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 # handles lei <q|ls-query|rm-query|mv-query> commands
5 package PublicInbox::LeiQuery;
10 sub prep_ext { # externals_each callback
11 my ($lxs, $exclude, $loc) = @_;
12 $lxs->prepare_external($loc) unless $exclude->{$loc};
18 require PublicInbox::LeiAuth;
19 $self->{auth} = PublicInbox::LeiAuth->new
21 $self->{lxs}->do_query($self);
24 sub qstr_add { # PublicInbox::InputPipe::consume callback for --stdin
25 my ($self) = @_; # $_[1] = $rbuf
27 $_[1] eq '' and return eval {
28 $self->{lse}->query_approxidate($self->{lse}->git,
29 $self->{mset_opt}->{qstr});
32 $self->{mset_opt}->{qstr} .= $_[1];
34 $self->fail("error reading stdin: $!");
40 require PublicInbox::LeiXSearch;
41 # prepare any number of LeiXSearch || LeiSearch || Inbox || URL
42 my $lxs = $self->{lxs} = PublicInbox::LeiXSearch->new;
43 my $opt = $self->{opt};
44 my @only = @{$opt->{only} // []};
45 # --local is enabled by default unless --only is used
46 # we'll allow "--only $LOCATION --local"
47 my $sto = $self->_lei_store(1); # FIXME: should not create
48 $self->{lse} = $sto->search;
49 if ($opt->{'local'} //= scalar(@only) ? 0 : 1) {
50 $lxs->prepare_external($self->{lse});
54 my @loc = $self->get_externals($loc) or return;
55 $lxs->prepare_external($_) for @loc;
58 my (@ilocals, @iremotes);
59 for my $loc (@{$opt->{include} // []}) {
60 my @loc = $self->get_externals($loc) or return;
61 $lxs->prepare_external($_) for @loc;
62 @ilocals = @{$lxs->{locals} // []};
63 @iremotes = @{$lxs->{remotes} // []};
65 # --external is enabled by default, but allow --no-external
66 if ($opt->{external} //= 1) {
68 for my $loc (@{$opt->{exclude} // []}) {
69 my @l = $self->get_externals($loc, 1) or return;
72 my $ne = $self->externals_each(\&prep_ext, $lxs, \%x);
73 $opt->{remote} //= !($lxs->locals - $opt->{'local'});
74 if ($opt->{'local'}) {
75 $lxs->{remotes} = \@iremotes if !$opt->{remote};
77 $lxs->{locals} = \@ilocals;
81 ($lxs->locals || $lxs->remotes) ? ($self->{lxs} = $lxs) :
82 $self->fail('no local or remote inboxes to search');
85 # the main "lei q SEARCH_TERMS" method
87 my ($self, @argv) = @_;
88 require PublicInbox::LeiOverview;
89 require PublicInbox::V2Writable;
90 PublicInbox::Config->json; # preload before forking
91 PublicInbox::LeiOverview->new($self) or return;
92 my $lxs = lxs_prepare($self) or return;
93 $self->ale->refresh_externals($lxs);
94 my $opt = $self->{opt};
95 my ($xj, $mj) = split(/,/, $opt->{jobs} // '');
96 if (defined($xj) && $xj ne '' && $xj !~ /\A[1-9][0-9]*\z/) {
97 return $self->fail("`$xj' search jobs must be >= 1");
99 $xj ||= $lxs->concurrency($opt); # allow: "--jobs ,$WRITER_ONLY"
100 my $nproc = $lxs->detect_nproc // 1; # don't memoize, schedtool(1) exists
101 $xj = $nproc if $xj > $nproc;
102 $lxs->{-wq_nr_workers} = $xj;
103 if (defined($mj) && $mj !~ /\A[1-9][0-9]*\z/) {
104 return $self->fail("`$mj' writer jobs must be >= 1");
106 if ($self->{l2m} && ($opt->{'import-remote'} //= 1) |
107 # we use \1 (a ref) to distinguish between
108 # user-supplied and default value
109 (($opt->{'import-before'} //= \1) ? 1 : 0)) {
110 $self->_lei_store(1)->write_prepare($self);
112 $self->{l2m} and $self->{l2m}->{-wq_nr_workers} = $mj // do {
113 $mj = POSIX::lround($nproc * 3 / 4); # keep some CPU for git
117 my %mset_opt = map { $_ => $opt->{$_} } qw(threads limit offset);
118 $mset_opt{asc} = $opt->{'reverse'} ? 1 : 0;
119 $mset_opt{limit} //= 10000;
120 if (defined(my $sort = $opt->{'sort'})) {
121 if ($sort eq 'relevance') {
122 $mset_opt{relevance} = 1;
123 } elsif ($sort eq 'docid') {
124 $mset_opt{relevance} = $mset_opt{asc} ? -1 : -2;
125 } elsif ($sort =~ /\Areceived(?:-?[aA]t)?\z/) {
128 die "unrecognized --sort=$sort\n";
131 # descending docid order is cheapest, MUA controls sorting order
132 $mset_opt{relevance} //= -2 if $self->{l2m} || $opt->{threads};
133 $self->{mset_opt} = \%mset_opt;
136 return $self->fail(<<'') if @argv;
137 no query allowed on command-line with --stdin
139 require PublicInbox::InputPipe;
140 PublicInbox::InputPipe::consume($self->{0}, \&qstr_add, $self);
144 $self->{lse}->query_argv_to_string($self->{lse}->git, \@argv);
148 # shell completion helper called by lei__complete
150 my ($self, @argv) = @_;
151 my $ext = qr/\A(?:-I|(?:--(?:include|exclude|only)))\z/;
154 if ($argv[-1] =~ $ext) {
155 my @c = $self->_complete_forget_external(@cur);
156 # try basename match:
157 if (scalar(@cur) == 1 && index($cur[0], '/') < 0) {
158 my $all = $self->externals_each;
160 for my $loc (keys %$all) {
161 my $bn = (split(m!/!, $loc))[-1];
165 $bn{$_} == 1 && /\A\Q$cur[0]/
170 unshift(@cur, pop @argv);
175 # Stuff we may pass through to curl (as of 7.64.0), see curl manpage for
176 # details, so most options which make sense for HTTP/HTTPS (including proxy
177 # support for Tor and other methods of getting past weird networks).
178 # Most of these are untested by us, some may not make sense for our use case
179 # and typos below are likely.
180 # n.b. some short options (-$NUMBER) are not supported since they conflict
181 # with other "lei q" switches.
182 # FIXME: Getopt::Long doesn't easily let us support support options with
183 # '.' in them (e.g. --http1.1)
184 # TODO: should we depend on "-c http.*" options for things which have
185 # analogues in git(1)? that would reduce likelyhood of conflicts with
186 # our other CLI options
187 # Note: some names are renamed to avoid potential conflicts,
188 # see %lei2mail in lib/PublicInbox/LeiCurl.pm
191 abstract-unix-socket=s anyauth basic cacert=s capath=s
192 cert-status cert-type cert=s ciphers=s
193 connect-timeout=s connect-to=s cookie-jar=s cookie=s crlfile=s
194 digest disable dns-interface=s dns-ipv4-addr=s dns-ipv6-addr=s
195 dns-servers=s doh-url=s egd-file=s engine=s false-start
196 happy-eyeballs-timeout-ms=s haproxy-protocol header=s@
197 http2-prior-knowledge http2 insecure
198 interface=s ipv4 ipv6 junk-session-cookies
199 key-type=s key=s limit-rate=s local-port=s location-trusted location
200 max-redirs=i max-time=s negotiate netrc-file=s netrc-optional netrc
201 no-alpn no-buffer no-npn no-sessionid noproxy=s ntlm-wb ntlm
202 pass=s pinnedpubkey=s post301 post302 post303 preproxy=s
203 proxy-anyauth proxy-basic proxy-cacert=s proxy-capath=s
204 proxy-cert-type=s proxy-cert=s proxy-ciphers=s proxy-crlfile=s
205 proxy-digest proxy-header=s@ proxy-insecure
206 proxy-key-type=s proxy-key proxy-negotiate proxy-ntlm proxy-pass=s
207 proxy-pinnedpubkey=s proxy-service-name=s proxy-ssl-allow-beast
208 proxy-tls13-ciphers=s proxy-tlsauthtype=s proxy-tlspassword=s
209 proxy-tlsuser=s proxy-tlsv1 proxy-user=s proxy=s
210 proxytunnel=s pubkey=s random-file=s referer=s resolve=s
211 retry-connrefused retry-delay=s retry-max-time=s retry=i
212 sasl-ir service-name=s socks4=s socks4a=s socks5-basic
213 socks5-gssapi-service-name=s socks5-gssapi socks5-hostname=s socks5=s
214 speed-limit speed-type ssl-allow-beast sslv2 sslv3
215 suppress-connect-headers tcp-fastopen tls-max=s
216 tls13-ciphers=s tlsauthtype=s tlspassword=s tlsuser=s
217 tlsv1 trace-ascii=s trace-time trace=s
218 unix-socket=s user-agent=s user=s