1 # Copyright (C) all contributors <meta@public-inbox.org>
2 # License: AGPL-3.0+ <https://www.gnu.org/licenses/agpl-3.0.txt>
4 # handles "lei q" command and provides internals for
5 # several other sub-commands (up, lcat, ...)
6 package PublicInbox::LeiQuery;
9 use PublicInbox::OverIdx;
11 sub prep_ext { # externals_each callback
12 my ($lxs, $exclude, $loc) = @_;
13 $lxs->prepare_external($loc) unless $exclude->{$loc};
16 sub _start_query { # used by "lei q" and "lei up"
18 require PublicInbox::LeiOverview;
19 PublicInbox::LeiOverview->new($self) or return;
20 my $opt = $self->{opt};
21 PublicInbox::OverIdx::fork_ok($opt);
22 my ($xj, $mj) = split(/,/, $opt->{jobs} // '');
23 (defined($xj) && $xj ne '' && $xj !~ /\A[1-9][0-9]*\z/) and
24 die "`$xj' search jobs must be >= 1\n";
25 my $lxs = $self->{lxs};
26 $xj ||= $lxs->concurrency($opt); # allow: "--jobs ,$WRITER_ONLY"
27 my $nproc = $lxs->detect_nproc || 1; # don't memoize, schedtool(1) exists
28 $xj = $nproc if $xj > $nproc;
29 $lxs->{-wq_nr_workers} = $xj;
30 (defined($mj) && $mj !~ /\A[1-9][0-9]*\z/) and
31 die "`$mj' writer jobs must be >= 1\n";
32 my $l2m = $self->{l2m};
33 # we use \1 (a ref) to distinguish between default vs. user-supplied
34 if ($l2m && grep { $opt->{$_} //= \1 } (qw(mail-sync import-remote
36 $self->_lei_store(1)->write_prepare($self);
37 if ($opt->{'mail-sync'}) {
38 my $lms = $l2m->{-lms_rw} = $self->lms(1);
39 $lms->lms_write_prepare->lms_pause; # just create
42 $l2m and $l2m->{-wq_nr_workers} //= $mj //
43 int($nproc * 0.75 + 0.5); # keep some CPU for git
45 # descending docid order is cheapest, MUA controls sorting order
46 $self->{mset_opt}->{relevance} //= -2 if $l2m || $opt->{threads};
48 my $tot = $self->{mset_opt}->{total} //= $self->{opt}->{limit} // 10000;
49 $self->{mset_opt}->{limit} = $tot > 10000 ? 10000 : $tot;
50 $self->{mset_opt}->{offset} //= 0;
51 $self->{mset_opt}->{threads} //= $opt->{threads};
54 require PublicInbox::LeiAuth;
55 $self->{auth} = PublicInbox::LeiAuth->new
57 $lxs->do_query($self);
60 sub qstr_add { # PublicInbox::InputPipe::consume callback for --stdin
61 my ($lei) = @_; # $_[1] = $rbuf
62 $_[1] // $lei->fail("error reading stdin: $!");
63 return $lei->{mset_opt}->{qstr} .= $_[1] if $_[1] ne '';
66 $lei->{mset_opt}->{q_raw} = $lei->{mset_opt}->{qstr};
67 $lei->{lse}->query_approxidate($lei->{lse}->git,
68 $lei->{mset_opt}->{qstr});
76 require PublicInbox::LeiXSearch;
77 # prepare any number of LeiXSearch || LeiSearch || Inbox || URL
78 my $lxs = $self->{lxs} = PublicInbox::LeiXSearch->new;
79 my $opt = $self->{opt};
80 my @only = @{$opt->{only} // []};
81 # --local is enabled by default unless --only is used
82 # we'll allow "--only $LOCATION --local"
83 my $sto = $self->_lei_store(1);
84 $self->{lse} = $sto->search;
85 if ($opt->{'local'} //= scalar(@only) ? 0 : 1) {
86 $lxs->prepare_external($self->{lse});
90 my @loc = $self->get_externals($loc) or return;
91 $lxs->prepare_external($_) for @loc;
94 my (@ilocals, @iremotes);
95 for my $loc (@{$opt->{include} // []}) {
96 my @loc = $self->get_externals($loc) or return;
97 $lxs->prepare_external($_) for @loc;
98 @ilocals = @{$lxs->{locals} // []};
99 @iremotes = @{$lxs->{remotes} // []};
101 # --external is enabled by default, but allow --no-external
102 if ($opt->{external} //= 1) {
103 my $ex = $self->canonicalize_excludes($opt->{exclude});
104 $self->externals_each(\&prep_ext, $lxs, $ex);
105 $opt->{remote} //= !($lxs->locals - $opt->{'local'});
106 $lxs->{locals} = \@ilocals if !$opt->{'local'};
107 $lxs->{remotes} = \@iremotes if !$opt->{remote};
110 ($lxs->locals || $lxs->remotes) ? ($self->{lxs} = $lxs) :
111 die("no local or remote inboxes to search\n");
114 # the main "lei q SEARCH_TERMS" method
116 my ($self, @argv) = @_;
117 PublicInbox::Config->json; # preload before forking
118 my $lxs = lxs_prepare($self) or return;
119 $self->ale->refresh_externals($lxs, $self);
120 my $opt = $self->{opt};
122 $mset_opt{asc} = $opt->{'reverse'} ? 1 : 0;
123 if (defined(my $sort = $opt->{'sort'})) {
124 if ($sort eq 'relevance') {
125 $mset_opt{relevance} = 1;
126 } elsif ($sort eq 'docid') {
127 $mset_opt{relevance} = $mset_opt{asc} ? -1 : -2;
128 } elsif ($sort =~ /\Areceived(?:-?[aA]t)?\z/) {
131 die "unrecognized --sort=$sort\n";
133 $opt->{save} and return
134 $self->fail('--save and --sort are incompatible');
136 $self->{mset_opt} = \%mset_opt;
139 return $self->fail(<<'') if @argv;
140 no query allowed on command-line with --stdin
142 require PublicInbox::InputPipe;
143 PublicInbox::InputPipe::consume($self->{0}, \&qstr_add, $self);
146 chomp(@argv) and $self->qerr("# trailing `\\n' removed");
147 $mset_opt{q_raw} = [ @argv ]; # copy
149 $self->{lse}->query_argv_to_string($self->{lse}->git, \@argv);
153 # shell completion helper called by lei__complete
155 my ($self, @argv) = @_;
157 my $cb = $self->lazy_cb(qw(forget-external _complete_));
159 if ($argv[-1] =~ /\A(?:-I|(?:--(?:include|exclude|only)))\z/) {
160 my @c = $cb->($self, @cur);
161 # try basename match:
162 if (scalar(@cur) == 1 && index($cur[0], '/') < 0) {
163 my $all = $self->externals_each;
165 for my $loc (keys %$all) {
166 my $bn = (split(m!/!, $loc))[-1];
170 $bn{$_} == 1 && /\A\Q$cur[0]/
175 unshift(@cur, pop @argv);
180 # Stuff we may pass through to curl (as of 7.64.0), see curl manpage for
181 # details, so most options which make sense for HTTP/HTTPS (including proxy
182 # support for Tor and other methods of getting past weird networks).
183 # Most of these are untested by us, some may not make sense for our use case
184 # and typos below are likely.
185 # n.b. some short options (-$NUMBER) are not supported since they conflict
186 # with other "lei q" switches.
187 # FIXME: Getopt::Long doesn't easily let us support support options with
188 # '.' in them (e.g. --http1.1)
189 # TODO: should we depend on "-c http.*" options for things which have
190 # analogues in git(1)? that would reduce likelihood of conflicts with
191 # our other CLI options
192 # Note: some names are renamed to avoid potential conflicts,
193 # see %lei2curl in lib/PublicInbox/LeiCurl.pm
196 abstract-unix-socket=s anyauth basic cacert=s capath=s
197 cert-status cert-type cert=s ciphers=s
198 connect-timeout=s connect-to=s cookie-jar=s cookie=s crlfile=s
199 digest disable dns-interface=s dns-ipv4-addr=s dns-ipv6-addr=s
200 dns-servers=s doh-url=s egd-file=s engine=s false-start
201 happy-eyeballs-timeout-ms=s haproxy-protocol header=s@
202 http2-prior-knowledge http2 insecure
203 interface=s ipv4 ipv6 junk-session-cookies
204 key-type=s key=s limit-rate=s local-port=s location-trusted location
205 max-redirs=i max-time=s negotiate netrc-file=s netrc-optional netrc
206 no-alpn no-buffer no-npn no-sessionid noproxy=s ntlm-wb ntlm
207 pass=s pinnedpubkey=s post301 post302 post303 preproxy=s
208 proxy-anyauth proxy-basic proxy-cacert=s proxy-capath=s
209 proxy-cert-type=s proxy-cert=s proxy-ciphers=s proxy-crlfile=s
210 proxy-digest proxy-header=s@ proxy-insecure
211 proxy-key-type=s proxy-key proxy-negotiate proxy-ntlm proxy-pass=s
212 proxy-pinnedpubkey=s proxy-service-name=s proxy-ssl-allow-beast
213 proxy-tls13-ciphers=s proxy-tlsauthtype=s proxy-tlspassword=s
214 proxy-tlsuser=s proxy-tlsv1 proxy-user=s proxy=s
215 proxytunnel=s pubkey=s random-file=s referer=s resolve=s
216 retry-connrefused retry-delay=s retry-max-time=s retry=i
217 sasl-ir service-name=s socks4=s socks4a=s socks5-basic
218 socks5-gssapi-service-name=s socks5-gssapi socks5-hostname=s socks5=s
219 speed-limit speed-type ssl-allow-beast sslv2 sslv3
220 suppress-connect-headers tcp-fastopen tls-max=s
221 tls13-ciphers=s tlsauthtype=s tlspassword=s tlsuser=s
222 tlsv1 trace-ascii=s trace-time trace=s
223 unix-socket=s user-agent=s user=s