X-Git-Url: http://git.vanrenterghem.biz/git.ikiwiki.info.git/blobdiff_plain/1b6c1895784db37438ab51a4aebb366aa2e8009d..e4516078cc7c5e9148a0f079e3d9a71eeccd4155:/IkiWiki/Plugin/git.pm diff --git a/IkiWiki/Plugin/git.pm b/IkiWiki/Plugin/git.pm index 3dd910cd5..2bc250052 100644 --- a/IkiWiki/Plugin/git.pm +++ b/IkiWiki/Plugin/git.pm @@ -5,6 +5,8 @@ use warnings; use strict; use IkiWiki; use Encode; +use File::Path qw{remove_tree}; +use URI::Escape q{uri_escape_utf8}; use open qw{:utf8 :std}; my $sha1_pattern = qr/[0-9a-fA-F]{40}/; # pattern to validate Git sha1sums @@ -28,6 +30,8 @@ sub import { hook(type => "rcs", id => "rcs_receive", call => \&rcs_receive); hook(type => "rcs", id => "rcs_preprevert", call => \&rcs_preprevert); hook(type => "rcs", id => "rcs_revert", call => \&rcs_revert); + hook(type => "rcs", id => "rcs_find_changes", call => \&rcs_find_changes); + hook(type => "rcs", id => "rcs_get_current_rev", call => \&rcs_get_current_rev); } sub checkconfig () { @@ -150,40 +154,65 @@ sub genwrapper { } } -my $git_dir=undef; -my $prefix=undef; +# Loosely based on git-new-workdir from git contrib. +sub create_temp_working_dir ($$) { + my $rootdir = shift; + my $branch = shift; + my $working = "$rootdir/.git/ikiwiki-temp-working"; + remove_tree($working); -sub in_git_dir ($$) { - $git_dir=shift; - my @ret=shift->(); - $git_dir=undef; - $prefix=undef; - return @ret; + foreach my $dir ("", ".git") { + if (!mkdir("$working/$dir")) { + error("Unable to create $working/$dir: $!"); + } + } + + # Hooks are deliberately not included: we will commit to the temporary + # branch that is used in the temporary working tree, and we don't want + # to run the post-commit hook there. + # + # logs/refs is not included because we don't use the reflog. + # remotes, rr-cache, svn are similarly excluded. + foreach my $link ("config", "refs", "objects", "info", "packed-refs") { + if (!symlink("../../$link", "$working/.git/$link")) { + error("Unable to create symlink $working/.git/$link: $!"); + } + } + + open (my $out, '>', "$working/.git/HEAD") or + error("failed to write $working.git/HEAD: $!"); + print $out "ref: refs/heads/$branch\n" or + error("failed to write $working.git/HEAD: $!"); + close $out or + error("failed to write $working.git/HEAD: $!"); + return $working; } -sub safe_git (&@) { +sub safe_git { # Start a child process safely without resorting to /bin/sh. # Returns command output (in list content) or success state # (in scalar context), or runs the specified data handler. - my ($error_handler, $data_handler, @cmdline) = @_; + my %params = @_; my $pid = open my $OUT, "-|"; + error("Working directory not specified") unless defined $params{chdir}; error("Cannot fork: $!") if !defined $pid; if (!$pid) { # In child. # Git commands want to be in wc. - if (! defined $git_dir) { - chdir $config{srcdir} - or error("cannot chdir to $config{srcdir}: $!"); + if ($params{chdir} ne '.') { + chdir $params{chdir} + or error("cannot chdir to $params{chdir}: $!"); } - else { - chdir $git_dir - or error("cannot chdir to $git_dir: $!"); + + if ($params{stdout}) { + open(STDOUT, '>&', $params{stdout}) or error("Cannot reopen stdout: $!"); } - exec @cmdline or error("Cannot exec '@cmdline': $!"); + + exec @{$params{cmdline}} or error("Cannot exec '@{$params{cmdline}}': $!"); } # In parent. @@ -198,25 +227,51 @@ sub safe_git (&@) { chomp; - if (! defined $data_handler) { + if (! defined $params{data_handler}) { push @lines, $_; } else { - last unless $data_handler->($_); + last unless $params{data_handler}->($_); } } close $OUT; - $error_handler->("'@cmdline' failed: $!") if $? && $error_handler; + $params{error_handler}->("'@{$params{cmdline}}' failed: $!") if $? && $params{error_handler}; return wantarray ? @lines : ($? == 0); } # Convenient wrappers. -sub run_or_die ($@) { safe_git(\&error, undef, @_) } -sub run_or_cry ($@) { safe_git(sub { warn @_ }, undef, @_) } -sub run_or_non ($@) { safe_git(undef, undef, @_) } +sub run_or_die_in ($$@) { + my $dir = shift; + safe_git(chdir => $dir, error_handler => \&error, cmdline => \@_); +} +sub run_or_cry_in ($$@) { + my $dir = shift; + safe_git(chdir => $dir, error_handler => sub { warn @_ }, cmdline => \@_); +} +sub run_or_non_in ($$@) { + my $dir = shift; + safe_git(chdir => $dir, cmdline => \@_); +} + +sub ensure_committer ($) { + my $dir = shift; + + if (! length $ENV{GIT_AUTHOR_NAME} || ! length $ENV{GIT_COMMITTER_NAME}) { + my $name = join('', run_or_non_in($dir, "git", "config", "user.name")); + if (! length $name) { + run_or_die_in($dir, "git", "config", "user.name", "IkiWiki"); + } + } + if (! length $ENV{GIT_AUTHOR_EMAIL} || ! length $ENV{GIT_COMMITTER_EMAIL}) { + my $email = join('', run_or_non_in($dir, "git", "config", "user.email")); + if (! length $email) { + run_or_die_in($dir, "git", "config", "user.email", "ikiwiki.info"); + } + } +} sub merge_past ($$$) { # Unlike with Subversion, Git cannot make a 'svn merge -rN:M file'. @@ -255,6 +310,8 @@ sub merge_past ($$$) { my @undo; # undo stack for cleanup in case of an error my $conflict; # file content with conflict markers + ensure_committer($config{srcdir}); + eval { # Hide local changes from Git by renaming the modified file. # Relative paths must be converted to absolute for renaming. @@ -273,30 +330,30 @@ sub merge_past ($$$) { my $branch = "throw_away_${sha1}"; # supposed to be unique # Create a throw-away branch and rewind backward. - push @undo, sub { run_or_cry('git', 'branch', '-D', $branch) }; - run_or_die('git', 'branch', $branch, $sha1); + push @undo, sub { run_or_cry_in($config{srcdir}, 'git', 'branch', '-D', $branch) }; + run_or_die_in($config{srcdir}, 'git', 'branch', $branch, $sha1); # Switch to throw-away branch for the merge operation. push @undo, sub { - if (!run_or_cry('git', 'checkout', $config{gitmaster_branch})) { - run_or_cry('git', 'checkout','-f',$config{gitmaster_branch}); + if (!run_or_cry_in($config{srcdir}, 'git', 'checkout', $config{gitmaster_branch})) { + run_or_cry_in($config{srcdir}, 'git', 'checkout','-f',$config{gitmaster_branch}); } }; - run_or_die('git', 'checkout', $branch); + run_or_die_in($config{srcdir}, 'git', 'checkout', $branch); # Put the modified file in _this_ branch. rename($hidden, $target) or error("rename '$hidden' to '$target' failed: $!"); # _Silently_ commit all modifications in the current branch. - run_or_non('git', 'commit', '-m', $message, '-a'); + run_or_non_in($config{srcdir}, 'git', 'commit', '-m', $message, '-a'); # ... and re-switch to master. - run_or_die('git', 'checkout', $config{gitmaster_branch}); + run_or_die_in($config{srcdir}, 'git', 'checkout', $config{gitmaster_branch}); # Attempt to merge without complaining. - if (!run_or_non('git', 'pull', '--no-commit', '.', $branch)) { + if (!run_or_non_in($config{srcdir}, 'git', 'pull', '--no-commit', '.', $branch)) { $conflict = readfile($target); - run_or_die('git', 'reset', '--hard'); + run_or_die_in($config{srcdir}, 'git', 'reset', '--hard'); } }; my $failure = $@; @@ -312,7 +369,11 @@ sub merge_past ($$$) { return $conflict; } -sub decode_git_file ($) { +{ +my %prefix_cache; + +sub decode_git_file ($$) { + my $dir=shift; my $file=shift; # git does not output utf-8 filenames, but instead @@ -323,25 +384,27 @@ sub decode_git_file ($) { } # strip prefix if in a subdir - if (! defined $prefix) { - ($prefix) = run_or_die('git', 'rev-parse', '--show-prefix'); - if (! defined $prefix) { - $prefix=""; + if (! defined $prefix_cache{$dir}) { + ($prefix_cache{$dir}) = run_or_die_in($dir, 'git', 'rev-parse', '--show-prefix'); + if (! defined $prefix_cache{$dir}) { + $prefix_cache{$dir}=""; } } - $file =~ s/^\Q$prefix\E//; + $file =~ s/^\Q$prefix_cache{$dir}\E//; return decode("utf8", $file); } +} -sub parse_diff_tree ($) { +sub parse_diff_tree ($$) { # Parse the raw diff tree chunk and return the info hash. # See git-diff-tree(1) for the syntax. + my $dir = shift; my $dt_ref = shift; # End of stream? - return if !defined @{ $dt_ref } || - !defined @{ $dt_ref }[0] || !length @{ $dt_ref }[0]; + return if ! @{ $dt_ref } || + !defined $dt_ref->[0] || !length $dt_ref->[0]; my %ci; # Header line. @@ -405,6 +468,17 @@ sub parse_diff_tree ($) { } shift @{ $dt_ref } if $dt_ref->[0] =~ /^$/; + $ci{details} = [parse_changed_files($dir, $dt_ref)]; + + return \%ci; +} + +sub parse_changed_files ($$) { + my $dir = shift; + my $dt_ref = shift; + + my @files; + # Modified files. while (my $line = shift @{ $dt_ref }) { if ($line =~ m{^ @@ -422,8 +496,8 @@ sub parse_diff_tree ($) { my $status = shift(@tmp); if (length $file) { - push @{ $ci{'details'} }, { - 'file' => decode_git_file($file), + push @files, { + 'file' => decode_git_file($dir, $file), 'sha1_from' => $sha1_from[0], 'sha1_to' => $sha1_to, 'mode_from' => $mode_from[0], @@ -436,23 +510,23 @@ sub parse_diff_tree ($) { last; } - return \%ci; + return @files; } -sub git_commit_info ($;$) { +sub git_commit_info ($$;$) { # Return an array of commit info hashes of num commits # starting from the given sha1sum. - my ($sha1, $num) = @_; + my ($dir, $sha1, $num) = @_; my @opts; push @opts, "--max-count=$num" if defined $num; - my @raw_lines = run_or_die('git', 'log', @opts, + my @raw_lines = run_or_die_in($dir, 'git', 'log', @opts, '--pretty=raw', '--raw', '--abbrev=40', '--always', '-c', - '-r', $sha1, '--', '.'); + '-r', $sha1, '--no-renames', '--', '.'); my @ci; - while (my $parsed = parse_diff_tree(\@raw_lines)) { + while (my $parsed = parse_diff_tree($dir, \@raw_lines)) { push @ci, $parsed; } @@ -461,27 +535,73 @@ sub git_commit_info ($;$) { return wantarray ? @ci : $ci[0]; } -sub git_sha1 (;$) { - # Return head sha1sum (of given file). - my $file = shift || q{--}; +sub rcs_find_changes ($) { + my $oldrev=shift; + # Note that git log will sometimes show files being added that + # don't exist. Particularly, git merge -s ours can result in a + # merge commit where some files were not really added. + # This is why the code below verifies that the files really + # exist. + my @raw_lines = run_or_die_in($config{srcdir}, 'git', 'log', + '--pretty=raw', '--raw', '--abbrev=40', '--always', '-c', + '--no-renames', , '--reverse', + '-r', "$oldrev..HEAD", '--', '.'); + + # Due to --reverse, we see changes in chronological order. + my %changed; + my %deleted; + my $nullsha = 0 x 40; + my $newrev=$oldrev; + while (my $ci = parse_diff_tree($config{srcdir}, \@raw_lines)) { + $newrev=$ci->{sha1}; + foreach my $i (@{$ci->{details}}) { + my $file=$i->{file}; + if ($i->{sha1_to} eq $nullsha) { + if (! -e "$config{srcdir}/$file") { + delete $changed{$file}; + $deleted{$file}=1; + } + } + else { + if (-e "$config{srcdir}/$file") { + delete $deleted{$file}; + $changed{$file}=1; + } + } + } + } + + return (\%changed, \%deleted, $newrev); +} + +sub git_sha1_file ($$) { + my $dir=shift; + my $file=shift; + return git_sha1($dir, $file); +} + +sub git_sha1 ($@) { + my $dir = shift; # Ignore error since a non-existing file might be given. - my ($sha1) = run_or_non('git', 'rev-list', '--max-count=1', 'HEAD', - '--', $file); - if ($sha1) { + my ($sha1) = run_or_non_in($dir, 'git', 'rev-list', '--max-count=1', 'HEAD', + '--', @_); + if (defined $sha1) { ($sha1) = $sha1 =~ m/($sha1_pattern)/; # sha1 is untainted now } - else { - debug("Empty sha1sum for '$file'."); - } - return defined $sha1 ? $sha1 : q{}; + return defined $sha1 ? $sha1 : ''; +} + +sub rcs_get_current_rev () { + return git_sha1($config{srcdir}); } sub rcs_update () { # Update working directory. + ensure_committer($config{srcdir}); if (length $config{gitorigin_branch}) { - run_or_cry('git', 'pull', '--prune', $config{gitorigin_branch}); + run_or_cry_in($config{srcdir}, 'git', 'pull', '--prune', $config{gitorigin_branch}); } } @@ -490,7 +610,7 @@ sub rcs_prepedit ($) { # This will be later used in rcs_commit if a merge is required. my ($file) = @_; - return git_sha1($file); + return git_sha1_file($config{srcdir}, $file); } sub rcs_commit (@) { @@ -501,8 +621,11 @@ sub rcs_commit (@) { # Check to see if the page has been changed by someone else since # rcs_prepedit was called. - my $cur = git_sha1($params{file}); - my ($prev) = $params{token} =~ /^($sha1_pattern)$/; # untaint + my $cur = git_sha1_file($config{srcdir}, $params{file}); + my $prev; + if (defined $params{token}) { + ($prev) = $params{token} =~ /^($sha1_pattern)$/; # untaint + } if (defined $cur && defined $prev && $cur ne $prev) { my $conflict = merge_past($prev, $params{file}, $dummy_commit_msg); @@ -532,27 +655,41 @@ sub rcs_commit_helper (@) { elsif (defined $params{session}->remote_addr()) { $u=$params{session}->remote_addr(); } - if (defined $u) { - $u=encode_utf8($u); + if (length $u) { + $u=encode_utf8(IkiWiki::cloak($u)); $ENV{GIT_AUTHOR_NAME}=$u; } + else { + $u = 'anonymous'; + } if (defined $params{session}->param("nickname")) { $u=encode_utf8($params{session}->param("nickname")); $u=~s/\s+/_/g; $u=~s/[^-_0-9[:alnum:]]+//g; } - if (defined $u) { + if (length $u) { $ENV{GIT_AUTHOR_EMAIL}="$u\@web"; } + else { + $ENV{GIT_AUTHOR_EMAIL}='anonymous@web'; + } } + ensure_committer($config{srcdir}); + $params{message} = IkiWiki::possibly_foolish_untaint($params{message}); my @opts; if ($params{message} !~ /\S/) { # Force git to allow empty commit messages. # (If this version of git supports it.) my ($version)=`git --version` =~ /git version (.*)/; - if ($version ge "1.5.4") { + if ($version ge "1.7.8") { + push @opts, "--allow-empty-message", "--no-edit"; + } + if ($version ge "1.7.2") { + push @opts, "--allow-empty-message"; + } + elsif ($version ge "1.5.4") { push @opts, '--cleanup=verbatim'; } else { @@ -563,10 +700,10 @@ sub rcs_commit_helper (@) { push @opts, '--', $params{file}; } # git commit returns non-zero if nothing really changed. - # So we should ignore its exit status (hence run_or_non). - if (run_or_non('git', 'commit', '-m', $params{message}, '-q', @opts)) { + # So we should ignore its exit status (hence run_or_non_in). + if (run_or_non_in($config{srcdir}, 'git', 'commit', '-m', $params{message}, '-q', @opts)) { if (length $config{gitorigin_branch}) { - run_or_cry('git', 'push', $config{gitorigin_branch}); + run_or_cry_in($config{srcdir}, 'git', 'push', $config{gitorigin_branch}, $config{gitmaster_branch}); } } @@ -579,7 +716,8 @@ sub rcs_add ($) { my ($file) = @_; - run_or_cry('git', 'add', $file); + ensure_committer($config{srcdir}); + run_or_cry_in($config{srcdir}, 'git', 'add', '--', $file); } sub rcs_remove ($) { @@ -587,13 +725,15 @@ sub rcs_remove ($) { my ($file) = @_; - run_or_cry('git', 'rm', '-f', $file); + ensure_committer($config{srcdir}); + run_or_cry_in($config{srcdir}, 'git', 'rm', '-f', '--', $file); } sub rcs_rename ($$) { my ($src, $dest) = @_; - run_or_cry('git', 'mv', '-f', $src, $dest); + ensure_committer($config{srcdir}); + run_or_cry_in($config{srcdir}, 'git', 'mv', '-f', '--', $src, $dest); } sub rcs_recentchanges ($) { @@ -605,7 +745,7 @@ sub rcs_recentchanges ($) { error($@) if $@; my @rets; - foreach my $ci (git_commit_info('HEAD', $num || 1)) { + foreach my $ci (git_commit_info($config{srcdir}, 'HEAD', $num || 1)) { # Skip redundant commits. next if ($ci->{'comment'} && @{$ci->{'comment'}}[0] eq $dummy_commit_msg); @@ -617,9 +757,12 @@ sub rcs_recentchanges ($) { my @pages; foreach my $detail (@{ $ci->{'details'} }) { my $file = $detail->{'file'}; + my $efile = join('/', + map { uri_escape_utf8($_) } split('/', $file) + ); my $diffurl = defined $config{'diffurl'} ? $config{'diffurl'} : ""; - $diffurl =~ s/\[\[file\]\]/$file/go; + $diffurl =~ s/\[\[file\]\]/$efile/go; $diffurl =~ s/\[\[sha1_parent\]\]/$ci->{'parent'}/go; $diffurl =~ s/\[\[sha1_from\]\]/$detail->{'sha1_from'}/go; $diffurl =~ s/\[\[sha1_to\]\]/$detail->{'sha1_to'}/go; @@ -688,7 +831,12 @@ sub rcs_diff ($;$) { if (@lines || $line=~/^diff --git/); return 1; }; - safe_git(undef, $addlines, "git", "show", $sha1); + safe_git( + chdir => $config{srcdir}, + error_handler => undef, + data_handler => $addlines, + cmdline => ["git", "show", $sha1], + ); if (wantarray) { return @lines; } @@ -706,7 +854,7 @@ sub findtimes ($$) { if (! keys %time_cache) { my $date; - foreach my $line (run_or_die('git', 'log', + foreach my $line (run_or_die_in($config{srcdir}, 'git', 'log', '--pretty=format:%at', '--name-only', '--relative')) { if (! defined $date && $line =~ /^(\d+)$/) { @@ -716,7 +864,7 @@ sub findtimes ($$) { $date=undef; } else { - my $f=decode_git_file($line); + my $f=decode_git_file($config{srcdir}, $line); if (! $time_cache{$f}) { $time_cache{$f}[0]=$date; # mtime @@ -768,7 +916,8 @@ sub git_find_root { } -sub git_parse_changes { +sub git_parse_changes ($$@) { + my $dir = shift; my $reverted = shift; my @changes = @_; @@ -818,11 +967,12 @@ sub git_parse_changes { die $@ if $@; my $fh; ($fh, $path)=File::Temp::tempfile(undef, UNLINK => 1); - my $cmd = "cd $git_dir && ". - "git show $detail->{sha1_to} > '$path'"; - if (system($cmd) != 0) { - error("failed writing temp file '$path'."); - } + safe_git( + chdir => $dir, + error_handler => sub { error("failed writing temp file '$path': ".shift."."); }, + stdout => $fh, + cmdline => ['git', 'show', $detail->{sha1_to}], + ); } push @rets, { @@ -852,9 +1002,7 @@ sub rcs_receive () { # (Also, if a subdir is involved, we don't want to chdir to # it and only see changes in it.) # The pre-receive hook already puts us in the right place. - in_git_dir(".", sub { - push @rets, git_parse_changes(0, git_commit_info($oldrev."..".$newrev)); - }); + push @rets, git_parse_changes('.', 0, git_commit_info('.', $oldrev."..".$newrev)); } return reverse @rets; @@ -864,12 +1012,17 @@ sub rcs_preprevert ($) { my $rev=shift; my ($sha1) = $rev =~ /^($sha1_pattern)$/; # untaint + my @undo; # undo stack for cleanup in case of an error + # Examine changes from root of git repo, not from any subdir, # in order to see all changes. my ($subdir, $rootdir) = git_find_root(); - in_git_dir($rootdir, sub { - my @commits=git_commit_info($sha1, 1); - + ensure_committer($rootdir); + + # preserve indentation of previous in_git_dir code for now + do { + my @commits=git_commit_info($rootdir, $sha1, 1); + if (! @commits) { error "unknown commit"; # just in case } @@ -880,8 +1033,60 @@ sub rcs_preprevert ($) { error gettext("you are not allowed to revert a merge"); } - git_parse_changes(1, @commits); - }); + # Due to the presence of rename-detection, we cannot actually + # see what will happen in a revert without trying it. + # But we can guess, which is enough to rule out most changes + # that we won't allow reverting. + git_parse_changes($rootdir, 1, @commits); + + my $failure; + my @ret; + eval { + my $branch = "ikiwiki_revert_${sha1}"; # supposed to be unique + + push @undo, sub { + run_or_cry_in($rootdir, 'git', 'branch', '-D', $branch) if $failure; + }; + if (run_or_non_in($rootdir, 'git', 'rev-parse', '--quiet', '--verify', $branch)) { + run_or_non_in($rootdir, 'git', 'branch', '-D', $branch); + } + run_or_die_in($rootdir, 'git', 'branch', $branch, $config{gitmaster_branch}); + + my $working = create_temp_working_dir($rootdir, $branch); + + push @undo, sub { + remove_tree($working); + }; + + run_or_die_in($working, 'git', 'checkout', '--quiet', '--force', $branch); + run_or_die_in($working, 'git', 'revert', '--no-commit', $sha1); + run_or_die_in($working, 'git', 'commit', '-m', "revert $sha1", '-a'); + + my @raw_lines; + @raw_lines = run_or_die_in($rootdir, 'git', 'diff', '--pretty=raw', + '--raw', '--abbrev=40', '--always', '--no-renames', + "..${branch}"); + + my $ci = { + details => [parse_changed_files($rootdir, \@raw_lines)], + }; + + @ret = git_parse_changes($rootdir, 0, $ci); + }; + $failure = $@; + + # Process undo stack (in reverse order). By policy cleanup + # actions should normally print a warning on failure. + while (my $handle = pop @undo) { + $handle->(); + } + + if ($failure) { + my $message = sprintf(gettext("Failed to revert commit %s"), $sha1); + error("$message\n$failure\n"); + } + return @ret; + }; } sub rcs_revert ($) { @@ -889,11 +1094,13 @@ sub rcs_revert ($) { my $rev = shift; my ($sha1) = $rev =~ /^($sha1_pattern)$/; # untaint - if (run_or_non('git', 'revert', '--no-commit', $sha1)) { + ensure_committer($config{srcdir}); + + if (run_or_non_in($config{srcdir}, 'git', 'cherry-pick', '--no-commit', "ikiwiki_revert_$sha1")) { return undef; } else { - run_or_die('git', 'reset', '--hard'); + run_or_non_in($config{srcdir}, 'git', 'branch', '-D', "ikiwiki_revert_$sha1"); return sprintf(gettext("Failed to revert commit %s"), $sha1); } }