X-Git-Url: http://www.git.stargrave.org/?a=blobdiff_plain;f=lib%2FPublicInbox%2FSearch.pm;h=82a6e5416240e19876203fc61fcd5b25d289d420;hb=feb4a75affbe5f940116dc4f6bcc5ddb52a110b4;hp=39b06b0afe6045fcd220eee77133a12144f4d476;hpb=7745ac38a4e2d4ae42e7192183f65c84a8e5662a;p=public-inbox.git diff --git a/lib/PublicInbox/Search.pm b/lib/PublicInbox/Search.pm index 39b06b0a..82a6e541 100644 --- a/lib/PublicInbox/Search.pm +++ b/lib/PublicInbox/Search.pm @@ -1,260 +1,220 @@ -# Copyright (C) 2015, all contributors +# Copyright (C) 2015 all contributors # License: AGPLv3 or later (https://www.gnu.org/licenses/agpl-3.0.txt) # based on notmuch, but with no concept of folders, files or flags +# +# Read-only search interface for use by the web and NNTP interfaces package PublicInbox::Search; use strict; use warnings; -use PublicInbox::SearchMsg; + +# values for searching +use constant TS => 0; # timestamp +use constant NUM => 1; # NNTP article number +use constant BYTES => 2; # :bytes as defined in RFC 3977 +use constant LINES => 3; # :lines as defined in RFC 3977 +use constant YYYYMMDD => 4; # for searching in the WWW UI + use Search::Xapian qw/:standard/; -require PublicInbox::View; -use Email::MIME; -use PublicInbox::MID qw/mid_clean mid_compressed/; +use PublicInbox::SearchMsg; +use PublicInbox::MIME; +use PublicInbox::MID qw/mid_clean id_compress/; + +# This is English-only, everything else is non-standard and may be confused as +# a prefix common in patch emails +our $REPLY_RE = qr/^re:\s+/i; +our $LANG = 'english'; use constant { - TS => 0, - SCHEMA_VERSION => 0, - LANG => 'english', + # SCHEMA_VERSION history + # 0 - initial + # 1 - subject_path is lower-cased + # 2 - subject_path is id_compress in the index, only + # 3 - message-ID is compressed if it includes '%' (hack!) + # 4 - change "Re: " normalization, avoid circular Reference ghosts + # 5 - subject_path drops trailing '.' + # 6 - preserve References: order in document data + # 7 - remove references and inreplyto terms + # 8 - remove redundant/unneeded document data + # 9 - disable Message-ID compression (SHA-1) + # 10 - optimize doc for NNTP overviews + # 11 - merge threads when vivifying ghosts + # 12 - change YYYYMMDD value column to numeric + # 13 - fix threading for empty References/In-Reply-To + # (commit 83425ef12e4b65cdcecd11ddcb38175d4a91d5a0) + # 14 - fix ghost root vivification + SCHEMA_VERSION => 14, + + # n.b. FLAG_PURE_NOT is expensive not suitable for a public website + # as it could become a denial-of-service vector QP_FLAGS => FLAG_PHRASE|FLAG_BOOLEAN|FLAG_LOVEHATE|FLAG_WILDCARD, }; # setup prefixes my %bool_pfx_internal = ( type => 'T', # "mail" or "ghost" - mid => 'Q', # uniQue id (Message-ID or mid_compressed) + thread => 'G', # newsGroup (or similar entity - e.g. a web forum name) ); my %bool_pfx_external = ( + # do we still need these? probably not.. path => 'XPATH', - thread => 'G', # newsGroup (or similar entity - e.g. a web forum name) - references => 'XREFS', - inreplyto => 'XIRT', + mid => 'Q', # uniQue id (Message-ID) ); my %prob_prefix = ( - subject => 'S', + # for mairix compatibility + s => 'S', + m => 'XMID', # 'mid:' (bool) is exact, 'm:' (prob) can do partial + f => 'A', + t => 'XTO', + tc => 'XTO XCC', + c => 'XCC', + tcf => 'XTO XCC A', + a => 'XTO XCC A', + b => 'XNQ XQUOT', + bs => 'XNQ XQUOT S', + n => 'XFN', + + q => 'XQUOT', + nq => 'XNQ', + + # default: + '' => 'XMID S A XNQ XQUOT XFN', ); +# not documenting m: and mid: for now, the using the URLs works w/o Xapian +our @HELP = ( + 's:' => 'match within Subject e.g. s:"a quick brown fox"', + 'd:' => < 'match within message body, including text attachments', + 'nq:' => 'match non-quoted text within message body', + 'q:' => 'match quoted text within message body', + 'n:' => 'match filename of attachment(s)', + 't:' => 'match within the To header', + 'c:' => 'match within the Cc header', + 'f:' => 'match within the From header', + 'a:' => 'match within the To, Cc, and From headers', + 'tc:' => 'match within the To and Cc headers', + 'bs:' => 'match within the Subject and body', +); +chomp @HELP; +# TODO: +# df (filenames from diff) +# da (diff a/ removed lines) +# db (diff b/ added lines) + my %all_pfx = (%bool_pfx_internal, %bool_pfx_external, %prob_prefix); sub xpfx { $all_pfx{$_[0]} } -our %PFX2TERM_RMAP; -while (my ($k, $v) = each %all_pfx) { - next if $prob_prefix{$k}; - $PFX2TERM_RMAP{$v} = $k; -} - my $mail_query = Search::Xapian::Query->new(xpfx('type') . 'mail'); -sub new { - my ($class, $git_dir, $writable) = @_; - # allow concurrent versions for easier rollback: - my $dir = "$git_dir/public-inbox/xapian" . SCHEMA_VERSION; - my $db; - - if ($writable) { # not used by the WWW interface - require Search::Xapian::WritableDatabase; - require File::Path; - File::Path::mkpath($dir); - $db = Search::Xapian::WritableDatabase->new($dir, - Search::Xapian::DB_CREATE_OR_OPEN); - } else { - $db = Search::Xapian::Database->new($dir); - } - bless { xdb => $db, git_dir => $git_dir }, $class; +sub xdir { + my (undef, $git_dir) = @_; + "$git_dir/public-inbox/xapian" . SCHEMA_VERSION; } -sub reopen { $_[0]->{xdb}->reopen } - -sub add_message { - my ($self, $mime) = @_; # mime = Email::MIME object - my $db = $self->{xdb}; - - my $doc_id; - my $mid = mid_clean($mime->header('Message-ID')); - $mid = mid_compressed($mid); - my $was_ghost = 0; - my $ct_msg = $mime->header('Content-Type') || 'text/plain'; - my $enc_msg = PublicInbox::View::enc_for($ct_msg); - - $db->begin_transaction; - eval { - my $smsg = $self->lookup_message($mid); - my $doc; - - if ($smsg) { - $smsg->ensure_metadata; - # convert a ghost to a regular message - # it will also clobber any existing regular message - $smsg->mime($mime); - $doc = $smsg->{doc}; - - my $type = xpfx('type'); - eval { - $doc->remove_term($type . 'ghost'); - $was_ghost = 1; - }; - - # probably does not exist: - eval { $doc->remove_term($type . 'mail') }; - $doc->add_term($type . 'mail'); - } else { - $smsg = PublicInbox::SearchMsg->new($mime); - $doc = $smsg->{doc}; - $doc->add_term(xpfx('mid') . $mid); - } - - my $subj = $smsg->subject; - - if (length $subj) { - $doc->add_term(xpfx('subject') . $subj); - - my $path = subject_path($subj); - $doc->add_term(xpfx('path') . $path); - } - - my $from = $smsg->from_name; - my $date = $smsg->date; - my $ts = Search::Xapian::sortable_serialise($smsg->ts); - $doc->add_value(PublicInbox::Search::TS, $ts); - - $doc->set_data($smsg->to_doc_data); - - my $tg = $self->term_generator; - - $tg->set_document($doc); - $tg->index_text($subj, 1, 'S') if $subj; - $tg->increase_termpos; - $tg->index_text($subj) if $subj; - $tg->increase_termpos; - - $tg->index_text($smsg->from->format); - $tg->increase_termpos; - - $mime->walk_parts(sub { - my ($part) = @_; - return if $part->subparts; # walk_parts already recurses - my $ct = $part->content_type || $ct_msg; - - # account for filter bugs... - $ct =~ m!\btext/plain\b!i or return; - - my $enc = PublicInbox::View::enc_for($ct, $enc_msg); - my (@orig, @quot); - foreach my $l (split(/\n/, $enc->decode($part->body))) { - if ($l =~ /^\s*>/) { - push @quot, $l; - } else { - push @orig, $l; - } - } - if (@quot) { - $tg->index_text(join("\n", @quot), 0); - $tg->increase_termpos; - } - if (@orig) { - $tg->index_text(join("\n", @orig)); - $tg->increase_termpos; - } - }); - - if ($was_ghost) { - $doc_id = $smsg->doc_id; - $self->link_message($smsg, 0); - $db->replace_document($doc_id, $doc); - } else { - $self->link_message($smsg, 0); - $doc_id = $db->add_document($doc); - } - }; - - if ($@) { - warn "failed to index message <$mid>: $@\n"; - $db->cancel_transaction; - } else { - $db->commit_transaction; - } - $doc_id; +sub new { + my ($class, $git_dir, $altid) = @_; + my $dir = $class->xdir($git_dir); + my $db = Search::Xapian::Database->new($dir); + bless { xdb => $db, git_dir => $git_dir, altid => $altid }, $class; } -# returns deleted doc_id on success, undef on missing -sub remove_message { - my ($self, $mid) = @_; - my $db = $self->{xdb}; - my $doc_id; - $mid = mid_clean($mid); - $mid = mid_compressed($mid); - - $db->begin_transaction; - eval { - $doc_id = $self->find_unique_doc_id('mid', $mid); - $db->delete_document($doc_id) if defined $doc_id; - }; - - if ($@) { - warn "failed to remove message <$mid>: $@\n"; - $db->cancel_transaction; - } else { - $db->commit_transaction; - } - $doc_id; -} +sub reopen { $_[0]->{xdb}->reopen } # read-only sub query { my ($self, $query_string, $opts) = @_; - my $query = $self->qp->parse_query($query_string, QP_FLAGS); - - $query = Search::Xapian::Query->new(OP_AND, $mail_query, $query); - $self->do_enquire($query, $opts); -} + my $query; -# given a message ID, get replies to a message -sub get_replies { - my ($self, $mid, $opts) = @_; - $mid = mid_clean($mid); - $mid = mid_compressed($mid); - my $qp = $self->qp; - my $irt = $qp->parse_query("inreplyto:$mid", 0); - my $ref = $qp->parse_query("references:$mid", 0); - my $query = Search::Xapian::Query->new(OP_OR, $irt, $ref); + $opts ||= {}; + unless ($query_string eq '') { + $query = $self->qp->parse_query($query_string, QP_FLAGS); + $opts->{relevance} = 1 unless exists $opts->{relevance}; + } - $self->do_enquire($query); + _do_enquire($self, $query, $opts); } sub get_thread { my ($self, $mid, $opts) = @_; my $smsg = eval { $self->lookup_message($mid) }; - return { count => 0, msgs => [] } unless $smsg; - my $qp = $self->qp; - my $qtid = $qp->parse_query('thread:'.$smsg->thread_id); - my $qsub = $qp->parse_query('path:'.$smsg->path); - my $query = Search::Xapian::Query->new(OP_OR, $qtid, $qsub); - $self->do_enquire($query); + return { total => 0, msgs => [] } unless $smsg; + my $qtid = Search::Xapian::Query->new(xpfx('thread').$smsg->thread_id); + my $path = $smsg->path; + if (defined $path && $path ne '') { + my $path = id_compress($smsg->path); + my $qsub = Search::Xapian::Query->new(xpfx('path').$path); + $qtid = Search::Xapian::Query->new(OP_OR, $qtid, $qsub); + } + $opts ||= {}; + $opts->{limit} ||= 1000; + + # always sort threads by timestamp, this makes life easier + # for the threading algorithm (in SearchThread.pm) + $opts->{asc} = 1; + + _do_enquire($self, $qtid, $opts); } -# private subs below +sub retry_reopen { + my ($self, $cb) = @_; + my $ret; + for (1..10) { + eval { $ret = $cb->() }; + return $ret unless $@; + # Exception: The revision being read has been discarded - + # you should call Xapian::Database::reopen() + if (ref($@) eq 'Search::Xapian::DatabaseModifiedError') { + reopen($self); + } else { + die; + } + } +} -sub do_enquire { +sub _do_enquire { my ($self, $query, $opts) = @_; - my $enquire = $self->enquire; + retry_reopen($self, sub { _enquire_once($self, $query, $opts) }); +} +sub _enquire_once { + my ($self, $query, $opts) = @_; + my $enquire = $self->enquire; + if (defined $query) { + $query = Search::Xapian::Query->new(OP_AND,$query,$mail_query); + } else { + $query = $mail_query; + } $enquire->set_query($query); - $enquire->set_sort_by_relevance_then_value(TS, 0); $opts ||= {}; + my $desc = !$opts->{asc}; + if ($opts->{relevance}) { + $enquire->set_sort_by_relevance_then_value(TS, $desc); + } elsif ($opts->{num}) { + $enquire->set_sort_by_value(NUM, 0); + } else { + $enquire->set_sort_by_value_then_relevance(TS, $desc); + } my $offset = $opts->{offset} || 0; my $limit = $opts->{limit} || 50; my $mset = $enquire->get_mset($offset, $limit); + return $mset if $opts->{mset}; my @msgs = map { PublicInbox::SearchMsg->load_doc($_->get_document); } $mset->items; - { count => $mset->get_matches_estimated, msgs => \@msgs } + { total => $mset->get_matches_estimated, msgs => \@msgs } } # read-write -sub stemmer { Search::Xapian::Stem->new(LANG) } +sub stemmer { Search::Xapian::Stem->new($LANG) } # read-only sub qp { @@ -269,113 +229,55 @@ sub qp { $qp->set_database($self->{xdb}); $qp->set_stemmer($self->stemmer); $qp->set_stemming_strategy(STEM_SOME); - $qp->add_valuerangeprocessor($self->ts_range_processor); - $qp->add_valuerangeprocessor($self->date_range_processor); + $qp->add_valuerangeprocessor( + Search::Xapian::NumberValueRangeProcessor->new(YYYYMMDD, 'd:')); while (my ($name, $prefix) = each %bool_pfx_external) { $qp->add_boolean_prefix($name, $prefix); } + # we do not actually create AltId objects, + # just parse the spec to avoid the extra DB handles for now. + if (my $altid = $self->{altid}) { + my $user_pfx = $self->{-user_pfx} ||= []; + for (@$altid) { + # $_ = 'serial:gmane:/path/to/gmane.msgmap.sqlite3' + /\Aserial:(\w+):/ or next; + my $pfx = $1; + push @$user_pfx, "$pfx:", < XGMANE + $qp->add_boolean_prefix($pfx, 'X'.uc($pfx)); + } + chomp @$user_pfx; + } + while (my ($name, $prefix) = each %prob_prefix) { - $qp->add_prefix($name, $prefix); + $qp->add_prefix($name, $_) foreach split(/ /, $prefix); } $self->{query_parser} = $qp; } -sub term_generator { # write-only - my ($self) = @_; - - my $tg = $self->{term_generator}; - return $tg if $tg; - - $tg = Search::Xapian::TermGenerator->new; - $tg->set_stemmer($self->stemmer); - - $self->{term_generator} = $tg; -} - -sub next_doc_id { $_[0]->{xdb}->get_lastdocid + 1 } - -# increments last_thread_id counter -# returns a 64-bit integer represented as a hex string -sub next_thread_id { - my ($self) = @_; - my $db = $self->{xdb}; - my $last_thread_id = int($db->get_metadata('last_thread_id') || 0); - - $db->set_metadata('last_thread_id', ++$last_thread_id); - - $last_thread_id; -} - -sub ts_range_processor { - $_[0]->{tsrp} ||= Search::Xapian::NumberValueRangeProcessor->new(TS); -} - -sub date_range_processor { - $_[0]->{drp} ||= Search::Xapian::DateValueRangeProcessor->new(TS); +sub num_range_processor { + $_[0]->{nrp} ||= Search::Xapian::NumberValueRangeProcessor->new(NUM); } -sub link_message { - my ($self, $smsg, $is_ghost) = @_; - - if ($is_ghost) { - $smsg->ensure_metadata; - } else { - $self->link_message_to_parents($smsg); - } -} - -sub link_message_to_parents { - my ($self, $smsg) = @_; - my $doc = $smsg->{doc}; - my $mid = mid_compressed($smsg->mid); - my $mime = $smsg->mime; - my $refs = $mime->header('References'); - my @refs = $refs ? ($refs =~ /<([^>]+)>/g) : (); - my $irt = $mime->header('In-Reply-To'); - if ($irt) { - if ($irt =~ /<([^>]+)>/) { - $irt = $1; - } - push @refs, $irt; - } +# only used for NNTP server +sub query_xover { + my ($self, $beg, $end, $offset) = @_; + my $qp = Search::Xapian::QueryParser->new; + $qp->set_database($self->{xdb}); + $qp->add_valuerangeprocessor($self->num_range_processor); + my $query = $qp->parse_query("$beg..$end", QP_FLAGS); - my $tid; - if (@refs) { - @refs = map { mid_compressed($_) } @refs; - my %uniq; - @refs = grep { !$uniq{$_}++ } @refs; # uniq - - $doc->add_term(xpfx('inreplyto') . $refs[-1]); - - my $ref_pfx = xpfx('references'); - - # first ref *should* be the thread root, - # but we can never trust clients to do the right thing - my $ref = shift @refs; - $doc->add_term($ref_pfx . $ref); - $tid = $self->_resolve_mid_to_tid($ref); - - # the rest of the refs should point to this tid: - foreach $ref (@refs) { - $doc->add_term($ref_pfx . $ref); - my $ptid = $self->_resolve_mid_to_tid($ref); - if ($tid ne $ptid) { - $self->merge_threads($tid, $ptid); - } - } - } else { - $tid = $self->next_thread_id; - } - $doc->add_term(xpfx('thread') . $tid); + _do_enquire($self, $query, {num => 1, limit => 200, offset => $offset}); } sub lookup_message { my ($self, $mid) = @_; $mid = mid_clean($mid); - $mid = mid_compressed($mid); my $doc_id = $self->find_unique_doc_id('mid', $mid); my $smsg; @@ -383,11 +285,19 @@ sub lookup_message { # raises on error: my $doc = $self->{xdb}->get_document($doc_id); $smsg = PublicInbox::SearchMsg->wrap($doc, $mid); - $smsg->doc_id($doc_id); + $smsg->{doc_id} = $doc_id; } $smsg; } +sub lookup_mail { # no ghosts! + my ($self, $mid) = @_; + retry_reopen($self, sub { + my $smsg = lookup_message($self, $mid) or return; + PublicInbox::SearchMsg->load_doc($smsg->{doc}); + }); +} + sub find_unique_doc_id { my ($self, $term, $value) = @_; @@ -418,80 +328,23 @@ sub find_doc_ids_for_term { ($db->postlist_begin($term), $db->postlist_end($term)); } -# this will create a ghost as necessary -sub _resolve_mid_to_tid { - my ($self, $mid) = @_; - - my $smsg = $self->lookup_message($mid) || $self->create_ghost($mid); - $smsg->thread_id; -} - -sub create_ghost { - my ($self, $mid, $tid) = @_; - - $mid = mid_compressed($mid); - $tid = $self->next_thread_id unless defined $tid; - - my $doc = Search::Xapian::Document->new; - $doc->add_term(xpfx('mid') . $mid); - $doc->add_term(xpfx('thread') . $tid); - $doc->add_term(xpfx('type') . 'ghost'); - - my $smsg = PublicInbox::SearchMsg->wrap($doc, $mid); - $self->link_message($smsg, 1); - $self->{xdb}->add_document($doc); - - $smsg; -} - -sub merge_threads { - my ($self, $winner_tid, $loser_tid) = @_; - my ($head, $tail) = $self->find_doc_ids('thread', $loser_tid); - my $thread_pfx = xpfx('thread'); - my $db = $self->{xdb}; - - for (; $head != $tail; $head->inc) { - my $docid = $head->get_docid; - my $doc = $db->get_document($docid); - $doc->remove_term($thread_pfx . $loser_tid); - $doc->add_term($thread_pfx . $winner_tid); - $db->replace_document($docid, $doc); - } -} - # normalize subjects so they are suitable as pathnames for URLs +# XXX: consider for removal sub subject_path { - my ($subj) = @_; - - $subj =~ s/\A\s+//; - $subj =~ s/\s+\z//; - $subj =~ s/^(?:re|aw):\s*//i; # remove reply prefix (aw: German) + my $subj = pop; + $subj = subject_normalized($subj); $subj =~ s![^a-zA-Z0-9_\.~/\-]+!_!g; - $subj; -} - -sub do_cat_mail { - my ($git, $blob) = @_; - my $mime = eval { - my $str = $git->cat_file($blob); - Email::MIME->new($str); - }; - $@ ? undef : $mime; -} - -sub index_blob { - my ($self, $git, $blob) = @_; - my $mime = do_cat_mail($git, $blob) or return; - eval { $self->add_message($mime) }; - warn "W: index_blob $blob: $@\n" if $@; + lc($subj); } -sub unindex_blob { - my ($self, $git, $blob) = @_; - my $mime = do_cat_mail($git, $blob) or return; - my $mid = $mime->header('Message-ID'); - eval { $self->remove_message($mid) } if defined $mid; - warn "W: unindex_blob $blob: $@\n" if $@; +sub subject_normalized { + my $subj = pop; + $subj =~ s/\A\s+//s; # no leading space + $subj =~ s/\s+\z//s; # no trailing space + $subj =~ s/\s+/ /gs; # no redundant spaces + $subj =~ s/\.+\z//; # no trailing '.' + $subj =~ s/$REPLY_RE//igo; # remove reply prefix + $subj; } sub enquire { @@ -499,42 +352,14 @@ sub enquire { $self->{enquire} ||= Search::Xapian::Enquire->new($self->{xdb}); } -# indexes all unindexed messages -sub index_sync { - my ($self, $git) = @_; - my $db = $self->{xdb}; - my $latest = $db->get_metadata('last_commit'); - my $range = length $latest ? "$latest..HEAD" : 'HEAD'; - $latest = undef; - - my $hex = '[a-f0-9]'; - my $h40 = $hex .'{40}'; - my $addmsg = qr!^:000000 100644 \S+ ($h40) A\t${hex}{2}/${hex}{38}$!; - my $delmsg = qr!^:100644 000000 ($h40) \S+ D\t${hex}{2}/${hex}{38}$!; - - # get indexed messages - my @cmd = ('git', "--git-dir=$git->{git_dir}", "log", - qw/--reverse --no-notes --no-color --raw -r --no-abbrev/, - $range); - - my $pid = open(my $log, '-|', @cmd) or - die('open` '.join(' ', @cmd) . " pipe failed: $!\n"); - my $last; - while (my $line = <$log>) { - if ($line =~ /$addmsg/o) { - $self->index_blob($git, $1); - } elsif ($line =~ /$delmsg/o) { - $self->unindex_blob($git, $1); - } elsif ($line =~ /^commit ($h40)/o) { - my $commit = $1; - if (defined $latest) { - $db->set_metadata('last_commit', $latest) - } - $latest = $commit; - } +sub help { + my ($self) = @_; + $self->qp; # parse altids + my @ret = @HELP; + if (my $user_pfx = $self->{-user_pfx}) { + push @ret, @$user_pfx; } - close $log; - $db->set_metadata('last_commit', $latest) if defined $latest; + \@ret; } 1;