diff options
Diffstat (limited to 'lib/PublicInbox')
-rw-r--r-- | lib/PublicInbox/ExtSearchIdx.pm | 90 | ||||
-rw-r--r-- | lib/PublicInbox/MiscIdx.pm | 16 | ||||
-rw-r--r-- | lib/PublicInbox/SearchIdx.pm | 23 | ||||
-rw-r--r-- | lib/PublicInbox/SearchIdxShard.pm | 21 |
4 files changed, 145 insertions, 5 deletions
diff --git a/lib/PublicInbox/ExtSearchIdx.pm b/lib/PublicInbox/ExtSearchIdx.pm index d780776f..4de47b58 100644 --- a/lib/PublicInbox/ExtSearchIdx.pm +++ b/lib/PublicInbox/ExtSearchIdx.pm @@ -87,6 +87,7 @@ sub _ibx_attach { # each_inbox callback sub attach_config { my ($self, $cfg) = @_; + $self->{cfg} = $cfg; $cfg->each_inbox(\&_ibx_attach, $self); } @@ -141,7 +142,8 @@ sub do_xpost ($$) { if ($nr == 0) { $idx->shard_remove($oid, $docid); } elsif ($rm_eidx_info) { - $idx->shard_remove_eidx_info($docid, $oid, $xibx, $eml); + $idx->shard_remove_eidx_info($docid, $oid, $eidx_key, + $eml); } } } @@ -324,6 +326,90 @@ sub _sync_inbox ($$$) { $ibx->git->cleanup; # done with this inbox, now } +sub unref_doc ($$$$) { + my ($self, $ibx_id, $eidx_key, $docid) = @_; + my $dbh = $self->{oidx}->dbh; + + # for debug/info purposes, oids may no longer be accessible + my $sth = $dbh->prepare_cached(<<'', undef, 1); +SELECT oidbin FROM xref3 WHERE docid = ? AND ibx_id = ? + + $sth->execute($docid, $ibx_id); + my @oid = map { unpack('H*', $_->[0]) } @{$sth->fetchall_arrayref}; + + $dbh->prepare_cached(<<'')->execute($docid, $ibx_id); +DELETE FROM xref3 WHERE docid = ? AND ibx_id = ? + + my $remain = $self->{oidx}->get_xref3($docid); + my $idx = $self->idx_shard($docid); + if (@$remain) { + for my $oid (@oid) { + warn "I: unref #$docid $eidx_key $oid\n"; + $idx->shard_remove_eidx_info($docid, $oid, $eidx_key); + } + } else { + for my $oid (@oid) { + warn "I: remove #$docid $eidx_key $oid\n"; + $idx->shard_remove($oid, $docid); + } + } +} + +sub eidx_gc { + my ($self, $opt) = @_; + $self->{cfg} or die "E: GC requires ->attach_config\n"; + $opt->{-idx_gc} = 1; + $self->idx_init($opt); # acquire lock via V2Writable::_idx_init + + my $dbh = $self->{oidx}->dbh; + my $x3_doc = $dbh->prepare('SELECT docid FROM xref3 WHERE ibx_id = ?'); + my $ibx_ck = $dbh->prepare('SELECT ibx_id,eidx_key FROM inboxes'); + my $lc_i = $dbh->prepare('SELECT key FROM eidx_meta WHERE key LIKE ?'); + + $ibx_ck->execute; + while (my ($ibx_id, $eidx_key) = $ibx_ck->fetchrow_array) { + next if $self->{ibx_map}->{$eidx_key}; + $self->{midx}->remove_eidx_key($eidx_key); + warn "I: deleting messages for $eidx_key...\n"; + $x3_doc->execute($ibx_id); + while (defined(my $docid = $x3_doc->fetchrow_array)) { + unref_doc($self, $ibx_id, $eidx_key, $docid); + } + $dbh->prepare_cached(<<'')->execute($ibx_id); +DELETE FROM inboxes WHERE ibx_id = ? + + # drop last_commit info + my $pat = $eidx_key; + $pat =~ s/([_%])/\\$1/g; + $lc_i->execute("lc-%:$pat//%"); + while (my ($key) = $lc_i->fetchrow_array) { + next if $key !~ m!\Alc-v[1-9]+:\Q$eidx_key\E//!; + warn "I: removing $key\n"; + $dbh->prepare_cached(<<'')->execute($key); +DELETE FROM eidx_meta WHERE key = ? + + } + + warn "I: $eidx_key removed\n"; + } + + # it's not real unless it's in `over', we use parallelism here, + # shards will be reading directly from over, so commit + $self->{oidx}->commit_lazy; + $self->{oidx}->begin_lazy; + + for my $idx (@{$self->{idx_shards}}) { + warn "I: cleaning up shard #$idx->{shard}\n"; + $idx->shard_over_check($self->{oidx}); + } + my $nr = $dbh->do(<<''); +DELETE FROM xref3 WHERE docid NOT IN (SELECT num FROM over) + + warn "I: eliminated $nr stale xref3 entries\n" if $nr != 0; + + done($self); +} + sub eidx_sync { # main entry point my ($self, $opt) = @_; $self->idx_init($opt); # acquire lock via V2Writable::_idx_init @@ -413,6 +499,7 @@ sub idx_init { # similar to V2Writable next if $seen{"$st[0]\0$st[1]"}++; } else { warn "W: stat($d) failed (from $alt): $!\n"; + next if $opt->{-idx_gc}; } push @old, $line; } @@ -424,6 +511,7 @@ sub idx_init { # similar to V2Writable next if $seen{"$st[0]\0$st[1]"}++; } else { warn "W: stat($d) failed (from $ibx->{inboxdir}): $!\n"; + next if $opt->{-idx_gc}; } push @new, $line; } diff --git a/lib/PublicInbox/MiscIdx.pm b/lib/PublicInbox/MiscIdx.pm index 642d920b..64591d05 100644 --- a/lib/PublicInbox/MiscIdx.pm +++ b/lib/PublicInbox/MiscIdx.pm @@ -53,6 +53,22 @@ sub commit_txn { delete($self->{xdb})->commit_transaction; } +sub remove_eidx_key { + my ($self, $eidx_key) = @_; + my $xdb = $self->{xdb}; + my $head = $xdb->postlist_begin('Q'.$eidx_key); + my $tail = $xdb->postlist_end('Q'.$eidx_key); + my @docids; # only one, unless we had bugs + for (; $head != $tail; $head++) { + push @docids, $head->get_docid; + } + for my $docid (@docids) { + $xdb->delete_document($docid); + warn "I: remove inbox docid #$docid ($eidx_key)\n"; + } +} + +# adds or updates according to $eidx_key sub index_ibx { my ($self, $ibx) = @_; my $eidx_key = $ibx->eidx_key; diff --git a/lib/PublicInbox/SearchIdx.pm b/lib/PublicInbox/SearchIdx.pm index d06c159b..c18c7c36 100644 --- a/lib/PublicInbox/SearchIdx.pm +++ b/lib/PublicInbox/SearchIdx.pm @@ -471,7 +471,7 @@ sub remove_eidx_info { my $doc = _get_doc($self, $docid, $oid) or return; eval { $doc->remove_term('O'.$eidx_key) }; warn "W: ->remove_term O$eidx_key: $@\n" if $@; - for my $l ($eml->header_raw('List-Id')) { + for my $l ($eml ? $eml->header_raw('List-Id') : ()) { $l =~ /<([^>]+)>/ or next; my $lid = lc $1; eval { $doc->remove_term('G' . $lid) }; @@ -970,4 +970,25 @@ sub eidx_shard_new { $self; } +# ensure there's no stale Xapian docs by treating $over as canonical +sub over_check { + my ($self, $over) = @_; + begin_txn_lazy($self); + my $sth = $over->dbh->prepare(<<''); +SELECT COUNT(*) FROM over WHERE num = ? + + my $xdb = $self->{xdb}; + my $cur = $xdb->postlist_begin(''); + my $end = $xdb->postlist_end(''); + my $xdir = $self->xdir; + for (; $cur != $end; $cur++) { + my $docid = $cur->get_docid; + $sth->execute($docid); + my $x = $sth->fetchrow_array; + next if $x > 0; + warn "I: removing $xdir #$docid, not in `over'\n"; + $xdb->delete_document($docid); + } +} + 1; diff --git a/lib/PublicInbox/SearchIdxShard.pm b/lib/PublicInbox/SearchIdxShard.pm index dcfeb0be..53fac9b6 100644 --- a/lib/PublicInbox/SearchIdxShard.pm +++ b/lib/PublicInbox/SearchIdxShard.pm @@ -57,6 +57,7 @@ sub spawn_worker { sub eml ($$) { my ($r, $len) = @_; + return if $len == 0; my $n = read($r, my $bref, $len) or die "read: $!\n"; $n == $len or die "short read: $n != $len\n"; PublicInbox::Eml->new(\$bref); @@ -92,6 +93,10 @@ sub shard_worker_loop ($$$$$) { chomp $eidx_key; $self->remove_eidx_info($docid, $oid, $eidx_key, eml($r, $len)); + } elsif ($line =~ s/\AO ([^\n]+)\n//) { + my $over_fn = $1; + $over_fn =~ tr/\0/\n/; + $self->over_check(PublicInbox::Over->new($over_fn)); } else { chomp $line; my $eidx_key; @@ -155,10 +160,9 @@ sub shard_add_eidx_info { } sub shard_remove_eidx_info { - my ($self, $docid, $oid, $xibx, $eml) = @_; - my $eidx_key = $xibx->eidx_key; + my ($self, $docid, $oid, $eidx_key, $eml) = @_; if (my $w = $self->{w}) { - my $hdr = $eml->header_obj->as_string; + my $hdr = $eml ? $eml->header_obj->as_string : ''; my $len = length($hdr); print $w "-X $len $docid $oid $eidx_key\n", $hdr or die "failed to write shard: $!"; @@ -212,4 +216,15 @@ sub shard_remove { } } +sub shard_over_check { + my ($self, $over) = @_; + if (my $w = $self->{w}) { # triggers remove_by_oid in a shard child + my ($over_fn) = $over->{dbh}->sqlite_db_filename; + $over_fn =~ tr/\n/\0/; + print $w "O $over_fn\n" or die "failed to write over $!"; + } else { + $self->over_check($over); + } +} + 1; |