1 # Copyright (C) 2021 all contributors <meta@public-inbox.org>
2 # License: AGPL-3.0+ <https://www.gnu.org/licenses/agpl-3.0.txt>
4 # "lei add-external --mirror" support
5 package PublicInbox::LeiMirror;
8 use parent qw(PublicInbox::IPC);
9 use IO::Uncompress::Gunzip qw(gunzip $GunzipError);
10 use PublicInbox::Spawn qw(popen_rd spawn);
11 use PublicInbox::PktOp;
13 sub mirror_done { # EOF callback for main daemon
15 my $mrr = delete $lei->{mrr};
16 $mrr->wq_wait_old($lei) if $mrr;
17 # FIXME: check $? before finish
18 $lei->add_external_finish($mrr->{dst});
22 # for old installations without manifest.js.gz
25 my $uri = URI->new($self->{src});
26 my $lei = $self->{lei};
27 my $curl = $self->{curl} //= PublicInbox::LeiCurl->new($lei) or return;
28 my $cmd = $curl->for_uri($lei, $uri);
29 my $opt = { 0 => $lei->{0}, 2 => $lei->{2} };
30 my $fh = popen_rd($cmd, $lei->{env}, $opt);
31 my $html = do { local $/; <$fh> } // die "read(curl $uri): $!";
32 close($fh) or return $lei->child_error($?, "@$cmd failed");
34 # we grep with URL below, we don't want Subject/From headers
35 # making us clone random URLs
36 my @urls = ($html =~ m!\bgit clone --mirror ([a-z\+]+://\S+)!g);
37 my $url = $uri->as_string;
38 chop($url) eq '/' or die "BUG: $uri not canonicalized";
40 # since this is for old instances w/o manifest.js.gz, try v1 first
41 return clone_v1($self) if grep(m!\A\Q$url\E/*\z!, @urls);
42 if (my @v2_urls = grep(m!\A\Q$url\E/[0-9]+\z!, @urls)) {
43 my %v2_uris = map { $_ => URI->new($_) } @v2_urls; # uniq
44 return clone_v2($self, [ values %v2_uris ]);
47 # filter out common URLs served by WWW (e.g /$MSGID/T/)
48 if (@urls && $url =~ s!/+[^/]+\@[^/]+/.*\z!! &&
49 grep(m!\A\Q$url\E/*\z!, @urls)) {
51 E: confused by scraping <$uri>, did you mean <$url>?
55 E: confused by scraping <$uri>, got ambiguous results:
58 die "E: scraping <$uri> revealed nothing\n";
64 # we support "-c $key=$val" for arbitrary git config options
65 # e.g.: git -c http.proxy=socks5h://127.0.0.1:9050
66 push(@cmd, '-c', $_) for @{$lei->{opt}->{c} // []};
67 push @cmd, qw(clone --mirror);
68 push @cmd, '-q' if $lei->{opt}->{quiet};
69 push @cmd, '-v' if $lei->{opt}->{verbose};
70 # XXX any other options to support?
71 # --reference is tricky with multiple epochs...
75 # tries the relatively new /$INBOX/_/text/config/raw endpoint
78 my $dst = $self->{dst};
79 if (!-d $dst || !mkdir($dst)) {
81 File::Path::mkpath($dst);
82 -d $dst or die "mkpath($dst): $!\n";
84 my $uri = URI->new($self->{src});
85 my $lei = $self->{lei};
86 my $path = $uri->path;
87 chop($path) eq '/' or die "BUG: $uri not canonicalized";
88 $uri->path($path . '/_/text/config/raw');
89 my $cmd = $self->{curl}->for_uri($lei, $uri);
90 push @$cmd, '--compressed'; # curl decompresses for us
91 my $ce = "$dst/inbox.config.example";
93 open(my $fh, '+>', $f) or return $lei->err("open $f: $! (non-fatal)");
94 my $opt = { 0 => $lei->{0}, 1 => $fh, 2 => $lei->{2} };
95 $lei->qerr("# @$cmd");
96 my $pid = spawn($cmd, $lei->{env}, $opt);
97 waitpid($pid, 0) == $pid or return $lei->err("waitpid @$cmd: $!");
98 if (($? >> 8) == 22) { # 404 missing
99 unlink($f) if -s $fh == 0;
102 return $lei->err("# @$cmd failed (non-fatal)") if $?;
103 rename($f, $ce) or return $lei->err("link($f, $ce): $! (non-fatal)");
104 my $cfg = PublicInbox::Config::git_config_dump($f);
105 my $ibx = $self->{ibx} = {};
106 for my $sec (grep(/\Apublicinbox\./, @{$cfg->{-section_order}})) {
107 for (qw(address newsgroup nntpmirror)) {
108 $ibx->{$_} = $cfg->{"$sec.$_"};
113 sub index_cloned_inbox {
114 my ($self, $iv) = @_;
115 my $ibx = delete($self->{ibx}) // {
116 address => [ 'lei@example.com' ],
119 $ibx->{inboxdir} = $self->{dst};
120 PublicInbox::Inbox->new($ibx);
121 PublicInbox::InboxWritable->new($ibx);
123 my $lei = $self->{lei};
124 for my $sw ($lei->index_opt) {
125 my ($k) = ($sw =~ /\A([\w-]+)/);
126 $opt->{$k} = $lei->{opt}->{$k};
128 # force synchronous dwaitpid for v2:
129 local $PublicInbox::DS::in_loop = 0;
130 my $cfg = PublicInbox::Config->new;
131 my $env = PublicInbox::Admin::index_prepare($opt, $cfg);
132 local %ENV = (%ENV, %$env) if $env;
133 PublicInbox::Admin::progress_prepare($opt, $lei->{2});
134 PublicInbox::Admin::index_inbox($ibx, undef, $opt);
139 my $lei = $self->{lei};
140 my $curl = $self->{curl} //= PublicInbox::LeiCurl->new($lei) or return;
141 my $uri = URI->new($self->{src});
142 my $pfx = $curl->torsocks($lei, $uri) or return;
143 my $cmd = [ @$pfx, clone_cmd($lei), $uri->as_string, $self->{dst} ];
144 $lei->qerr("# @$cmd");
145 my $pid = spawn($cmd, $lei->{env}, $lei);
146 waitpid($pid, 0) == $pid or die "BUG: waitpid @$cmd: $!";
147 $? == 0 or return $lei->child_error($?, "@$cmd failed");
149 index_cloned_inbox($self, 1);
153 my ($self, $v2_uris) = @_;
154 my $lei = $self->{lei};
155 my $curl = $self->{curl} //= PublicInbox::LeiCurl->new($lei) or return;
156 my $pfx //= $curl->torsocks($lei, $v2_uris->[0]) or return;
158 my $dst = $self->{dst};
160 for my $uri (@$v2_uris) {
161 my $src = $uri->as_string;
163 $src =~ m!/([0-9]+)(?:\.git)?\z! or die <<"";
164 failed to extract epoch number from $src
167 $edst .= "/git/$nr.git";
168 push @src_edst, [ $src, $edst ];
170 my $lk = bless { lock_path => "$dst/inbox.lock" }, 'PublicInbox::Lock';
172 my $on_destroy = $lk->lock_for_scope($$);
173 my @cmd = clone_cmd($lei);
174 while (my $pair = shift(@src_edst)) {
175 my $cmd = [ @$pfx, @cmd, @$pair ];
176 $lei->qerr("# @$cmd");
177 my $pid = spawn($cmd, $lei->{env}, $lei);
178 waitpid($pid, 0) == $pid or die "BUG: waitpid @$cmd: $!";
179 $? == 0 or return $lei->child_error($?, "@$cmd failed");
181 undef $on_destroy; # unlock
182 index_cloned_inbox($self, 2);
187 my $uri = URI->new($self->{src});
188 my $lei = $self->{lei};
189 my $curl = $self->{curl} //= PublicInbox::LeiCurl->new($lei) or return;
190 my $path = $uri->path;
191 chop($path) eq '/' or die "BUG: $uri not canonicalized";
192 $uri->path($path . '/manifest.js.gz');
193 my $cmd = $curl->for_uri($lei, $uri);
194 $lei->qerr("# @$cmd");
195 my $opt = { 0 => $lei->{0}, 2 => $lei->{2} };
196 my $fh = popen_rd($cmd, $lei->{env}, $opt);
197 my $gz = do { local $/; <$fh> } // die "read(curl $uri): $!";
199 return try_scrape($self) if ($? >> 8) == 22; # 404 missing
200 return $lei->child_error($?, "@$cmd failed");
203 gunzip(\$gz => \$js, MultiStream => 1) or
204 die "gunzip($uri): $GunzipError";
205 my $m = eval { PublicInbox::Config->json->decode($js) };
206 die "$uri: error decoding `$js': $@" if $@;
207 ref($m) eq 'HASH' or die "$uri unknown type: ".ref($m);
209 my $v1_bare = $m->{$path};
210 my @v2_epochs = grep(m!\A\Q$path\E/git/[0-9]+\.git\z!, keys %$m);
212 # It may be possible to have v1 + v2 in parallel someday:
213 $lei->err(<<EOM) if defined $v1_bare;
214 # `$v1_bare' appears to be a v1 inbox while v2 epochs exist:
216 # ignoring $v1_bare (use --inbox-version=1 to force v1 instead)
218 @v2_epochs = map { $uri->path($_); $uri->clone } @v2_epochs;
219 clone_v2($self, \@v2_epochs);
222 } elsif (my @maybe = grep(m!\Q$path\E!, keys %$m)) {
223 die "E: confused by <$uri>, possible matches:\n@maybe";
225 die "E: confused by <$uri>";
229 sub start_clone_url {
231 return try_manifest($self) if $self->{src} =~ m!\Ahttps?://!;
232 die "TODO: non-HTTP/HTTPS clone of $self->{src} not supported, yet";
235 sub do_mirror { # via wq_do
237 my $lei = $self->{lei};
239 my $iv = $lei->{opt}->{'inbox-version'};
241 return clone_v1($self) if $iv == 1;
242 return try_scrape($self) if $iv == 2;
243 die "bad --inbox-version=$iv\n";
245 return start_clone_url($self) if $self->{src} =~ m!://!;
246 die "TODO: cloning local directories not supported, yet";
248 return $lei->fail($@) if $@;
249 $lei->qerr("# mirrored $self->{src} => $self->{dst}");
253 my ($cls, $lei, $src, $dst) = @_;
254 my $self = bless { lei => $lei, src => $src, dst => $dst }, $cls;
256 if ($src =~ m!https?://!) {
258 require PublicInbox::LeiCurl;
260 require PublicInbox::Lock;
261 require PublicInbox::Inbox;
262 require PublicInbox::Admin;
263 require PublicInbox::InboxWritable;
265 '!' => [ $lei->can('fail_handler'), $lei ],
266 'x_it' => [ $lei->can('x_it'), $lei ],
267 'child_error' => [ $lei->can('child_error'), $lei ],
268 '' => [ \&mirror_done, $lei ],
270 ($lei->{pkt_op_c}, $lei->{pkt_op_p}) = PublicInbox::PktOp->pair($ops);
271 $self->wq_workers_start('lei_mirror', 1, $lei->oldset, {lei => $lei});
272 my $op = delete $lei->{pkt_op_c};
273 delete $lei->{pkt_op_p};
274 $self->wq_do('do_mirror', []);
276 $lei->event_step_init; # wait for shutdowns
277 if ($lei->{oneshot}) {
278 while ($op->{sock}) { $op->event_step }
282 sub ipc_atfork_child {
284 $self->{lei}->lei_atfork_child;
285 $self->SUPER::ipc_atfork_child;