X-Git-Url: http://www.git.stargrave.org/?a=blobdiff_plain;f=lib%2FPublicInbox%2FOverIdx.pm;h=6cc86d5d038d3669b0991d270aa74234f3937560;hb=e28f33dc1b6e6a296946c0c58b43900c6cbd07f7;hp=985abbf4e693de8f01939556abca07504fa21eae;hpb=e46b24f437639e3c2cc19df4a2c875b56fe6fea7;p=public-inbox.git diff --git a/lib/PublicInbox/OverIdx.pm b/lib/PublicInbox/OverIdx.pm index 985abbf4..6cc86d5d 100644 --- a/lib/PublicInbox/OverIdx.pm +++ b/lib/PublicInbox/OverIdx.pm @@ -1,4 +1,4 @@ -# Copyright (C) 2018-2021 all contributors +# Copyright (C) all contributors # License: AGPL-3.0+ # for XOVER, OVER in NNTP, and feeds/homepage/threads in PSGI @@ -434,6 +434,7 @@ sub commit_lazy { my ($self) = @_; delete $self->{txn} or return; $self->{dbh}->commit; + eval { $self->{dbh}->do('PRAGMA optimize') }; } sub begin_lazy { @@ -508,12 +509,12 @@ EOF next; } $pr->(<{num} <$mid> THREADID=$r->{tid} culled +# ghost $r->{num} <$mid> THREADID=$r->{tid} culled EOM } delete_by_num($self, $r->{num}); } - $pr->("I: rethread culled $total ghosts\n") if $pr && $total; + $pr->("# rethread culled $total ghosts\n") if $pr && $total; } # used for cross-inbox search @@ -543,9 +544,13 @@ CREATE TABLE IF NOT EXISTS xref3 ( $dbh->do('CREATE INDEX IF NOT EXISTS idx_docid ON xref3 (docid)'); # performance critical, this is not UNIQUE since we may need to - # tolerate some old bugs from indexing mirrors - $dbh->do('CREATE INDEX IF NOT EXISTS idx_nntp ON '. - 'xref3 (oidbin,xnum,ibx_id)'); + # tolerate some old bugs from indexing mirrors. n.b. we used + # to index oidbin here, but leaving it out speeds up reindexing + # and "XHDR Xref <$MSGID>" isn't any slower w/o oidbin + $dbh->do('CREATE INDEX IF NOT EXISTS idx_reindex ON '. + 'xref3 (xnum,ibx_id)'); + + $dbh->do('CREATE INDEX IF NOT EXISTS idx_oidbin ON xref3 (oidbin)'); $dbh->do(<<''); CREATE TABLE IF NOT EXISTS eidx_meta ( @@ -602,50 +607,6 @@ INSERT OR IGNORE INTO xref3 (docid, ibx_id, xnum, oidbin) VALUES (?, ?, ?, ?) $sth->execute; } -# returns remaining reference count to $docid -sub remove_xref3 { - my ($self, $docid, $oidhex, $eidx_key, $rm_eidx_info) = @_; - begin_lazy($self); - my $oidbin = pack('H*', $oidhex); - my ($sth, $ibx_id); - if (defined $eidx_key) { - $ibx_id = ibx_id($self, $eidx_key); - $sth = $self->{dbh}->prepare_cached(<<''); -DELETE FROM xref3 WHERE docid = ? AND ibx_id = ? AND oidbin = ? - - $sth->bind_param(1, $docid); - $sth->bind_param(2, $ibx_id); - $sth->bind_param(3, $oidbin, SQL_BLOB); - } else { - $sth = $self->{dbh}->prepare_cached(<<''); -DELETE FROM xref3 WHERE docid = ? AND oidbin = ? - - $sth->bind_param(1, $docid); - $sth->bind_param(2, $oidbin, SQL_BLOB); - } - $sth->execute; - $sth = $self->{dbh}->prepare_cached(<<'', undef, 1); -SELECT COUNT(*) FROM xref3 WHERE docid = ? - - $sth->execute($docid); - my $nr = $sth->fetchrow_array; - if ($nr == 0) { - delete_by_num($self, $docid); - } elsif (defined($ibx_id) && $rm_eidx_info) { - # if deduplication rules in ContentHash change, it's - # possible a docid can have multiple rows with the - # same ibx_id. This governs whether or not we call - # ->remove_eidx_info in ExtSearchIdx. - $sth = $self->{dbh}->prepare_cached(<<'', undef, 1); -SELECT COUNT(*) FROM xref3 WHERE docid = ? AND ibx_id = ? - - $sth->execute($docid, $ibx_id); - my $count = $sth->fetchrow_array; - $$rm_eidx_info = ($count == 0); - } - $nr; -} - # for when an xref3 goes missing, this does NOT update {ts} sub update_blob { my ($self, $smsg, $oidhex) = @_; @@ -659,8 +620,7 @@ UPDATE over SET ddd = ? WHERE num = ? } sub merge_xref3 { # used for "-extindex --dedupe" - my ($self, $keep_docid, $drop_docid, $oidhex) = @_; - my $oidbin = pack('H*', $oidhex); + my ($self, $keep_docid, $drop_docid, $oidbin) = @_; my $sth = $self->{dbh}->prepare_cached(<<''); UPDATE OR IGNORE xref3 SET docid = ? WHERE docid = ? AND oidbin = ? @@ -710,4 +670,16 @@ sub vivify_xvmd { $smsg->{-vivify_xvmd} = \@vivify_xvmd; } +sub fork_ok { + return 1 if $DBD::SQLite::sqlite_version >= 3008003; + my ($opt) = @_; + my @j = split(/,/, $opt->{jobs} // ''); + state $warned; + grep { $_ > 1 } @j and $warned //= warn('DBD::SQLite version is ', + $DBD::SQLite::sqlite_version, + ", need >= 3008003 (3.8.3) for --jobs > 1\n"); + $opt->{jobs} = '1,1'; + undef; +} + 1;