925 print STDERR "Warning: no common ancestor, pushing complete history\n";
926 my $history = run_git("rev-list --first-parent --children $local");
927 my @history = split('\n', $history);
928 @history = @history[1..$#history];
929 foreach my $line (reverse @history) {
930 my @commit_info_split = split(/ |\n/, $line);
931 push(@commit_pairs, \@commit_info_split);
932 }
933 }
934
935 foreach my $commit_info_split (@commit_pairs) {
936 my $sha1_child = @{$commit_info_split}[0];
937 my $sha1_commit = @{$commit_info_split}[1];
938 my $diff_infos = run_git("diff-tree -r --raw -z $sha1_child $sha1_commit");
939 # TODO: we could detect rename, and encode them with a #redirect on the wiki.
940 # TODO: for now, it's just a delete+add
941 my @diff_info_list = split(/\0/, $diff_infos);
28c24bd7
MM
942 # Keep the subject line of the commit message as mediawiki comment for the revision
943 my $commit_msg = run_git("log --no-walk --format=\"%s\" $sha1_commit");
428c995c
JN
944 chomp($commit_msg);
945 # Push every blob
946 while (@diff_info_list) {
3c1ed90e 947 my $status;
428c995c
JN
948 # git diff-tree -z gives an output like
949 # <metadata>\0<filename1>\0
950 # <metadata>\0<filename2>\0
951 # and we've split on \0.
952 my $info = shift(@diff_info_list);
953 my $file = shift(@diff_info_list);
3c1ed90e
MM
954 ($mw_revision, $status) = mw_push_file($info, $file, $commit_msg, $mw_revision);
955 if ($status eq "non-fast-forward") {
956 # we may already have sent part of the
957 # commit to MediaWiki, but it's too
958 # late to cancel it. Stop the push in
959 # the middle, but still give an
960 # accurate error message.
961 return error_non_fast_forward($remote);
962 }
963 if ($status ne "ok") {
964 die("Unknown error from mw_push_file()");
965 }
93f0d338
MM
966 }
967 unless ($dumb_push) {
968 run_git("notes --ref=$remotename/mediawiki add -m \"mediawiki_revision: $mw_revision\" $sha1_commit");
969 run_git("update-ref -m \"Git-MediaWiki push\" refs/mediawiki/$remotename/master $sha1_commit $sha1_child");
428c995c
JN
970 }
971 }
972
973 print STDOUT "ok $remote\n";
974 return 1;
975}
b3d98595
NKT
976
977sub get_allowed_file_extensions {
978 mw_connect_maybe();
979
980 my $query = {
981 action => 'query',
982 meta => 'siteinfo',
983 siprop => 'fileextensions'
984 };
985 my $result = $mediawiki->api($query);
986 my @file_extensions= map $_->{ext},@{$result->{query}->{fileextensions}};
987 my %hashFile = map {$_ => 1}@file_extensions;
988
989 return %hashFile;
990}