1 # Copyright (C) 2020-2021 all contributors <meta@public-inbox.org>
2 # License: AGPL-3.0+ <https://www.gnu.org/licenses/agpl-3.0.txt>
4 # Backend for `lei' (local email interface). Unlike the C10K-oriented
5 # PublicInbox::Daemon, this is designed exclusively to handle trusted
6 # local clients with read/write access to the FS and use as many
7 # system resources as the local user has access to.
8 package PublicInbox::LEI;
11 use parent qw(PublicInbox::DS PublicInbox::LeiExternal
12 PublicInbox::LeiQuery);
14 use Socket qw(AF_UNIX SOCK_SEQPACKET MSG_EOR pack_sockaddr_un);
15 use Errno qw(EPIPE EAGAIN EINTR ECONNREFUSED ENOENT ECONNRESET);
17 use POSIX qw(strftime);
19 use Fcntl qw(SEEK_SET);
20 use PublicInbox::Config;
21 use PublicInbox::Syscall qw(SFD_NONBLOCK EPOLLIN EPOLLET);
22 use PublicInbox::Sigfd;
23 use PublicInbox::DS qw(now dwaitpid);
24 use PublicInbox::Spawn qw(spawn popen_rd);
25 use PublicInbox::Lock;
26 use Time::HiRes qw(stat); # ctime comparisons for config cache
27 use File::Path qw(mkpath);
29 our $quit = \&CORE::exit;
30 our ($current_lei, $errors_log, $listener, $oldset);
31 my ($recv_cmd, $send_cmd);
32 my $GLP = Getopt::Long::Parser->new;
33 $GLP->configure(qw(gnu_getopt no_ignore_case auto_abbrev));
34 my $GLP_PASS = Getopt::Long::Parser->new;
35 $GLP_PASS->configure(qw(gnu_getopt no_ignore_case auto_abbrev pass_through));
37 our %PATH2CFG; # persistent for socket daemon
39 # TBD: this is a documentation mechanism to show a subcommand
40 # (may) pass options through to another command:
41 sub pass_through { $GLP_PASS }
45 my ($spec, $re_str) = @_; # 'limit|n=i', '([0-9]+)'
46 my ($key) = ($spec =~ m/\A([a-z]+)/g);
47 my $cb = sub { # Getopt::Long "<>" catch-all handler
49 if ($arg =~ /\A-($re_str)\z/) {
51 } elsif ($arg eq '--') { # "--" arg separator, ignore first
52 push @{$OPT->{-argv}}, $arg if $OPT->{'--'}++;
53 # lone (single) dash is handled elsewhere
54 } elsif (substr($arg, 0, 1) eq '-') {
56 push @{$OPT->{-argv}}, $arg;
58 die "bad argument: $arg\n";
61 push @{$OPT->{-argv}}, $arg;
64 ($spec, '<>' => $cb, $GLP_PASS) # for Getopt::Long
69 return $p if index($p, '/') == 0; # already absolute
70 my $pwd = $self->{env}->{PWD};
72 my $cwd = $self->{3} // getcwd() // die "getcwd(PWD=$pwd): $!";
73 if (my @st_pwd = stat($pwd)) {
74 my @st_cwd = stat($cwd) or die "stat($cwd): $!";
75 "@st_pwd[1,0]" eq "@st_cwd[1,0]" or
76 $self->{env}->{PWD} = $pwd = $cwd;
77 } else { # PWD was invalid
78 delete $self->{env}->{PWD};
82 $pwd //= $self->{env}->{PWD} = getcwd() // die "getcwd(PWD=$pwd): $!";
83 File::Spec->rel2abs($p, $pwd);
88 rel2abs($self, ($self->{env}->{XDG_DATA_HOME} //
89 ($self->{env}->{HOME} // '/nonexistent').'/.local/share')
93 sub _config_path ($) {
95 rel2abs($self, ($self->{env}->{XDG_CONFIG_HOME} //
96 ($self->{env}->{HOME} // '/nonexistent').'/.config')
102 rel2abs($self, ($self->{env}->{XDG_CACHE_HOME} //
103 ($self->{env}->{HOME} // '/nonexistent').'/.cache')
109 $self->{ale} //= do {
110 require PublicInbox::LeiALE;
111 $self->_lei_cfg(1)->{ale} //= PublicInbox::LeiALE->new($self);
116 # TODO: drop underscore variants everywhere, they're undocumented
117 qw(fsync|sync! jobs|j=i indexlevel|L=s compact
118 max_size|max-size=s sequential_shard|sequential-shard
119 batch_size|batch-size=s skip-docdata)
122 my @c_opt = qw(c=s@ C=s@ quiet|q);
124 # we generate shell completion + help using %CMD and %OPTDESC,
125 # see lei__complete() and PublicInbox::LeiHelp
126 # command => [ positional_args, 1-line description, Getopt::Long option spec ]
127 our %CMD = ( # sorted in order of importance/use:
128 'q' => [ '--stdin|SEARCH_TERMS...', 'search for messages matching terms',
129 'stdin|', # /|\z/ must be first for lone dash
130 qw(save-as=s output|mfolder|o=s format|f=s dedupe|d=s threads|t+
131 sort|s=s reverse|r offset=i remote! local! external! pretty
132 include|I=s@ exclude=s@ only=s@ jobs|j=s globoff|g augment|a
133 import-remote! import-before! lock=s@ rsyncable
134 alert=s@ mua=s no-torsocks torsocks=s verbose|v+), @c_opt,
135 PublicInbox::LeiQuery::curl_opt(), opt_dash('limit|n=i', '[0-9]+') ],
137 'show' => [ 'MID|OID', 'show a given object (Message-ID or object ID)',
138 qw(type=s solve! format|f=s dedupe|d=s threads|t remote local!
139 verbose|v+), @c_opt, pass_through('git show') ],
141 'add-external' => [ 'LOCATION',
142 'add/set priority of a publicinbox|extindex for extra matches',
143 qw(boost=i mirror=s no-torsocks torsocks=s inbox-version=i
144 verbose|v+), @c_opt, index_opt(),
145 PublicInbox::LeiQuery::curl_opt() ],
146 'ls-external' => [ '[FILTER]', 'list publicinbox|extindex locations',
147 qw(format|f=s z|0 globoff|g invert-match|v local remote), @c_opt ],
148 'ls-label' => [ '', 'list labels', qw(z|0 stats:s), @c_opt ],
149 'forget-external' => [ 'LOCATION...|--prune',
150 'exclude further results from a publicinbox|extindex',
153 'ls-query' => [ '[FILTER...]', 'list saved search queries',
154 qw(name-only format|f=s), @c_opt ],
155 'rm-query' => [ 'QUERY_NAME', 'remove a saved search', @c_opt ],
156 'mv-query' => [ qw(OLD_NAME NEW_NAME), 'rename a saved search', @c_opt ],
158 'plonk' => [ '--threads|--from=IDENT',
159 'exclude mail matching From: or threads from non-Message-ID searches',
160 qw(stdin| threads|t from|f=s mid=s oid=s), @c_opt ],
161 'mark' => [ 'KEYWORDS...',
162 'set/unset keywords on message(s)',
163 qw(stdin| in-format|F=s input|i=s@ oid=s@ mid=s@), @c_opt,
164 pass_through('-kw:foo for delete') ],
165 'forget' => [ '[--stdin|--oid=OID|--by-mid=MID]',
166 "exclude message(s) on stdin from `q' search results",
167 qw(stdin| oid=s exact by-mid|mid:s), @c_opt ],
169 'purge-mailsource' => [ 'LOCATION|--all',
170 'remove imported messages from IMAP, Maildirs, and MH',
171 qw(exact! all jobs:i indexed), @c_opt ],
173 # code repos are used for `show' to solve blobs from patch mails
174 'add-coderepo' => [ 'DIRNAME', 'add or set priority of a git code repo',
175 qw(boost=i), @c_opt ],
176 'ls-coderepo' => [ '[FILTER_TERMS...]',
177 'list known code repos', qw(format|f=s z), @c_opt ],
178 'forget-coderepo' => [ 'DIRNAME',
179 'stop using repo to solve blobs from patches',
182 'add-watch' => [ 'LOCATION', 'watch for new messages and flag changes',
183 qw(import! kw|keywords|flags! interval=s recursive|r
184 exclude=s include=s), @c_opt ],
185 'ls-watch' => [ '[FILTER...]', 'list active watches with numbers and status',
186 qw(format|f=s z), @c_opt ],
187 'pause-watch' => [ '[WATCH_NUMBER_OR_FILTER]', qw(all local remote), @c_opt ],
188 'resume-watch' => [ '[WATCH_NUMBER_OR_FILTER]', qw(all local remote), @c_opt ],
189 'forget-watch' => [ '{WATCH_NUMBER|--prune}', 'stop and forget a watch',
192 'import' => [ 'LOCATION...|--stdin',
193 'one-time import/update from URL or filesystem',
194 qw(stdin| offset=i recursive|r exclude=s include|I=s
195 lock=s@ in-format|F=s kw|keywords|flags! verbose|v+), @c_opt ],
196 'convert' => [ 'LOCATION...|--stdin',
197 'one-time conversion from URL or filesystem to another format',
198 qw(stdin| in-format|F=s out-format|f=s output|mfolder|o=s
199 lock=s@ kw|keywords|flags!), @c_opt ],
200 'p2q' => [ 'FILE|COMMIT_OID|--stdin',
201 "use a patch to generate a query for `lei q --stdin'",
202 qw(stdin| want|w=s@ uri debug), @c_opt ],
203 'config' => [ '[...]', sub {
204 'git-config(1) wrapper for '._config_path($_[0]);
205 }, qw(config-file|system|global|file|f=s), # for conflict detection
206 qw(c=s@ C=s@), pass_through('git config') ],
207 'init' => [ '[DIRNAME]', sub {
208 "initialize storage, default: ".store_path($_[0]);
210 'daemon-kill' => [ '[-SIGNAL]', 'signal the lei-daemon',
211 # "-C DIR" conflicts with -CHLD, here, and chdir makes no sense, here
212 opt_dash('signal|s=s', '[0-9]+|(?:[A-Z][A-Z0-9]+)') ],
213 'daemon-pid' => [ '', 'show the PID of the lei-daemon' ],
214 'help' => [ '[SUBCOMMAND]', 'show help' ],
216 # XXX do we need this?
217 # 'git' => [ '[ANYTHING...]', 'git(1) wrapper', pass_through('git') ],
219 'reorder-local-store-and-break-history' => [ '[REFNAME]',
220 'rewrite git history in an attempt to improve compression',
223 # internal commands are prefixed with '_'
224 '_complete' => [ '[...]', 'internal shell completion helper',
225 pass_through('everything') ],
228 # switch descriptions, try to keep consistent across commands
229 # $spec: Getopt::Long option specification
230 # $spec => [@ALLOWED_VALUES (default is first), $description],
231 # $spec => $description
232 # "$SUB_COMMAND TAB $spec" => as above
233 my $stdin_formats = [ 'MAIL_FORMAT|eml|mboxrd|mboxcl2|mboxcl|mboxo',
234 'specify message input format' ];
235 my $ls_format = [ 'OUT|plain|json|null', 'listing output format' ];
237 # we use \x{a0} (non-breaking SP) to avoid wrapping in PublicInbox::LeiHelp
239 'help|h' => 'show this built-in help',
240 'c=s@' => [ 'NAME=VALUE', 'set config option' ],
241 'C=s@' => [ 'DIR', 'chdir to specify to directory' ],
242 'quiet|q' => 'be quiet',
243 'lock=s@' => [ 'METHOD|dotlock|fcntl|flock|none',
244 'mbox(5) locking method(s) to use (default: fcntl,dotlock)' ],
246 'globoff|g' => "do not match locations using '*?' wildcards ".
247 "and\xa0'[]'\x{a0}ranges",
248 'verbose|v+' => 'be more verbose',
249 'external!' => 'do not use externals',
250 'solve!' => 'do not attempt to reconstruct blobs from emails',
251 'torsocks=s' => ['VAL|auto|no|yes',
252 'whether or not to wrap git and curl commands with torsocks'],
253 'no-torsocks' => 'alias for --torsocks=no',
254 'save-as=s' => ['NAME', 'save a search terms by given name'],
255 'import-remote!' => 'do not memoize remote messages into local store',
257 'type=s' => [ 'any|mid|git', 'disambiguate type' ],
259 'dedupe|d=s' => ['STRATEGY|content|oid|mid|none',
260 'deduplication strategy'],
262 'return all messages in the same threads as the actual match(es)',
264 'want|w=s@' => [ 'PREFIX|dfpost|dfn', # common ones in help...
265 'search prefixes to extract (default: dfpost7)' ],
267 'alert=s@' => ['CMD,:WINCH,:bell,<any command>',
268 'run command(s) or perform ops when done writing to output ' .
269 '(default: ":WINCH,:bell" with --mua and Maildir/IMAP output, ' .
270 'nothing otherwise)' ],
272 'augment|a' => 'augment --output destination instead of clobbering',
274 'output|mfolder|o=s' => [ 'MFOLDER',
275 "destination (e.g.\xa0`/path/to/Maildir', ".
276 "or\xa0`-'\x{a0}for\x{a0}stdout)" ],
278 "MUA to run on --output Maildir or mbox (e.g.\xa0`mutt\xa0-f\xa0%f')" ],
280 'inbox-version=i' => [ 'NUM|1|2',
281 'force a public-inbox version with --mirror'],
282 'mirror=s' => [ 'URL', 'mirror a public-inbox'],
284 # public-inbox-index options
285 'fsync!' => 'speed up indexing after --mirror, risk index corruption',
286 'compact' => 'run compact index after mirroring',
287 'indexlevel|L=s' => [ 'LEVEL|full|medium|basic',
288 "indexlevel with --mirror (default: full)" ],
289 'max_size|max-size=s' => [ 'SIZE',
290 'do not index messages larger than SIZE (default: infinity)' ],
291 'batch_size|batch-size=s' => [ 'SIZE',
292 'flush changes to OS after given number of bytes (default: 1m)' ],
293 'sequential_shard|sequential-shard' =>
294 'index Xapian shards sequentially for slow storage',
296 'drop compatibility w/ public-inbox <1.6 to save ~1.5% space',
299 'OUT|maildir|mboxrd|mboxcl2|mboxcl|mboxo|html|json|jsonl|concatjson',
300 'specify output format, default depends on --output'],
301 'exclude=s@ q' => [ 'LOCATION',
302 'exclude specified external(s) from search' ],
303 'include|I=s@ q' => [ 'LOCATION',
304 'include specified external(s) in search' ],
305 'only=s@ q' => [ 'LOCATION',
306 'only use specified external(s) for search' ],
307 'jobs=s q' => [ '[SEARCH_JOBS][,WRITER_JOBS]',
308 'control number of search and writer jobs' ],
309 'jobs|j=i add-external' => 'set parallelism when indexing after --mirror',
311 'in-format|F=s' => $stdin_formats,
312 'format|f=s ls-query' => $ls_format,
313 'format|f=s ls-external' => $ls_format,
315 'limit|n=i@' => ['NUM', 'limit on number of matches (default: 10000)' ],
316 'offset=i' => ['OFF', 'search result offset (default: 0)'],
318 'sort|s=s' => [ 'VAL|received|relevance|docid',
319 "order of results is `--output'-dependent"],
320 'reverse|r' => 'reverse search results', # like sort(1)
322 'boost=i' => 'increase/decrease priority of results (default: 0)',
324 'local' => 'limit operations to the local filesystem',
325 'local!' => 'exclude results from the local filesystem',
326 'remote' => 'limit operations to those requiring network access',
327 'remote!' => 'prevent operations requiring network access',
329 'mid=s' => 'specify the Message-ID of a message',
330 'oid=s' => 'specify the git object ID of a message',
332 'recursive|r' => 'scan directories/mailboxes/newsgroups recursively',
333 'exclude=s' => 'exclude mailboxes/newsgroups based on pattern',
334 'include=s' => 'include mailboxes/newsgroups based on pattern',
336 'exact' => 'operate on exact header matches only',
337 'exact!' => 'rely on content match instead of exact header matches',
339 'by-mid|mid:s' => [ 'MID', 'match only by Message-ID, ignoring contents' ],
341 'kw|keywords|flags!' => 'disable/enable importing flags',
343 # xargs, env, use "-0", git(1) uses "-z". We support z|0 everywhere
344 'z|0' => 'use NUL \\0 instead of newline (CR) to delimit lines',
346 'signal|s=s' => [ 'SIG', 'signal to send lei-daemon (default: TERM)' ],
350 'leistore.dir' => 'top-level storage location',
353 my @WQ_KEYS = qw(lxs l2m imp mrr cnv p2q mark); # internal workers
355 # pronounced "exit": x_it(1 << 8) => exit(1); x_it(13) => SIGPIPE
357 my ($self, $code) = @_;
358 # make sure client sees stdout before exit
359 $self->{1}->autoflush(1) if $self->{1};
360 dump_and_clear_log();
361 if (my $s = $self->{pkt_op_p} // $self->{sock}) {
362 send($s, "x_it $code", MSG_EOR);
363 } elsif ($self->{oneshot}) {
364 # don't want to end up using $? from child processes
365 for my $f (@WQ_KEYS) {
366 my $wq = delete $self->{$f} or next;
369 # cleanup anything that has tempfiles or open file handles
371 delete @$self{qw(ovv dedupe sto cfg)};
372 if (my $signum = ($code & 127)) { # usually SIGPIPE (13)
373 $SIG{PIPE} = 'DEFAULT'; # $SIG{$signum} doesn't work
375 sleep(1) while 1; # wait for signal
379 } # else ignore if client disconnected
384 my $err = $self->{2} // ($self->{pgr} // [])->[2] // *STDERR{GLOB};
385 my @eor = (substr($_[-1]//'', -1, 1) eq "\n" ? () : ("\n"));
386 print $err @_, @eor and return;
387 my $old_err = delete $self->{2};
388 close($old_err) if $! == EPIPE && $old_err;
389 $err = $self->{2} = ($self->{pgr} // [])->[2] // *STDERR{GLOB};
390 print $err @_, @eor or print STDERR @_, @eor;
393 sub qerr ($;@) { $_[0]->{opt}->{quiet} or err(shift, @_) }
395 sub fail_handler ($;$$) {
396 my ($lei, $code, $io) = @_;
397 for my $f (@WQ_KEYS) {
398 my $wq = delete $lei->{$f} or next;
399 $wq->wq_wait_old(undef, $lei) if $wq->wq_kill_old; # lei-daemon
401 close($io) if $io; # needed to avoid warnings on SIGPIPE
402 x_it($lei, $code // (1 << 8));
405 sub sigpipe_handler { # handles SIGPIPE from @WQ_KEYS workers
406 fail_handler($_[0], 13, delete $_[0]->{1});
409 # PublicInbox::OnDestroy callback for SIGINT to take out the entire pgid
412 dwaitpid($pgid) if kill('-INT', $pgid);
416 my ($self, $buf, $exit_code) = @_;
417 err($self, $buf) if defined $buf;
418 # calls fail_handler:
419 send($self->{pkt_op_p}, '!', MSG_EOR) if $self->{pkt_op_p};
420 x_it($self, ($exit_code // 1) << 8);
426 return if print { $self->{1} // return } @_; # likely
427 return note_sigpipe($self, 1) if $! == EPIPE;
428 my $err = "error writing to output: $!";
433 sub puts ($;@) { out(shift, map { "$_\n" } @_) }
435 sub child_error { # passes non-fatal curl exit codes to user
436 my ($self, $child_error, $msg) = @_; # child_error is $?
437 $self->err($msg) if $msg;
438 if (my $s = $self->{pkt_op_p} // $self->{sock}) {
439 # send to the parent lei-daemon or to lei(1) client
440 send($s, "child_error $child_error", MSG_EOR);
441 } elsif (!$PublicInbox::DS::in_loop) {
442 $self->{child_error} = $child_error;
443 } # else noop if client disconnected
446 sub note_sigpipe { # triggers sigpipe_handler
447 my ($self, $fd) = @_;
448 close(delete($self->{$fd})); # explicit close silences Perl warning
449 send($self->{pkt_op_p}, '|', MSG_EOR) if $self->{pkt_op_p};
453 sub _lei_atfork_child {
454 my ($self, $persist) = @_;
455 # we need to explicitly close things which are on stack
457 chdir '/' or die "chdir(/): $!";
458 my @io = delete @$self{qw(0 1 2 sock)};
459 unless ($self->{oneshot}) {
462 if (my $cfg = $self->{cfg}) {
463 delete $cfg->{-lei_store};
465 } else { # worker, Net::NNTP (Net::Cmd) uses STDERR directly
466 open STDERR, '+>&='.fileno($self->{2}) or warn "open $!";
469 for (delete @$self{qw(3 old_1 au_done)}) {
470 close($_) if defined($_);
472 if (my $op_c = delete $self->{pkt_op_c}) {
473 close(delete $op_c->{sock});
475 if (my $pgr = delete $self->{pgr}) {
476 close($_) for (@$pgr[1,2]);
478 close $listener if $listener;
482 $quit = \&CORE::exit;
483 $current_lei = $persist ? undef : $self; # for SIG{__WARN__}
486 sub _delete_pkt_op { # OnDestroy callback to prevent leaks on die
488 if (my $op = delete $self->{pkt_op_c}) { # in case of die
489 $op->close; # PublicInbox::PktOp::close
491 my $unclosed_after_die = delete($self->{pkt_op_p}) or return;
492 close $unclosed_after_die;
496 my ($self, $ops) = @_;
497 require PublicInbox::OnDestroy;
498 require PublicInbox::PktOp;
499 my $end = PublicInbox::OnDestroy->new($$, \&_delete_pkt_op, $self);
500 @$self{qw(pkt_op_c pkt_op_p)} = PublicInbox::PktOp->pair($ops);
505 my ($lei, $wq, $ident, $jobs, $ops) = @_;
507 '!' => [ \&fail_handler, $lei ],
508 '|' => [ \&sigpipe_handler, $lei ],
509 'x_it' => [ \&x_it, $lei ],
510 'child_error' => [ \&child_error, $lei ],
513 $ops->{''} //= [ \&dclose, $lei ];
514 my $end = $lei->pkt_op_pair($ops);
515 $wq->wq_workers_start($ident, $jobs, $lei->oldset, { lei => $lei });
516 delete $lei->{pkt_op_p};
517 my $op = delete $lei->{pkt_op_c};
519 $lei->event_step_init;
520 # oneshot needs $op, daemon-mode uses DS->EventLoop to handle $op
521 $lei->{oneshot} ? $op : undef;
525 require PublicInbox::LeiHelp;
526 PublicInbox::LeiHelp::call($_[0], $_[1], \%CMD, \%OPTDESC);
530 my ($self, $cmd, $argv) = @_;
531 # allow _complete --help to complete, not show help
532 return 1 if substr($cmd, 0, 1) eq '_';
534 $OPT = $self->{opt} //= {};
535 my $info = $CMD{$cmd} // [ '[...]' ];
536 my ($proto, undef, @spec) = @$info;
537 my $glp = ref($spec[-1]) eq ref($GLP) ? pop(@spec) : $GLP;
538 push @spec, qw(help|h);
540 if ($spec[0] =~ s/\|\z//s) { # "stdin|" or "clear|" allows "-" alias
541 $lone_dash = $spec[0];
542 $OPT->{$spec[0]} = \(my $var);
543 push @spec, '' => \$var;
545 $glp->getoptionsfromarray($argv, $OPT, @spec) or
546 return _help($self, "bad arguments or options for $cmd");
547 return _help($self) if $OPT->{help};
549 push @$argv, @{$OPT->{-argv}} if defined($OPT->{-argv});
551 # "-" aliases "stdin" or "clear"
552 $OPT->{$lone_dash} = ${$OPT->{$lone_dash}} if defined $lone_dash;
555 my $POS_ARG = '[A-Z][A-Z0-9_]+';
557 my @args = split(/ /, $proto);
558 for my $var (@args) {
559 if ($var =~ /\A$POS_ARG\.\.\.\z/o) { # >= 1 args;
560 $inf = defined($argv->[$i]) and last;
561 $var =~ s/\.\.\.\z//;
562 $err = "$var not supplied";
563 } elsif ($var =~ /\A$POS_ARG\z/o) { # required arg at $i
564 $argv->[$i++] // ($err = "$var not supplied");
565 } elsif ($var =~ /\.\.\.\]\z/) { # optional args start
568 } elsif ($var =~ /\A\[-?$POS_ARG\]\z/) { # one optional arg
570 } elsif ($var =~ /\A.+?\|/) { # required FOO|--stdin
571 $inf = 1 if index($var, '...') > 0;
572 my @or = split(/\|/, $var);
575 if ($o =~ /\A--([a-z0-9\-]+)/) {
576 $ok = defined($OPT->{$1});
578 } elsif (defined($argv->[$i])) {
582 } # else continue looping
586 $err = join(', ', @or) . " or $last must be set";
588 warn "BUG: can't parse `$var' in $proto";
592 if (!$inf && scalar(@$argv) > scalar(@args)) {
593 $err //= 'too many arguments';
595 $err ? fail($self, "usage: lei $cmd $proto\nE: $err") : 1;
598 sub _tmp_cfg { # for lei -c <name>=<value> ...
600 my $cfg = _lei_cfg($self, 1);
602 my $ft = File::Temp->new(TEMPLATE => 'lei_cfg-XXXX', TMPDIR => 1);
603 my $tmp = { '-f' => $ft->filename, -tmp => $ft };
605 print $ft <<EOM or return fail($self, "$tmp->{-f}: $!");
609 $tmp = $self->{cfg} = bless { %$cfg, %$tmp }, ref($cfg);
610 for (@{$self->{opt}->{c}}) {
611 /\A([^=\.]+\.[^=]+)(?:=(.*))?\z/ or return fail($self, <<EOM);
612 `-c $_' is not of the form -c <name>=<value>'
616 _config($self, '--add', $name, $value);
617 if (defined(my $v = $tmp->{$name})) {
618 if (ref($v) eq 'ARRAY') {
621 $tmp->{$name} = [ $v, $value ];
624 $tmp->{$name} = $value;
630 my ($self, $cmd, $pfx) = @_;
634 $cb = $self->can($pfx.$ucmd) and return $cb;
636 $base =~ s/_([a-z])/\u$1/g;
637 my $pkg = "PublicInbox::Lei\u$base";
638 ($INC{"PublicInbox/Lei\u$base.pm"} // eval("require $pkg")) ?
639 $pkg->can($pfx.$ucmd) : undef;
643 my ($self, $cmd, @argv) = @_;
644 local $current_lei = $self; # for __WARN__
645 dump_and_clear_log("from previous run\n");
646 return _help($self, 'no command given') unless defined($cmd);
647 # do not support Getopt bundling for this
648 while ($cmd eq '-C' || $cmd eq '-c') {
649 my $v = shift(@argv) // return fail($self, $cmd eq '-C' ?
650 '-C DIRECTORY' : '-c <name>=<value>');
651 push @{$self->{opt}->{substr($cmd, 1, 1)}}, $v;
652 $cmd = shift(@argv) // return _help($self, 'no command given');
654 if (my $cb = lazy_cb(__PACKAGE__, $cmd, 'lei_')) {
655 optparse($self, $cmd, \@argv) or return;
656 $self->{opt}->{c} and (_tmp_cfg($self) // return);
657 if (my $chdir = $self->{opt}->{C}) {
658 for my $d (@$chdir) {
659 next if $d eq ''; # same as git(1)
660 chdir $d or return fail($self, "cd $d: $!");
662 if (delete $self->{3}) { # update cwd for rel2abs
663 opendir my $dh, '.' or
664 return fail($self, "opendir . $!");
669 } elsif (grep(/\A-/, $cmd, @argv)) { # --help or -h only
670 $GLP->getoptionsfromarray([$cmd, @argv], {}, qw(help|h C=s@))
671 or return _help($self, 'bad arguments or options');
674 fail($self, "`$cmd' is not an lei command");
679 my ($self, $creat) = @_;
680 return $self->{cfg} if $self->{cfg};
681 my $f = _config_path($self);
683 my $cur_st = @st ? pack('dd', $st[10], $st[7]) : ''; # 10:ctime, 7:size
685 if (my $cfg = $PATH2CFG{$f}) { # reuse existing object in common case
686 return ($self->{cfg} = $cfg) if $cur_st eq $cfg->{-st};
687 ($sto, $sto_dir) = @$cfg{qw(-lei_store leistore.dir)};
692 return bless {}, 'PublicInbox::Config';
694 my (undef, $cfg_dir, undef) = File::Spec->splitpath($f);
695 -d $cfg_dir or mkpath($cfg_dir) or die "mkpath($cfg_dir): $!\n";
696 open my $fh, '>>', $f or die "open($f): $!\n";
697 @st = stat($fh) or die "fstat($f): $!\n";
698 $cur_st = pack('dd', $st[10], $st[7]);
699 qerr($self, "# $f created") if $self->{cmd} ne 'config';
701 my $cfg = PublicInbox::Config::git_config_dump($f);
702 bless $cfg, 'PublicInbox::Config';
703 $cfg->{-st} = $cur_st;
705 if ($sto && File::Spec->canonpath($sto_dir // store_path($self))
706 eq File::Spec->canonpath($cfg->{'leistore.dir'} //
707 store_path($self))) {
708 $cfg->{-lei_store} = $sto;
710 if (scalar(keys %PATH2CFG) > 5) {
711 # FIXME: use inotify/EVFILT_VNODE to detect unlinked configs
712 for my $k (keys %PATH2CFG) {
713 delete($PATH2CFG{$k}) unless -f $k
716 $self->{cfg} = $PATH2CFG{$f} = $cfg;
719 sub _lei_store ($;$) {
720 my ($self, $creat) = @_;
721 my $cfg = _lei_cfg($self, $creat);
722 $cfg->{-lei_store} //= do {
723 require PublicInbox::LeiStore;
724 my $dir = $cfg->{'leistore.dir'} // store_path($self);
725 return unless $creat || -d $dir;
726 PublicInbox::LeiStore->new($dir, { creat => $creat });
731 my ($self, @argv) = @_;
735 my ($self, @argv) = @_;
736 my %env = (%{$self->{env}}, GIT_CONFIG => undef);
737 my $cfg = _lei_cfg($self, 1);
738 my $cmd = [ qw(git config -f), $cfg->{'-f'}, @argv ];
739 my %rdr = map { $_ => $self->{$_} } (0..2);
740 waitpid(spawn($cmd, \%env, \%rdr), 0);
744 my ($self, @argv) = @_;
745 $self->{opt}->{'config-file'} and return fail $self,
746 "config file switches not supported by `lei config'";
748 x_it($self, $?) if $?;
752 my ($self, $dir) = @_;
753 my $cfg = _lei_cfg($self, 1);
754 my $cur = $cfg->{'leistore.dir'};
755 $dir //= store_path($self);
756 $dir = rel2abs($self, $dir);
757 my @cur = stat($cur) if defined($cur);
758 $cur = File::Spec->canonpath($cur // $dir);
759 my @dir = stat($dir);
760 my $exists = "# leistore.dir=$cur already initialized" if @dir;
763 _lei_store($self, 1)->done;
764 return qerr($self, $exists);
767 # some folks like symlinks and bind mounts :P
768 if (@dir && "@cur[1,0]" eq "@dir[1,0]") {
769 lei_config($self, 'leistore.dir', $dir);
770 _lei_store($self, 1)->done;
771 return qerr($self, "$exists (as $cur)");
773 return fail($self, <<"");
774 E: leistore.dir=$cur already initialized and it is not $dir
777 lei_config($self, 'leistore.dir', $dir);
778 _lei_store($self, 1)->done;
779 $exists //= "# leistore.dir=$dir newly initialized";
780 return qerr($self, $exists);
783 sub lei_daemon_pid { puts shift, $$ }
785 sub lei_daemon_kill {
787 my $sig = $self->{opt}->{signal} // 'TERM';
788 kill($sig, $$) or fail($self, "kill($sig, $$): $!");
791 sub lei_help { _help($_[0]) }
793 # Shell completion helper. Used by lei-completion.bash and hopefully
794 # other shells. Try to do as much here as possible to avoid redundancy
795 # and improve maintainability.
797 my ($self, @argv) = @_; # argv = qw(lei and any other args...)
798 shift @argv; # ignore "lei", the entire command is sent
799 @argv or return puts $self, grep(!/^_/, keys %CMD), qw(--help -h -C);
800 my $cmd = shift @argv;
801 my $info = $CMD{$cmd} // do { # filter matching commands
802 @argv or puts $self, grep(/\A\Q$cmd\E/, keys %CMD);
805 my ($proto, undef, @spec) = @$info;
807 my $re = defined($cur) ? qr/\A\Q$cur\E/ : qr/./;
808 if (substr(my $_cur = $cur // '-', 0, 1) eq '-') { # --switches
809 # gross special case since the only git-config options
810 # Consider moving to a table if we need more special cases
811 # we use Getopt::Long for are the ones we reject, so these
812 # are the ones we don't reject:
813 if ($cmd eq 'config') {
814 puts $self, grep(/$re/, keys %CONFIG_KEYS);
815 @spec = qw(add z|null get get-all unset unset-all
816 replace-all get-urlmatch
817 remove-section rename-section
818 name-only list|l edit|e
819 get-color-name get-colorbool);
822 # generate short/long names from Getopt::Long specs
823 puts $self, grep(/$re/, qw(--help -h -C), map {
824 if (s/[:=].+\z//) { # req/optional args, e.g output|o=i
825 } elsif (s/\+\z//) { # verbose|v+
827 # negation: solve! => no-solve|solve
828 s/([\w\-]+)/$1|no-$1/g
831 my $x = length > 1 ? "--$_" : "-$_";
832 $x eq $_cur ? () : $x;
833 } grep(!/_/, split(/\|/, $_, -1)) # help|h
834 } grep { $OPTDESC{"$_\t$cmd"} || $OPTDESC{$_} } @spec);
835 } elsif ($cmd eq 'config' && !@argv && !$CONFIG_KEYS{$cur}) {
836 puts $self, grep(/$re/, keys %CONFIG_KEYS);
839 # switch args (e.g. lei q -f mbox<TAB>)
840 if (($argv[-1] // $cur // '') =~ /\A--?([\w\-]+)\z/) {
841 my $opt = quotemeta $1;
843 my $v = $OPTDESC{$_};
844 my @v = ref($v) ? split(/\|/, $v->[0]) : ();
845 # get rid of ALL CAPS placeholder (e.g "OUT")
846 # (TODO: completion for external paths)
847 shift(@v) if scalar(@v) && uc($v[0]) eq $v[0];
849 } grep(/\A(?:[\w-]+\|)*$opt\b.*?(?:\t$cmd)?\z/, keys %OPTDESC);
851 if (my $cb = lazy_cb($self, $cmd, '_complete_')) {
852 puts $self, $cb->($self, @argv, $cur ? ($cur) : ());
854 # TODO: URLs, pathnames, OIDs, MIDs, etc... See optparse() for
859 my ($argv, $env) = @_;
860 my $argc = scalar @$argv;
861 my $buf = 'exec '.join("\0", scalar(@$argv), @$argv);
862 while (my ($k, $v) = each %$env) { $buf .= "\0$k=$v" };
868 my $mua = $self->{opt}->{mua} // return;
869 my $mfolder = $self->{ovv}->{dst};
870 my (@cmd, $replaced);
871 if ($mua =~ /\A(?:mutt|mailx|mail|neomutt)\z/) {
873 # TODO: help wanted: other common FOSS MUAs
875 require Text::ParseWords;
876 @cmd = Text::ParseWords::shellwords($mua);
877 # mutt uses '%f' for open-hook with compressed mbox, we follow
878 @cmd = map { $_ eq '%f' ? ($replaced = $mfolder) : $_ } @cmd;
880 push @cmd, $mfolder unless defined($replaced);
881 if (my $sock = $self->{sock}) { # lei(1) client process runs it
882 send($sock, exec_buf(\@cmd, {}), MSG_EOR);
883 } elsif ($self->{oneshot}) {
884 $self->{"pid.$self.$$"}->{spawn(\@cmd)} = \@cmd;
886 if ($self->{lxs} && $self->{au_done}) { # kick wait_startq
887 syswrite($self->{au_done}, 'q' x ($self->{lxs}->{jobs} // 0));
889 $self->{opt}->{quiet} = 1;
890 delete $self->{-progress};
891 delete $self->{opt}->{verbose};
894 sub send_exec_cmd { # tell script/lei to execute a command
895 my ($self, $io, $cmd, $env) = @_;
896 my $sock = $self->{sock} // die 'lei client gone';
897 my $fds = [ map { fileno($_) } @$io ];
898 $send_cmd->($sock, $fds, exec_buf($cmd, $env), MSG_EOR);
901 sub poke_mua { # forces terminal MUAs to wake up and hopefully notice new mail
903 my $alerts = $self->{opt}->{alert} // return;
904 while (my $op = shift(@$alerts)) {
905 if ($op eq ':WINCH') {
906 # hit the process group that started the MUA
908 send($self->{sock}, '-WINCH', MSG_EOR);
909 } elsif ($self->{oneshot}) {
912 } elsif ($op eq ':bell') {
914 } elsif ($op =~ /(?<!\\),/) { # bare ',' (not ',,')
915 push @$alerts, split(/(?<!\\),/, $op);
916 } elsif ($op =~ m!\A([/a-z0-9A-Z].+)!) {
917 my $cmd = $1; # run an arbitrary command
918 require Text::ParseWords;
919 $cmd = [ Text::ParseWords::shellwords($cmd) ];
920 if (my $s = $self->{sock}) {
921 send($s, exec_buf($cmd, {}), MSG_EOR);
922 } elsif ($self->{oneshot}) {
923 $self->{"pid.$self.$$"}->{spawn($cmd)} = $cmd;
926 err($self, "W: unsupported --alert=$op"); # non-fatal
931 my %path_to_fd = ('/dev/stdin' => 0, '/dev/stdout' => 1, '/dev/stderr' => 2);
932 $path_to_fd{"/dev/fd/$_"} = $path_to_fd{"/proc/self/fd/$_"} for (0..2);
934 my ($self, $mode, $path) = @_;
935 rel2abs($self, $path);
937 if (defined(my $fd = $path_to_fd{$path})) {
940 if ($path =~ m!\A/(?:dev|proc/self)/fd/[0-9]+\z!) {
941 return fail($self, "cannot open $path from daemon");
943 open my $fh, $mode, $path or return;
947 # caller needs to "-t $self->{1}" to check if tty
950 my $fh = popen_rd([qw(git var GIT_PAGER)]);
951 chomp(my $pager = <$fh> // '');
952 close($fh) or warn "`git var PAGER' error: \$?=$?";
953 return if $pager eq 'cat' || $pager eq '';
954 my $new_env = { LESS => 'FRX', LV => '-c' };
955 $new_env->{MORE} = 'FRX' if $^O eq 'freebsd';
956 pipe(my ($r, $wpager)) or return warn "pipe: $!";
957 my $rdr = { 0 => $r, 1 => $self->{1}, 2 => $self->{2} };
958 my $pgr = [ undef, @$rdr{1, 2} ];
959 my $env = $self->{env};
960 if ($self->{sock}) { # lei(1) process runs it
961 delete @$new_env{keys %$env}; # only set iff unset
962 send_exec_cmd($self, [ @$rdr{0..2} ], [$pager], $new_env);
963 } elsif ($self->{oneshot}) {
965 $self->{"pid.$self.$$"}->{spawn($cmd, $new_env, $rdr)} = $cmd;
967 die 'BUG: start_pager w/o socket';
969 $self->{1} = $wpager;
970 $self->{2} = $wpager if -t $self->{2};
971 $env->{GIT_PAGER_IN_USE} = 'true'; # we may spawn git
977 my $pgr = delete($self->{pgr}) or return;
978 $self->{2} = $pgr->[2];
979 # do not restore original stdout, just close it so we error out
980 close(delete($self->{1})) if $self->{1};
983 sub accept_dispatch { # Listener {post_accept} callback
984 my ($sock) = @_; # ignore other
986 my $self = bless { sock => $sock }, __PACKAGE__;
987 vec(my $rvec = '', fileno($sock), 1) = 1;
988 select($rvec, undef, undef, 60) or
989 return send($sock, 'timed out waiting to recv FDs', MSG_EOR);
990 # (4096 * 33) >MAX_ARG_STRLEN
991 my @fds = $recv_cmd->($sock, my $buf, 4096 * 33) or return; # EOF
992 if (!defined($fds[0])) {
993 warn(my $msg = "recv_cmd failed: $!");
994 return send($sock, $msg, MSG_EOR);
998 open($self->{$i++}, '+<&=', $fd) and next;
999 send($sock, "open(+<&=$fd) (FD=$i): $!", MSG_EOR);
1001 return if scalar(@fds) != 4;
1003 $self->{2}->autoflush(1); # keep stdout buffered until x_it|DESTROY
1004 # $ENV_STR = join('', map { "\0$_=$ENV{$_}" } keys %ENV);
1005 # $buf = "$argc\0".join("\0", @ARGV).$ENV_STR."\0\0";
1006 substr($buf, -2, 2, '') eq "\0\0" or # s/\0\0\z//
1007 return send($sock, 'request command truncated', MSG_EOR);
1008 my ($argc, @argv) = split(/\0/, $buf, -1);
1010 my %env = map { split(/=/, $_, 2) } splice(@argv, $argc);
1011 if (chdir($self->{3})) {
1013 $self->{env} = \%env;
1014 eval { dispatch($self, @argv) };
1015 send($sock, $@, MSG_EOR) if $@;
1017 send($sock, "fchdir: $!", MSG_EOR); # implicit close
1023 delete $self->{-progress};
1024 for my $f (@WQ_KEYS) {
1025 my $wq = delete $self->{$f} or next;
1027 $wq->wq_close(0, undef, $self);
1028 } elsif ($wq->wq_kill_old) {
1029 $wq->wq_wait_old(undef, $self);
1032 close(delete $self->{1}) if $self->{1}; # may reap_compress
1033 if (my $sto = delete $self->{sto}) {
1034 $sto->ipc_do('done');
1036 $self->close if $self->{-event_init_done}; # PublicInbox::DS::close
1039 # for long-running results
1042 local %ENV = %{$self->{env}};
1043 my $sock = $self->{sock};
1044 local $current_lei = $self;
1046 while (my @fds = $recv_cmd->($sock, my $buf, 4096)) {
1047 if (scalar(@fds) == 1 && !defined($fds[0])) {
1048 return if $! == EAGAIN;
1049 next if $! == EINTR;
1050 last if $! == ECONNRESET;
1054 open my $rfh, '+<&=', $fd;
1056 die "unrecognized client signal: $buf";
1061 eval { $self->fail($err) };
1066 sub event_step_init {
1068 return if $self->{-event_init_done}++;
1069 if (my $sock = $self->{sock}) { # using DS->EventLoop
1070 $self->SUPER::new($sock, EPOLLIN|EPOLLET);
1076 sub oldset { $oldset }
1078 sub dump_and_clear_log {
1079 if (defined($errors_log) && -s STDIN && seek(STDIN, 0, SEEK_SET)) {
1081 unshift(@pfx, "$errors_log ") if @pfx;
1082 warn @pfx, do { local $/; <STDIN> };
1083 truncate(STDIN, 0) or warn "ftruncate ($errors_log): $!";
1087 # lei(1) calls this when it can't connect
1089 my ($path, $errno, $narg) = @_;
1090 local ($errors_log, $listener);
1091 ($errors_log) = ($path =~ m!\A(.+?/)[^/]+\z!);
1092 $errors_log .= 'errors.log';
1093 my $addr = pack_sockaddr_un($path);
1094 my $lk = bless { lock_path => $errors_log }, 'PublicInbox::Lock';
1096 socket($listener, AF_UNIX, SOCK_SEQPACKET, 0) or die "socket: $!";
1097 if ($errno == ECONNREFUSED || $errno == ENOENT) {
1098 return if connect($listener, $addr); # another process won
1099 if ($errno == ECONNREFUSED && -S $path) {
1100 unlink($path) or die "unlink($path): $!";
1103 $! = $errno; # allow interpolation to stringify in die
1104 die "connect($path): $!";
1106 umask(077) // die("umask(077): $!");
1107 bind($listener, $addr) or die "bind($path): $!";
1108 listen($listener, 1024) or die "listen: $!";
1111 my @st = stat($path) or die "stat($path): $!";
1112 my $dev_ino_expect = pack('dd', $st[0], $st[1]); # dev+ino
1113 local $oldset = PublicInbox::DS::block_signals();
1115 $send_cmd = PublicInbox::Spawn->can('send_cmd4');
1116 $recv_cmd = PublicInbox::Spawn->can('recv_cmd4') // do {
1117 require PublicInbox::CmdIPC4;
1118 $send_cmd = PublicInbox::CmdIPC4->can('send_cmd4');
1119 PublicInbox::CmdIPC4->can('recv_cmd4');
1122 $recv_cmd or die <<"";
1123 (Socket::MsgHdr || Inline::C) missing/unconfigured (narg=$narg);
1125 require PublicInbox::Listener;
1126 require PublicInbox::EOFpipe;
1127 (-p STDOUT) or die "E: stdout must be a pipe\n";
1128 open(STDIN, '+>>', $errors_log) or die "open($errors_log): $!";
1129 STDIN->autoflush(1);
1130 dump_and_clear_log("from previous daemon process:\n");
1131 POSIX::setsid() > 0 or die "setsid: $!";
1132 my $pid = fork // die "fork: $!";
1134 $0 = "lei-daemon $path";
1136 $listener->blocking(0);
1138 my $pil = PublicInbox::Listener->new($listener, \&accept_dispatch);
1140 pipe(my ($eof_r, $eof_w)) or die "pipe: $!";
1141 PublicInbox::EOFpipe->new($eof_r, \&noop, undef);
1143 $exit_code //= shift;
1144 my $lis = $pil or exit($exit_code);
1145 # closing eof_w triggers \&noop wakeup
1146 $listener = $eof_w = $pil = $path = undef;
1147 $lis->close; # DS::close
1148 PublicInbox::DS->SetLoopTimeout(1000);
1152 CHLD => \&PublicInbox::DS::enqueue_reap,
1160 my $sigfd = PublicInbox::Sigfd->new($sig, SFD_NONBLOCK);
1161 local @SIG{keys %$sig} = values(%$sig) unless $sigfd;
1163 local $SIG{PIPE} = 'IGNORE';
1164 if ($sigfd) { # TODO: use inotify/kqueue to detect unlinked sockets
1166 PublicInbox::DS->SetLoopTimeout(5000);
1168 # wake up every second to accept signals if we don't
1169 # have signalfd or IO::KQueue:
1170 PublicInbox::DS::sig_setmask($oldset);
1171 PublicInbox::DS->SetLoopTimeout(1000);
1173 PublicInbox::DS->SetPostLoopCallback(sub {
1174 my ($dmap, undef) = @_;
1175 if (@st = defined($path) ? stat($path) : ()) {
1176 if ($dev_ino_expect ne pack('dd', $st[0], $st[1])) {
1177 warn "$path dev/ino changed, quitting\n";
1180 } elsif (defined($path)) { # ENOENT is common
1181 warn "stat($path): $!, quitting ...\n" if $! != ENOENT;
1185 return 1 if defined($path);
1188 for my $s (values %$dmap) {
1189 $s->can('busy') or next;
1190 if ($s->busy($now)) {
1196 $n; # true: continue, false: stop
1199 # STDIN was redirected to /dev/null above, closing STDERR and
1200 # STDOUT will cause the calling `lei' client process to finish
1201 # reading the <$daemon> pipe.
1202 local $SIG{__WARN__} = sub {
1203 $current_lei ? err($current_lei, @_) : warn(
1204 strftime('%Y-%m-%dT%H:%M:%SZ', gmtime(time))," $$ ", @_);
1206 open STDERR, '>&STDIN' or die "redirect stderr failed: $!";
1207 open STDOUT, '>&STDIN' or die "redirect stdout failed: $!";
1208 # $daemon pipe to `lei' closed, main loop begins:
1209 PublicInbox::DS->EventLoop;
1210 exit($exit_code // 0);
1213 sub busy { 1 } # prevent daemon-shutdown if client is connected
1215 # for users w/o Socket::Msghdr installed or Inline::C enabled
1217 my ($main_pkg) = @_;
1218 my $exit = $main_pkg->can('exit'); # caller may override exit()
1219 local $quit = $exit if $exit;
1221 umask(077) // die("umask(077): $!");
1229 dispatch($self, @ARGV);
1230 x_it($self, $self->{child_error}) if $self->{child_error};
1233 # ensures stdout hits the FS before sock disconnects so a client
1234 # can immediately reread it
1237 $self->{1}->autoflush(1) if $self->{1};
1240 my $oneshot_pids = delete $self->{"pid.$self.$$"} or return;
1241 waitpid($_, 0) for keys %$oneshot_pids;
1242 $? = $err if $err; # preserve ->fail or ->x_it code