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 'forget-external' => [ 'LOCATION...|--prune',
149 'exclude further results from a publicinbox|extindex',
152 'ls-query' => [ '[FILTER...]', 'list saved search queries',
153 qw(name-only format|f=s), @c_opt ],
154 'rm-query' => [ 'QUERY_NAME', 'remove a saved search', @c_opt ],
155 'mv-query' => [ qw(OLD_NAME NEW_NAME), 'rename a saved search', @c_opt ],
157 'plonk' => [ '--threads|--from=IDENT',
158 'exclude mail matching From: or threads from non-Message-ID searches',
159 qw(stdin| threads|t from|f=s mid=s oid=s), @c_opt ],
160 'mark' => [ 'KEYWORDS...',
161 'set/unset keywords on message(s)',
162 qw(stdin| in-format|F=s input|i=s@ oid=s@ mid=s@), @c_opt,
163 pass_through('-kw:foo for delete') ],
164 'forget' => [ '[--stdin|--oid=OID|--by-mid=MID]',
165 "exclude message(s) on stdin from `q' search results",
166 qw(stdin| oid=s exact by-mid|mid:s), @c_opt ],
168 'purge-mailsource' => [ 'LOCATION|--all',
169 'remove imported messages from IMAP, Maildirs, and MH',
170 qw(exact! all jobs:i indexed), @c_opt ],
172 # code repos are used for `show' to solve blobs from patch mails
173 'add-coderepo' => [ 'DIRNAME', 'add or set priority of a git code repo',
174 qw(boost=i), @c_opt ],
175 'ls-coderepo' => [ '[FILTER_TERMS...]',
176 'list known code repos', qw(format|f=s z), @c_opt ],
177 'forget-coderepo' => [ 'DIRNAME',
178 'stop using repo to solve blobs from patches',
181 'add-watch' => [ 'LOCATION', 'watch for new messages and flag changes',
182 qw(import! kw|keywords|flags! interval=s recursive|r
183 exclude=s include=s), @c_opt ],
184 'ls-watch' => [ '[FILTER...]', 'list active watches with numbers and status',
185 qw(format|f=s z), @c_opt ],
186 'pause-watch' => [ '[WATCH_NUMBER_OR_FILTER]', qw(all local remote), @c_opt ],
187 'resume-watch' => [ '[WATCH_NUMBER_OR_FILTER]', qw(all local remote), @c_opt ],
188 'forget-watch' => [ '{WATCH_NUMBER|--prune}', 'stop and forget a watch',
191 'import' => [ 'LOCATION...|--stdin',
192 'one-time import/update from URL or filesystem',
193 qw(stdin| offset=i recursive|r exclude=s include|I=s
194 lock=s@ in-format|F=s kw|keywords|flags! verbose|v+), @c_opt ],
195 'convert' => [ 'LOCATION...|--stdin',
196 'one-time conversion from URL or filesystem to another format',
197 qw(stdin| in-format|F=s out-format|f=s output|mfolder|o=s
198 lock=s@ kw|keywords|flags!), @c_opt ],
199 'p2q' => [ 'FILE|COMMIT_OID|--stdin',
200 "use a patch to generate a query for `lei q --stdin'",
201 qw(stdin| want|w=s@ uri debug), @c_opt ],
202 'config' => [ '[...]', sub {
203 'git-config(1) wrapper for '._config_path($_[0]);
204 }, qw(config-file|system|global|file|f=s), # for conflict detection
205 qw(c=s@ C=s@), pass_through('git config') ],
206 'init' => [ '[DIRNAME]', sub {
207 "initialize storage, default: ".store_path($_[0]);
209 'daemon-kill' => [ '[-SIGNAL]', 'signal the lei-daemon',
210 # "-C DIR" conflicts with -CHLD, here, and chdir makes no sense, here
211 opt_dash('signal|s=s', '[0-9]+|(?:[A-Z][A-Z0-9]+)') ],
212 'daemon-pid' => [ '', 'show the PID of the lei-daemon' ],
213 'help' => [ '[SUBCOMMAND]', 'show help' ],
215 # XXX do we need this?
216 # 'git' => [ '[ANYTHING...]', 'git(1) wrapper', pass_through('git') ],
218 'reorder-local-store-and-break-history' => [ '[REFNAME]',
219 'rewrite git history in an attempt to improve compression',
222 # internal commands are prefixed with '_'
223 '_complete' => [ '[...]', 'internal shell completion helper',
224 pass_through('everything') ],
227 # switch descriptions, try to keep consistent across commands
228 # $spec: Getopt::Long option specification
229 # $spec => [@ALLOWED_VALUES (default is first), $description],
230 # $spec => $description
231 # "$SUB_COMMAND TAB $spec" => as above
232 my $stdin_formats = [ 'MAIL_FORMAT|eml|mboxrd|mboxcl2|mboxcl|mboxo',
233 'specify message input format' ];
234 my $ls_format = [ 'OUT|plain|json|null', 'listing output format' ];
236 # we use \x{a0} (non-breaking SP) to avoid wrapping in PublicInbox::LeiHelp
238 'help|h' => 'show this built-in help',
239 'c=s@' => [ 'NAME=VALUE', 'set config option' ],
240 'C=s@' => [ 'DIR', 'chdir to specify to directory' ],
241 'quiet|q' => 'be quiet',
242 'lock=s@' => [ 'METHOD|dotlock|fcntl|flock|none',
243 'mbox(5) locking method(s) to use (default: fcntl,dotlock)' ],
245 'globoff|g' => "do not match locations using '*?' wildcards ".
246 "and\xa0'[]'\x{a0}ranges",
247 'verbose|v+' => 'be more verbose',
248 'external!' => 'do not use externals',
249 'solve!' => 'do not attempt to reconstruct blobs from emails',
250 'torsocks=s' => ['VAL|auto|no|yes',
251 'whether or not to wrap git and curl commands with torsocks'],
252 'no-torsocks' => 'alias for --torsocks=no',
253 'save-as=s' => ['NAME', 'save a search terms by given name'],
254 'import-remote!' => 'do not memoize remote messages into local store',
256 'type=s' => [ 'any|mid|git', 'disambiguate type' ],
258 'dedupe|d=s' => ['STRATEGY|content|oid|mid|none',
259 'deduplication strategy'],
261 'return all messages in the same threads as the actual match(es)',
263 'want|w=s@' => [ 'PREFIX|dfpost|dfn', # common ones in help...
264 'search prefixes to extract (default: dfpost7)' ],
266 'alert=s@' => ['CMD,:WINCH,:bell,<any command>',
267 'run command(s) or perform ops when done writing to output ' .
268 '(default: ":WINCH,:bell" with --mua and Maildir/IMAP output, ' .
269 'nothing otherwise)' ],
271 'augment|a' => 'augment --output destination instead of clobbering',
273 'output|mfolder|o=s' => [ 'MFOLDER',
274 "destination (e.g.\xa0`/path/to/Maildir', ".
275 "or\xa0`-'\x{a0}for\x{a0}stdout)" ],
277 "MUA to run on --output Maildir or mbox (e.g.\xa0`mutt\xa0-f\xa0%f')" ],
279 'inbox-version=i' => [ 'NUM|1|2',
280 'force a public-inbox version with --mirror'],
281 'mirror=s' => [ 'URL', 'mirror a public-inbox'],
283 # public-inbox-index options
284 'fsync!' => 'speed up indexing after --mirror, risk index corruption',
285 'compact' => 'run compact index after mirroring',
286 'indexlevel|L=s' => [ 'LEVEL|full|medium|basic',
287 "indexlevel with --mirror (default: full)" ],
288 'max_size|max-size=s' => [ 'SIZE',
289 'do not index messages larger than SIZE (default: infinity)' ],
290 'batch_size|batch-size=s' => [ 'SIZE',
291 'flush changes to OS after given number of bytes (default: 1m)' ],
292 'sequential_shard|sequential-shard' =>
293 'index Xapian shards sequentially for slow storage',
295 'drop compatibility w/ public-inbox <1.6 to save ~1.5% space',
298 'OUT|maildir|mboxrd|mboxcl2|mboxcl|mboxo|html|json|jsonl|concatjson',
299 'specify output format, default depends on --output'],
300 'exclude=s@ q' => [ 'LOCATION',
301 'exclude specified external(s) from search' ],
302 'include|I=s@ q' => [ 'LOCATION',
303 'include specified external(s) in search' ],
304 'only=s@ q' => [ 'LOCATION',
305 'only use specified external(s) for search' ],
306 'jobs=s q' => [ '[SEARCH_JOBS][,WRITER_JOBS]',
307 'control number of search and writer jobs' ],
308 'jobs|j=i add-external' => 'set parallelism when indexing after --mirror',
310 'in-format|F=s' => $stdin_formats,
311 'format|f=s ls-query' => $ls_format,
312 'format|f=s ls-external' => $ls_format,
314 'limit|n=i@' => ['NUM', 'limit on number of matches (default: 10000)' ],
315 'offset=i' => ['OFF', 'search result offset (default: 0)'],
317 'sort|s=s' => [ 'VAL|received|relevance|docid',
318 "order of results is `--output'-dependent"],
319 'reverse|r' => 'reverse search results', # like sort(1)
321 'boost=i' => 'increase/decrease priority of results (default: 0)',
323 'local' => 'limit operations to the local filesystem',
324 'local!' => 'exclude results from the local filesystem',
325 'remote' => 'limit operations to those requiring network access',
326 'remote!' => 'prevent operations requiring network access',
328 'mid=s' => 'specify the Message-ID of a message',
329 'oid=s' => 'specify the git object ID of a message',
331 'recursive|r' => 'scan directories/mailboxes/newsgroups recursively',
332 'exclude=s' => 'exclude mailboxes/newsgroups based on pattern',
333 'include=s' => 'include mailboxes/newsgroups based on pattern',
335 'exact' => 'operate on exact header matches only',
336 'exact!' => 'rely on content match instead of exact header matches',
338 'by-mid|mid:s' => [ 'MID', 'match only by Message-ID, ignoring contents' ],
340 'kw|keywords|flags!' => 'disable/enable importing flags',
342 # xargs, env, use "-0", git(1) uses "-z". We support z|0 everywhere
343 'z|0' => 'use NUL \\0 instead of newline (CR) to delimit lines',
345 'signal|s=s' => [ 'SIG', 'signal to send lei-daemon (default: TERM)' ],
349 'leistore.dir' => 'top-level storage location',
352 my @WQ_KEYS = qw(lxs l2m imp mrr cnv p2q mark); # internal workers
354 # pronounced "exit": x_it(1 << 8) => exit(1); x_it(13) => SIGPIPE
356 my ($self, $code) = @_;
357 # make sure client sees stdout before exit
358 $self->{1}->autoflush(1) if $self->{1};
359 dump_and_clear_log();
360 if (my $s = $self->{pkt_op_p} // $self->{sock}) {
361 send($s, "x_it $code", MSG_EOR);
362 } elsif ($self->{oneshot}) {
363 # don't want to end up using $? from child processes
364 for my $f (@WQ_KEYS) {
365 my $wq = delete $self->{$f} or next;
368 # cleanup anything that has tempfiles or open file handles
370 delete @$self{qw(ovv dedupe sto cfg)};
371 if (my $signum = ($code & 127)) { # usually SIGPIPE (13)
372 $SIG{PIPE} = 'DEFAULT'; # $SIG{$signum} doesn't work
374 sleep(1) while 1; # wait for signal
378 } # else ignore if client disconnected
383 my $err = $self->{2} // ($self->{pgr} // [])->[2] // *STDERR{GLOB};
384 my @eor = (substr($_[-1]//'', -1, 1) eq "\n" ? () : ("\n"));
385 print $err @_, @eor and return;
386 my $old_err = delete $self->{2};
387 close($old_err) if $! == EPIPE && $old_err;
388 $err = $self->{2} = ($self->{pgr} // [])->[2] // *STDERR{GLOB};
389 print $err @_, @eor or print STDERR @_, @eor;
392 sub qerr ($;@) { $_[0]->{opt}->{quiet} or err(shift, @_) }
394 sub fail_handler ($;$$) {
395 my ($lei, $code, $io) = @_;
396 for my $f (@WQ_KEYS) {
397 my $wq = delete $lei->{$f} or next;
398 $wq->wq_wait_old(undef, $lei) if $wq->wq_kill_old; # lei-daemon
400 close($io) if $io; # needed to avoid warnings on SIGPIPE
401 x_it($lei, $code // (1 << 8));
404 sub sigpipe_handler { # handles SIGPIPE from @WQ_KEYS workers
405 fail_handler($_[0], 13, delete $_[0]->{1});
408 # PublicInbox::OnDestroy callback for SIGINT to take out the entire pgid
411 dwaitpid($pgid) if kill('-INT', $pgid);
415 my ($self, $buf, $exit_code) = @_;
416 err($self, $buf) if defined $buf;
417 # calls fail_handler:
418 send($self->{pkt_op_p}, '!', MSG_EOR) if $self->{pkt_op_p};
419 x_it($self, ($exit_code // 1) << 8);
425 return if print { $self->{1} // return } @_; # likely
426 return note_sigpipe($self, 1) if $! == EPIPE;
427 my $err = "error writing to output: $!";
432 sub puts ($;@) { out(shift, map { "$_\n" } @_) }
434 sub child_error { # passes non-fatal curl exit codes to user
435 my ($self, $child_error, $msg) = @_; # child_error is $?
436 $self->err($msg) if $msg;
437 if (my $s = $self->{pkt_op_p} // $self->{sock}) {
438 # send to the parent lei-daemon or to lei(1) client
439 send($s, "child_error $child_error", MSG_EOR);
440 } elsif (!$PublicInbox::DS::in_loop) {
441 $self->{child_error} = $child_error;
442 } # else noop if client disconnected
445 sub note_sigpipe { # triggers sigpipe_handler
446 my ($self, $fd) = @_;
447 close(delete($self->{$fd})); # explicit close silences Perl warning
448 send($self->{pkt_op_p}, '|', MSG_EOR) if $self->{pkt_op_p};
452 sub _lei_atfork_child {
453 my ($self, $persist) = @_;
454 # we need to explicitly close things which are on stack
456 my @io = delete @$self{qw(0 1 2 sock)};
457 unless ($self->{oneshot}) {
460 } else { # worker, Net::NNTP (Net::Cmd) uses STDERR directly
461 open STDERR, '+>&='.fileno($self->{2}) or warn "open $!";
464 delete @$self{qw(cnv mark imp)};
465 for (delete @$self{qw(3 old_1 au_done)}) {
466 close($_) if defined($_);
468 if (my $op_c = delete $self->{pkt_op_c}) {
469 close(delete $op_c->{sock});
471 if (my $pgr = delete $self->{pgr}) {
472 close($_) for (@$pgr[1,2]);
474 close $listener if $listener;
478 $quit = \&CORE::exit;
479 $current_lei = $persist ? undef : $self; # for SIG{__WARN__}
483 my ($lei, $wq, $ident, $jobs, $ops) = @_;
485 '!' => [ \&fail_handler, $lei ],
486 '|' => [ \&sigpipe_handler, $lei ],
487 'x_it' => [ \&x_it, $lei ],
488 'child_error' => [ \&child_error, $lei ],
491 $ops->{''} //= [ \&dclose, $lei ];
492 require PublicInbox::PktOp;
493 ($lei->{pkt_op_c}, $lei->{pkt_op_p}) = PublicInbox::PktOp->pair($ops);
494 $wq->wq_workers_start($ident, $jobs, $lei->oldset, { lei => $lei });
495 delete $lei->{pkt_op_p};
496 my $op = delete $lei->{pkt_op_c};
497 $lei->event_step_init;
498 # oneshot needs $op, daemon-mode uses DS->EventLoop to handle $op
499 $lei->{oneshot} ? $op : undef;
503 require PublicInbox::LeiHelp;
504 PublicInbox::LeiHelp::call($_[0], $_[1], \%CMD, \%OPTDESC);
508 my ($self, $cmd, $argv) = @_;
509 # allow _complete --help to complete, not show help
510 return 1 if substr($cmd, 0, 1) eq '_';
512 $OPT = $self->{opt} //= {};
513 my $info = $CMD{$cmd} // [ '[...]' ];
514 my ($proto, undef, @spec) = @$info;
515 my $glp = ref($spec[-1]) eq ref($GLP) ? pop(@spec) : $GLP;
516 push @spec, qw(help|h);
518 if ($spec[0] =~ s/\|\z//s) { # "stdin|" or "clear|" allows "-" alias
519 $lone_dash = $spec[0];
520 $OPT->{$spec[0]} = \(my $var);
521 push @spec, '' => \$var;
523 $glp->getoptionsfromarray($argv, $OPT, @spec) or
524 return _help($self, "bad arguments or options for $cmd");
525 return _help($self) if $OPT->{help};
527 push @$argv, @{$OPT->{-argv}} if defined($OPT->{-argv});
529 # "-" aliases "stdin" or "clear"
530 $OPT->{$lone_dash} = ${$OPT->{$lone_dash}} if defined $lone_dash;
533 my $POS_ARG = '[A-Z][A-Z0-9_]+';
535 my @args = split(/ /, $proto);
536 for my $var (@args) {
537 if ($var =~ /\A$POS_ARG\.\.\.\z/o) { # >= 1 args;
538 $inf = defined($argv->[$i]) and last;
539 $var =~ s/\.\.\.\z//;
540 $err = "$var not supplied";
541 } elsif ($var =~ /\A$POS_ARG\z/o) { # required arg at $i
542 $argv->[$i++] // ($err = "$var not supplied");
543 } elsif ($var =~ /\.\.\.\]\z/) { # optional args start
546 } elsif ($var =~ /\A\[-?$POS_ARG\]\z/) { # one optional arg
548 } elsif ($var =~ /\A.+?\|/) { # required FOO|--stdin
549 $inf = 1 if index($var, '...') > 0;
550 my @or = split(/\|/, $var);
553 if ($o =~ /\A--([a-z0-9\-]+)/) {
554 $ok = defined($OPT->{$1});
556 } elsif (defined($argv->[$i])) {
560 } # else continue looping
564 $err = join(', ', @or) . " or $last must be set";
566 warn "BUG: can't parse `$var' in $proto";
570 if (!$inf && scalar(@$argv) > scalar(@args)) {
571 $err //= 'too many arguments';
573 $err ? fail($self, "usage: lei $cmd $proto\nE: $err") : 1;
576 sub _tmp_cfg { # for lei -c <name>=<value> ...
578 my $cfg = _lei_cfg($self, 1);
580 my $ft = File::Temp->new(TEMPLATE => 'lei_cfg-XXXX', TMPDIR => 1);
581 my $tmp = { '-f' => $ft->filename, -tmp => $ft };
583 print $ft <<EOM or return fail($self, "$tmp->{-f}: $!");
587 $tmp = $self->{cfg} = bless { %$cfg, %$tmp }, ref($cfg);
588 for (@{$self->{opt}->{c}}) {
589 /\A([^=\.]+\.[^=]+)(?:=(.*))?\z/ or return fail($self, <<EOM);
590 `-c $_' is not of the form -c <name>=<value>'
594 _config($self, '--add', $name, $value);
595 if (defined(my $v = $tmp->{$name})) {
596 if (ref($v) eq 'ARRAY') {
599 $tmp->{$name} = [ $v, $value ];
602 $tmp->{$name} = $value;
608 my ($self, $cmd, $pfx) = @_;
612 $cb = $self->can($pfx.$ucmd) and return $cb;
614 $base =~ s/_([a-z])/\u$1/g;
615 my $pkg = "PublicInbox::Lei\u$base";
616 ($INC{"PublicInbox/Lei\u$base.pm"} // eval("require $pkg")) ?
617 $pkg->can($pfx.$ucmd) : undef;
621 my ($self, $cmd, @argv) = @_;
622 local $current_lei = $self; # for __WARN__
623 dump_and_clear_log("from previous run\n");
624 return _help($self, 'no command given') unless defined($cmd);
625 # do not support Getopt bundling for this
626 while ($cmd eq '-C' || $cmd eq '-c') {
627 my $v = shift(@argv) // return fail($self, $cmd eq '-C' ?
628 '-C DIRECTORY' : '-c <name>=<value>');
629 push @{$self->{opt}->{substr($cmd, 1, 1)}}, $v;
630 $cmd = shift(@argv) // return _help($self, 'no command given');
632 if (my $cb = lazy_cb(__PACKAGE__, $cmd, 'lei_')) {
633 optparse($self, $cmd, \@argv) or return;
634 $self->{opt}->{c} and (_tmp_cfg($self) // return);
635 if (my $chdir = $self->{opt}->{C}) {
636 for my $d (@$chdir) {
637 next if $d eq ''; # same as git(1)
638 chdir $d or return fail($self, "cd $d: $!");
642 } elsif (grep(/\A-/, $cmd, @argv)) { # --help or -h only
643 $GLP->getoptionsfromarray([$cmd, @argv], {}, qw(help|h C=s@))
644 or return _help($self, 'bad arguments or options');
647 fail($self, "`$cmd' is not an lei command");
652 my ($self, $creat) = @_;
653 return $self->{cfg} if $self->{cfg};
654 my $f = _config_path($self);
656 my $cur_st = @st ? pack('dd', $st[10], $st[7]) : ''; # 10:ctime, 7:size
658 if (my $cfg = $PATH2CFG{$f}) { # reuse existing object in common case
659 return ($self->{cfg} = $cfg) if $cur_st eq $cfg->{-st};
660 ($sto, $sto_dir) = @$cfg{qw(-lei_store leistore.dir)};
665 return bless {}, 'PublicInbox::Config';
667 my (undef, $cfg_dir, undef) = File::Spec->splitpath($f);
668 -d $cfg_dir or mkpath($cfg_dir) or die "mkpath($cfg_dir): $!\n";
669 open my $fh, '>>', $f or die "open($f): $!\n";
670 @st = stat($fh) or die "fstat($f): $!\n";
671 $cur_st = pack('dd', $st[10], $st[7]);
672 qerr($self, "# $f created") if $self->{cmd} ne 'config';
674 my $cfg = PublicInbox::Config::git_config_dump($f);
675 bless $cfg, 'PublicInbox::Config';
676 $cfg->{-st} = $cur_st;
678 if ($sto && File::Spec->canonpath($sto_dir) eq
679 File::Spec->canonpath($cfg->{'leistore.dir'})) {
680 $cfg->{-lei_store} = $sto;
682 $self->{cfg} = $PATH2CFG{$f} = $cfg;
685 sub _lei_store ($;$) {
686 my ($self, $creat) = @_;
687 my $cfg = _lei_cfg($self, $creat);
688 $cfg->{-lei_store} //= do {
689 require PublicInbox::LeiStore;
690 my $dir = $cfg->{'leistore.dir'};
691 $dir //= $creat ? store_path($self) : return;
692 PublicInbox::LeiStore->new($dir, { creat => $creat });
697 my ($self, @argv) = @_;
701 my ($self, @argv) = @_;
702 my %env = (%{$self->{env}}, GIT_CONFIG => undef);
703 my $cfg = _lei_cfg($self, 1);
704 my $cmd = [ qw(git config -f), $cfg->{'-f'}, @argv ];
705 my %rdr = map { $_ => $self->{$_} } (0..2);
706 waitpid(spawn($cmd, \%env, \%rdr), 0);
710 my ($self, @argv) = @_;
711 $self->{opt}->{'config-file'} and return fail $self,
712 "config file switches not supported by `lei config'";
714 x_it($self, $?) if $?;
718 my ($self, $dir) = @_;
719 my $cfg = _lei_cfg($self, 1);
720 my $cur = $cfg->{'leistore.dir'};
721 $dir //= store_path($self);
722 $dir = rel2abs($self, $dir);
723 my @cur = stat($cur) if defined($cur);
724 $cur = File::Spec->canonpath($cur // $dir);
725 my @dir = stat($dir);
726 my $exists = "# leistore.dir=$cur already initialized" if @dir;
729 _lei_store($self, 1)->done;
730 return qerr($self, $exists);
733 # some folks like symlinks and bind mounts :P
734 if (@dir && "@cur[1,0]" eq "@dir[1,0]") {
735 lei_config($self, 'leistore.dir', $dir);
736 _lei_store($self, 1)->done;
737 return qerr($self, "$exists (as $cur)");
739 return fail($self, <<"");
740 E: leistore.dir=$cur already initialized and it is not $dir
743 lei_config($self, 'leistore.dir', $dir);
744 _lei_store($self, 1)->done;
745 $exists //= "# leistore.dir=$dir newly initialized";
746 return qerr($self, $exists);
749 sub lei_daemon_pid { puts shift, $$ }
751 sub lei_daemon_kill {
753 my $sig = $self->{opt}->{signal} // 'TERM';
754 kill($sig, $$) or fail($self, "kill($sig, $$): $!");
757 sub lei_help { _help($_[0]) }
759 # Shell completion helper. Used by lei-completion.bash and hopefully
760 # other shells. Try to do as much here as possible to avoid redundancy
761 # and improve maintainability.
763 my ($self, @argv) = @_; # argv = qw(lei and any other args...)
764 shift @argv; # ignore "lei", the entire command is sent
765 @argv or return puts $self, grep(!/^_/, keys %CMD), qw(--help -h -C);
766 my $cmd = shift @argv;
767 my $info = $CMD{$cmd} // do { # filter matching commands
768 @argv or puts $self, grep(/\A\Q$cmd\E/, keys %CMD);
771 my ($proto, undef, @spec) = @$info;
773 my $re = defined($cur) ? qr/\A\Q$cur\E/ : qr/./;
774 if (substr(my $_cur = $cur // '-', 0, 1) eq '-') { # --switches
775 # gross special case since the only git-config options
776 # Consider moving to a table if we need more special cases
777 # we use Getopt::Long for are the ones we reject, so these
778 # are the ones we don't reject:
779 if ($cmd eq 'config') {
780 puts $self, grep(/$re/, keys %CONFIG_KEYS);
781 @spec = qw(add z|null get get-all unset unset-all
782 replace-all get-urlmatch
783 remove-section rename-section
784 name-only list|l edit|e
785 get-color-name get-colorbool);
788 # generate short/long names from Getopt::Long specs
789 puts $self, grep(/$re/, qw(--help -h -C), map {
790 if (s/[:=].+\z//) { # req/optional args, e.g output|o=i
791 } elsif (s/\+\z//) { # verbose|v+
793 # negation: solve! => no-solve|solve
794 s/([\w\-]+)/$1|no-$1/g
797 my $x = length > 1 ? "--$_" : "-$_";
798 $x eq $_cur ? () : $x;
799 } grep(!/_/, split(/\|/, $_, -1)) # help|h
800 } grep { $OPTDESC{"$_\t$cmd"} || $OPTDESC{$_} } @spec);
801 } elsif ($cmd eq 'config' && !@argv && !$CONFIG_KEYS{$cur}) {
802 puts $self, grep(/$re/, keys %CONFIG_KEYS);
805 # switch args (e.g. lei q -f mbox<TAB>)
806 if (($argv[-1] // $cur // '') =~ /\A--?([\w\-]+)\z/) {
807 my $opt = quotemeta $1;
809 my $v = $OPTDESC{$_};
810 my @v = ref($v) ? split(/\|/, $v->[0]) : ();
811 # get rid of ALL CAPS placeholder (e.g "OUT")
812 # (TODO: completion for external paths)
813 shift(@v) if scalar(@v) && uc($v[0]) eq $v[0];
815 } grep(/\A(?:[\w-]+\|)*$opt\b.*?(?:\t$cmd)?\z/, keys %OPTDESC);
817 if (my $cb = lazy_cb($self, $cmd, '_complete_')) {
818 puts $self, $cb->($self, @argv, $cur ? ($cur) : ());
820 # TODO: URLs, pathnames, OIDs, MIDs, etc... See optparse() for
825 my ($argv, $env) = @_;
826 my $argc = scalar @$argv;
827 my $buf = 'exec '.join("\0", scalar(@$argv), @$argv);
828 while (my ($k, $v) = each %$env) { $buf .= "\0$k=$v" };
834 my $mua = $self->{opt}->{mua} // return;
835 my $mfolder = $self->{ovv}->{dst};
836 my (@cmd, $replaced);
837 if ($mua =~ /\A(?:mutt|mailx|mail|neomutt)\z/) {
839 # TODO: help wanted: other common FOSS MUAs
841 require Text::ParseWords;
842 @cmd = Text::ParseWords::shellwords($mua);
843 # mutt uses '%f' for open-hook with compressed mbox, we follow
844 @cmd = map { $_ eq '%f' ? ($replaced = $mfolder) : $_ } @cmd;
846 push @cmd, $mfolder unless defined($replaced);
847 if (my $sock = $self->{sock}) { # lei(1) client process runs it
848 send($sock, exec_buf(\@cmd, {}), MSG_EOR);
849 } elsif ($self->{oneshot}) {
850 $self->{"pid.$self.$$"}->{spawn(\@cmd)} = \@cmd;
852 if ($self->{lxs} && $self->{au_done}) { # kick wait_startq
853 syswrite($self->{au_done}, 'q' x ($self->{lxs}->{jobs} // 0));
855 $self->{opt}->{quiet} = 1;
856 delete $self->{-progress};
857 delete $self->{opt}->{verbose};
860 sub send_exec_cmd { # tell script/lei to execute a command
861 my ($self, $io, $cmd, $env) = @_;
862 my $sock = $self->{sock} // die 'lei client gone';
863 my $fds = [ map { fileno($_) } @$io ];
864 $send_cmd->($sock, $fds, exec_buf($cmd, $env), MSG_EOR);
867 sub poke_mua { # forces terminal MUAs to wake up and hopefully notice new mail
869 my $alerts = $self->{opt}->{alert} // return;
870 while (my $op = shift(@$alerts)) {
871 if ($op eq ':WINCH') {
872 # hit the process group that started the MUA
874 send($self->{sock}, '-WINCH', MSG_EOR);
875 } elsif ($self->{oneshot}) {
878 } elsif ($op eq ':bell') {
880 } elsif ($op =~ /(?<!\\),/) { # bare ',' (not ',,')
881 push @$alerts, split(/(?<!\\),/, $op);
882 } elsif ($op =~ m!\A([/a-z0-9A-Z].+)!) {
883 my $cmd = $1; # run an arbitrary command
884 require Text::ParseWords;
885 $cmd = [ Text::ParseWords::shellwords($cmd) ];
886 if (my $s = $self->{sock}) {
887 send($s, exec_buf($cmd, {}), MSG_EOR);
888 } elsif ($self->{oneshot}) {
889 $self->{"pid.$self.$$"}->{spawn($cmd)} = $cmd;
892 err($self, "W: unsupported --alert=$op"); # non-fatal
897 my %path_to_fd = ('/dev/stdin' => 0, '/dev/stdout' => 1, '/dev/stderr' => 2);
898 $path_to_fd{"/dev/fd/$_"} = $path_to_fd{"/proc/self/fd/$_"} for (0..2);
900 my ($self, $mode, $path) = @_;
901 rel2abs($self, $path);
903 if (defined(my $fd = $path_to_fd{$path})) {
906 if ($path =~ m!\A/(?:dev|proc/self)/fd/[0-9]+\z!) {
907 return fail($self, "cannot open $path from daemon");
909 open my $fh, $mode, $path or return;
913 # caller needs to "-t $self->{1}" to check if tty
916 my $fh = popen_rd([qw(git var GIT_PAGER)]);
917 chomp(my $pager = <$fh> // '');
918 close($fh) or warn "`git var PAGER' error: \$?=$?";
919 return if $pager eq 'cat' || $pager eq '';
920 my $new_env = { LESS => 'FRX', LV => '-c' };
921 $new_env->{MORE} = 'FRX' if $^O eq 'freebsd';
922 pipe(my ($r, $wpager)) or return warn "pipe: $!";
923 my $rdr = { 0 => $r, 1 => $self->{1}, 2 => $self->{2} };
924 my $pgr = [ undef, @$rdr{1, 2} ];
925 my $env = $self->{env};
926 if ($self->{sock}) { # lei(1) process runs it
927 delete @$new_env{keys %$env}; # only set iff unset
928 send_exec_cmd($self, [ @$rdr{0..2} ], [$pager], $new_env);
929 } elsif ($self->{oneshot}) {
931 $self->{"pid.$self.$$"}->{spawn($cmd, $new_env, $rdr)} = $cmd;
933 die 'BUG: start_pager w/o socket';
935 $self->{1} = $wpager;
936 $self->{2} = $wpager if -t $self->{2};
937 $env->{GIT_PAGER_IN_USE} = 'true'; # we may spawn git
943 my $pgr = delete($self->{pgr}) or return;
944 $self->{2} = $pgr->[2];
945 # do not restore original stdout, just close it so we error out
946 close(delete($self->{1})) if $self->{1};
949 sub accept_dispatch { # Listener {post_accept} callback
950 my ($sock) = @_; # ignore other
952 my $self = bless { sock => $sock }, __PACKAGE__;
953 vec(my $rvec = '', fileno($sock), 1) = 1;
954 select($rvec, undef, undef, 60) or
955 return send($sock, 'timed out waiting to recv FDs', MSG_EOR);
956 # (4096 * 33) >MAX_ARG_STRLEN
957 my @fds = $recv_cmd->($sock, my $buf, 4096 * 33) or return; # EOF
958 if (scalar(@fds) == 4) {
960 my $fd = shift(@fds);
961 open($self->{$i}, '+<&=', $fd) and next;
962 send($sock, "open(+<&=$fd) (FD=$i): $!", MSG_EOR);
964 } elsif (!defined($fds[0])) {
965 warn(my $msg = "recv_cmd failed: $!");
966 return send($sock, $msg, MSG_EOR);
970 $self->{2}->autoflush(1); # keep stdout buffered until x_it|DESTROY
971 # $ENV_STR = join('', map { "\0$_=$ENV{$_}" } keys %ENV);
972 # $buf = "$argc\0".join("\0", @ARGV).$ENV_STR."\0\0";
973 substr($buf, -2, 2, '') eq "\0\0" or # s/\0\0\z//
974 return send($sock, 'request command truncated', MSG_EOR);
975 my ($argc, @argv) = split(/\0/, $buf, -1);
977 my %env = map { split(/=/, $_, 2) } splice(@argv, $argc);
978 if (chdir($self->{3})) {
980 $self->{env} = \%env;
981 eval { dispatch($self, @argv) };
982 send($sock, $@, MSG_EOR) if $@;
984 send($sock, "fchdir: $!", MSG_EOR); # implicit close
990 delete $self->{-progress};
991 for my $f (@WQ_KEYS) {
992 my $wq = delete $self->{$f} or next;
994 $wq->wq_close(0, undef, $self);
995 } elsif ($wq->wq_kill_old) {
996 $wq->wq_wait_old(undef, $self);
999 close(delete $self->{1}) if $self->{1}; # may reap_compress
1000 if (my $sto = delete $self->{sto}) {
1001 $sto->ipc_do('done');
1003 $self->close if $self->{-event_init_done}; # PublicInbox::DS::close
1006 # for long-running results
1009 local %ENV = %{$self->{env}};
1010 my $sock = $self->{sock};
1011 local $current_lei = $self;
1013 while (my @fds = $recv_cmd->($sock, my $buf, 4096)) {
1014 if (scalar(@fds) == 1 && !defined($fds[0])) {
1015 return if $! == EAGAIN;
1016 next if $! == EINTR;
1017 last if $! == ECONNRESET;
1021 open my $rfh, '+<&=', $fd;
1023 die "unrecognized client signal: $buf";
1028 eval { $self->fail($err) };
1033 sub event_step_init {
1035 return if $self->{-event_init_done}++;
1036 if (my $sock = $self->{sock}) { # using DS->EventLoop
1037 $self->SUPER::new($sock, EPOLLIN|EPOLLET);
1043 sub oldset { $oldset }
1045 sub dump_and_clear_log {
1046 if (defined($errors_log) && -s STDIN && seek(STDIN, 0, SEEK_SET)) {
1048 unshift(@pfx, "$errors_log ") if @pfx;
1049 warn @pfx, do { local $/; <STDIN> };
1050 truncate(STDIN, 0) or warn "ftruncate ($errors_log): $!";
1054 # lei(1) calls this when it can't connect
1056 my ($path, $errno, $narg) = @_;
1057 local ($errors_log, $listener);
1058 ($errors_log) = ($path =~ m!\A(.+?/)[^/]+\z!);
1059 $errors_log .= 'errors.log';
1060 my $addr = pack_sockaddr_un($path);
1061 my $lk = bless { lock_path => $errors_log }, 'PublicInbox::Lock';
1063 socket($listener, AF_UNIX, SOCK_SEQPACKET, 0) or die "socket: $!";
1064 if ($errno == ECONNREFUSED || $errno == ENOENT) {
1065 return if connect($listener, $addr); # another process won
1066 if ($errno == ECONNREFUSED && -S $path) {
1067 unlink($path) or die "unlink($path): $!";
1070 $! = $errno; # allow interpolation to stringify in die
1071 die "connect($path): $!";
1073 umask(077) // die("umask(077): $!");
1074 bind($listener, $addr) or die "bind($path): $!";
1075 listen($listener, 1024) or die "listen: $!";
1078 my @st = stat($path) or die "stat($path): $!";
1079 my $dev_ino_expect = pack('dd', $st[0], $st[1]); # dev+ino
1080 local $oldset = PublicInbox::DS::block_signals();
1082 $send_cmd = PublicInbox::Spawn->can('send_cmd4');
1083 $recv_cmd = PublicInbox::Spawn->can('recv_cmd4') // do {
1084 require PublicInbox::CmdIPC4;
1085 $send_cmd = PublicInbox::CmdIPC4->can('send_cmd4');
1086 PublicInbox::CmdIPC4->can('recv_cmd4');
1089 $recv_cmd or die <<"";
1090 (Socket::MsgHdr || Inline::C) missing/unconfigured (narg=$narg);
1092 require PublicInbox::Listener;
1093 require PublicInbox::EOFpipe;
1094 (-p STDOUT) or die "E: stdout must be a pipe\n";
1095 open(STDIN, '+>>', $errors_log) or die "open($errors_log): $!";
1096 STDIN->autoflush(1);
1097 dump_and_clear_log("from previous daemon process:\n");
1098 POSIX::setsid() > 0 or die "setsid: $!";
1099 my $pid = fork // die "fork: $!";
1101 $0 = "lei-daemon $path";
1103 $listener->blocking(0);
1105 my $pil = PublicInbox::Listener->new($listener, \&accept_dispatch);
1107 pipe(my ($eof_r, $eof_w)) or die "pipe: $!";
1108 PublicInbox::EOFpipe->new($eof_r, \&noop, undef);
1110 $exit_code //= shift;
1111 my $lis = $pil or exit($exit_code);
1112 # closing eof_w triggers \&noop wakeup
1113 $listener = $eof_w = $pil = $path = undef;
1114 $lis->close; # DS::close
1115 PublicInbox::DS->SetLoopTimeout(1000);
1119 CHLD => \&PublicInbox::DS::enqueue_reap,
1127 my $sigfd = PublicInbox::Sigfd->new($sig, SFD_NONBLOCK);
1128 local @SIG{keys %$sig} = values(%$sig) unless $sigfd;
1130 local $SIG{PIPE} = 'IGNORE';
1131 if ($sigfd) { # TODO: use inotify/kqueue to detect unlinked sockets
1133 PublicInbox::DS->SetLoopTimeout(5000);
1135 # wake up every second to accept signals if we don't
1136 # have signalfd or IO::KQueue:
1137 PublicInbox::DS::sig_setmask($oldset);
1138 PublicInbox::DS->SetLoopTimeout(1000);
1140 PublicInbox::DS->SetPostLoopCallback(sub {
1141 my ($dmap, undef) = @_;
1142 if (@st = defined($path) ? stat($path) : ()) {
1143 if ($dev_ino_expect ne pack('dd', $st[0], $st[1])) {
1144 warn "$path dev/ino changed, quitting\n";
1147 } elsif (defined($path)) { # ENOENT is common
1148 warn "stat($path): $!, quitting ...\n" if $! != ENOENT;
1152 return 1 if defined($path);
1155 for my $s (values %$dmap) {
1156 $s->can('busy') or next;
1157 if ($s->busy($now)) {
1163 $n; # true: continue, false: stop
1166 # STDIN was redirected to /dev/null above, closing STDERR and
1167 # STDOUT will cause the calling `lei' client process to finish
1168 # reading the <$daemon> pipe.
1169 local $SIG{__WARN__} = sub {
1170 $current_lei ? err($current_lei, @_) : warn(
1171 strftime('%Y-%m-%dT%H:%M:%SZ', gmtime(time))," $$ ", @_);
1173 open STDERR, '>&STDIN' or die "redirect stderr failed: $!";
1174 open STDOUT, '>&STDIN' or die "redirect stdout failed: $!";
1175 # $daemon pipe to `lei' closed, main loop begins:
1176 PublicInbox::DS->EventLoop;
1177 exit($exit_code // 0);
1180 sub busy { 1 } # prevent daemon-shutdown if client is connected
1182 # for users w/o Socket::Msghdr installed or Inline::C enabled
1184 my ($main_pkg) = @_;
1185 my $exit = $main_pkg->can('exit'); # caller may override exit()
1186 local $quit = $exit if $exit;
1188 umask(077) // die("umask(077): $!");
1196 dispatch($self, @ARGV);
1197 x_it($self, $self->{child_error}) if $self->{child_error};
1200 # ensures stdout hits the FS before sock disconnects so a client
1201 # can immediately reread it
1204 $self->{1}->autoflush(1) if $self->{1};
1207 my $oneshot_pids = delete $self->{"pid.$self.$$"} or return;
1208 waitpid($_, 0) for keys %$oneshot_pids;
1209 $? = $err if $err; # preserve ->fail or ->x_it code