1 # Copyright (C) 2016 all contributors <meta@public-inbox.org>
2 # License: AGPL-3.0+ <https://www.gnu.org/licenses/agpl-3.0.txt>
4 # git fast-import-based ssoma-mda MDA replacement
5 # This is only ever run by public-inbox-mda and public-inbox-learn,
6 # not the WWW or NNTP code which only requires read-only access.
7 package PublicInbox::Import;
10 use Fcntl qw(:flock :DEFAULT);
12 use PublicInbox::Spawn qw(spawn);
13 use PublicInbox::MID qw(mid_mime mid2path);
16 my ($class, $git, $name, $email) = @_;
19 ident => "$name <$email>",
21 ref => 'refs/heads/master',
25 # idempotent start function
29 return ($self->{in}, $self->{out}) if $self->{pid};
31 my ($in_r, $in_w, $out_r, $out_w);
32 pipe($in_r, $in_w) or die "pipe failed: $!";
33 pipe($out_r, $out_w) or die "pipe failed: $!";
34 my $git = $self->{git};
35 my $git_dir = $git->{git_dir};
36 my $lockpath = "$git_dir/ssoma.lock";
37 sysopen(my $lockfh, $lockpath, O_WRONLY|O_CREAT) or
38 die "failed to open lock $lockpath: $!";
40 # wait for other processes to be done
41 flock($lockfh, LOCK_EX) or die "lock failed: $!\n";
42 chomp($self->{tip} = $git->qx(qw(rev-parse --revs-only), $self->{ref}));
44 my @cmd = ('git', "--git-dir=$git_dir", qw(fast-import
45 --quiet --done --date-format=rfc2822));
46 my $rdr = { 0 => fileno($out_r), 1 => fileno($in_w) };
47 my $pid = spawn(\@cmd, undef, $rdr);
48 die "spawn failed: $!" unless defined $pid;
51 $self->{out} = $out_w;
52 $self->{lockfh} = $lockfh;
57 sub wfail () { die "write to fast-import failed: $!" }
61 my $day = qw(Sun Mon Tue Wed Thu Fri Sat)[$t[6]];
62 my $mon = qw(Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec)[$t[4]];
64 sprintf('%s, %2d %s %d %02d:%02d:%02d +0000',
65 $day, $t[3], $mon, $t[5] + 1900, $t[2], $t[1], $t[0]);
68 # returns undef on non-existent
69 # (-1, msg) on mismatch
70 # (:MARK, msg) on success
72 my ($self, $mime) = @_; # mime = Email::MIME
74 my $mid = mid_mime($mime);
75 my $path = mid2path($mid);
77 my ($r, $w) = $self->gfi_start;
78 my $tip = $self->{tip};
81 print $w "ls $tip $path\n" or wfail;
84 defined $check or die "EOF from fast-import / ls: $!";
85 return if $check =~ /\Amissing /;
86 $check =~ m!\A100644 blob ([a-f0-9]{40})\t!s or die "not blob: $check";
88 print $w "cat-blob $blob\n" or wfail;
90 defined $check or die "EOF from fast-import / cat-blob: $!";
91 $check =~ /\A[a-f0-9]{40} blob (\d+)\n\z/ or
92 die "unexpected cat-blob response: $check";
97 my $n = read($r, $buf, $left, $offset);
98 defined($n) or die "read cat-blob failed: $!";
99 $n == 0 and die 'fast-export (cat-blob) died';
104 die "bad read on final byte: <$lf>" if $lf ne "\n";
105 my $cur = Email::MIME->new($buf);
106 if ($cur->header('Subject') ne $mime->header('Subject') ||
107 $cur->body ne $mime->body) {
111 my $ref = $self->{ref};
112 my $commit = $self->{mark}++;
113 my $parent = $tip =~ /\A:/ ? $tip : undef;
115 print $w "reset $ref\n" or wfail;
117 my $ident = $self->{ident};
119 print $w "commit $ref\nmark :$commit\n",
120 "author $ident $now\n",
121 "committer $ident $now\n",
123 'from ', ($parent ? $parent : $tip), "\n" or wfail;
124 print $w "D $path\n\n" or wfail;
125 (($self->{tip} = ":$commit"), $cur);
128 # returns undef on duplicate
130 my ($self, $mime) = @_; # mime = Email::MIME
132 my $from = $mime->header('From');
133 my @from = Email::Address->parse($from);
134 my $name = $from[0]->name;
135 my $email = $from[0]->address;
136 my $date = $mime->header('Date');
137 my $subject = $mime->header('Subject');
138 $subject = '(no subject)' unless defined $subject;
139 my $mid = mid_mime($mime);
140 my $path = mid2path($mid);
142 my ($r, $w) = $self->gfi_start;
143 my $tip = $self->{tip};
145 print $w "ls $tip $path\n" or wfail;
148 defined $check or die "EOF from fast-import: $!";
149 return unless $check =~ /\Amissing /;
152 # kill potentially confusing/misleading headers
153 $mime->header_set($_) for qw(bytes lines content-length status);
154 $mime = $mime->as_string;
155 my $blob = $self->{mark}++;
156 print $w "blob\nmark :$blob\ndata ", length($mime), "\n" or wfail;
157 print $w $mime, "\n" or wfail;
158 my $ref = $self->{ref};
159 my $commit = $self->{mark}++;
160 my $parent = $tip =~ /\A:/ ? $tip : undef;
163 print $w "reset $ref\n" or wfail;
166 # quiet down wide character warnings:
167 binmode $w, ':utf8' or die "binmode :utf8 failed: $!";
168 print $w "commit $ref\nmark :$commit\n",
169 "author $name <$email> $date\n",
170 "committer $self->{ident} ", now2822(), "\n",
171 "data ", (bytes::length($subject) + 1), "\n",
172 $subject, "\n\n" or wfail;
173 binmode $w, ':raw' or die "binmode :raw failed: $!";
176 print $w 'from ', ($parent ? $parent : $tip), "\n" or wfail;
178 print $w "M 100644 :$blob $path\n\n" or wfail;
179 $self->{tip} = ":$commit";
184 my $w = delete $self->{out} or return;
185 my $r = delete $self->{in} or die 'BUG: missing {in} when done';
186 print $w "done\n" or wfail;
187 my $pid = delete $self->{pid} or die 'BUG: missing {pid} when done';
188 waitpid($pid, 0) == $pid or die 'fast-import did not finish';
189 $? == 0 or die "fast-import failed: $?";
190 my $lockfh = delete $self->{lockfh} or die "BUG: not locked: $!";
191 flock($lockfh, LOCK_UN) or die "unlock failed: $!";
192 close $lockfh or die "close lock failed: $!";