X-Git-Url: http://www.git.stargrave.org/?a=blobdiff_plain;f=lib%2FPublicInbox%2FSearchIdx.pm;h=1b1a21b767b1443367045a845a6c8a3acb37ef22;hb=9bd675d33ad1e49bd2ebe12a1d216216e61380de;hp=db0495bcd8a90d313c1ea08cafab564d16baf541;hpb=65323f060a3db731bb9fafa004336eeb4bbb8f00;p=public-inbox.git diff --git a/lib/PublicInbox/SearchIdx.pm b/lib/PublicInbox/SearchIdx.pm index db0495bc..1b1a21b7 100644 --- a/lib/PublicInbox/SearchIdx.pm +++ b/lib/PublicInbox/SearchIdx.pm @@ -1,4 +1,4 @@ -# Copyright (C) 2015-2018 all contributors +# Copyright (C) 2015-2019 all contributors # License: AGPL-3.0+ # based on notmuch, but with no concept of folders, files or flags # @@ -19,7 +19,6 @@ use POSIX qw(strftime); use PublicInbox::OverIdx; use PublicInbox::Spawn qw(spawn); use PublicInbox::Git qw(git_unquote); -use Compress::Zlib qw(compress); use constant { BATCH_BYTES => defined($ENV{XAPIAN_FLUSH_THRESHOLD}) ? @@ -30,35 +29,25 @@ use constant { my $xapianlevels = qr/\A(?:full|medium)\z/; sub new { - my ($class, $ibx, $creat, $part) = @_; + my ($class, $ibx, $creat, $shard) = @_; + ref $ibx or die "BUG: expected PublicInbox::Inbox object: $ibx"; my $levels = qr/\A(?:full|medium|basic)\z/; - my $mainrepo = $ibx; # for "public-inbox-index" w/o entry in config - my $git_dir = $mainrepo; - my ($altid, $git); - my $version = 1; + my $mainrepo = $ibx->{mainrepo}; + my $version = $ibx->{version} || 1; my $indexlevel = 'full'; - if (ref $ibx) { - $mainrepo = $ibx->{mainrepo}; - $altid = $ibx->{altid}; - $version = $ibx->{version} || 1; - if ($altid) { - require PublicInbox::AltId; - $altid = [ map { - PublicInbox::AltId->new($ibx, $_); - } @$altid ]; - } - if ($ibx->{indexlevel}) { - if ($ibx->{indexlevel} =~ $levels) { - $indexlevel = $ibx->{indexlevel}; - } else { - die("Invalid indexlevel $ibx->{indexlevel}\n"); - } + my $altid = $ibx->{altid}; + if ($altid) { + require PublicInbox::AltId; + $altid = [ map { PublicInbox::AltId->new($ibx, $_); } @$altid ]; + } + if ($ibx->{indexlevel}) { + if ($ibx->{indexlevel} =~ $levels) { + $indexlevel = $ibx->{indexlevel}; + } else { + die("Invalid indexlevel $ibx->{indexlevel}\n"); } - } else { # v1 - $ibx = { mainrepo => $git_dir, version => 1 }; } $ibx = PublicInbox::InboxWritable->new($ibx); - require Search::Xapian::WritableDatabase; my $self = bless { mainrepo => $mainrepo, -inbox => $ibx, @@ -73,9 +62,9 @@ sub new { my $dir = $self->xdir; $self->{over} = PublicInbox::OverIdx->new("$dir/over.sqlite3"); } elsif ($version == 2) { - defined $part or die "partition is required for v2\n"; - # partition is a number - $self->{partition} = $part; + defined $shard or die "shard is required for v2\n"; + # shard is a number + $self->{shard} = $shard; $self->{lock_path} = undef; } else { die "unsupported inbox version=$version\n"; @@ -84,26 +73,46 @@ sub new { $self; } +sub need_xapian ($) { $_[0]->{indexlevel} =~ $xapianlevels } + sub _xdb_release { my ($self) = @_; - my $xdb = delete $self->{xdb} or croak 'not acquired'; - $xdb->close; + if (need_xapian($self)) { + my $xdb = delete $self->{xdb} or croak 'not acquired'; + $xdb->close; + } $self->lock_release if $self->{creat}; undef; } sub _xdb_acquire { my ($self) = @_; - croak 'already acquired' if $self->{xdb}; + my $flag; my $dir = $self->xdir; - my $flag = Search::Xapian::DB_OPEN; + if (need_xapian($self)) { + croak 'already acquired' if $self->{xdb}; + PublicInbox::Search::load_xapian(); + require Search::Xapian::WritableDatabase; + $flag = $self->{creat} ? + Search::Xapian::DB_CREATE_OR_OPEN() : + Search::Xapian::DB_OPEN(); + } if ($self->{creat}) { require File::Path; $self->lock_acquire; - File::Path::mkpath($dir); - $flag = Search::Xapian::DB_CREATE_OR_OPEN; + + # don't create empty Xapian directories if we don't need Xapian + my $is_shard = defined($self->{shard}); + if (!$is_shard || ($is_shard && need_xapian($self))) { + File::Path::mkpath($dir); + } + } + return unless defined $flag; + my $xdb = eval { Search::Xapian::WritableDatabase->new($dir, $flag) }; + if ($@) { + die "Failed opening $dir: ", $@; } - $self->{xdb} = Search::Xapian::WritableDatabase->new($dir, $flag); + $self->{xdb} = $xdb; } sub add_val ($$$) { @@ -342,7 +351,7 @@ sub add_message { $num = index_mm($self, $mime); } eval { - if ($self->{indexlevel} =~ $xapianlevels) { + if (need_xapian($self)) { $self->add_xapian($mime, $num, $oid, $mids, $mid0) } if (my $over = $self->{over}) { @@ -365,6 +374,7 @@ sub find_doc_ids { ($db->postlist_begin($termval), $db->postlist_end($termval)); } +# v1 only sub batch_do { my ($self, $termval, $cb) = @_; my $batch_size = 1000; # don't let @ids grow too large to avoid OOM @@ -379,35 +389,45 @@ sub batch_do { } } +# v1 only, where $mid is unique sub remove_message { my ($self, $mid) = @_; - my $db = $self->{xdb}; - my $called; $mid = mid_clean($mid); - my $over = $self->{over}; + if (my $over = $self->{over}) { + my $nr = eval { $over->remove_oid(undef, $mid) }; + if ($@) { + warn "failed to remove <$mid> from overview: $@\n"; + } elsif ($nr == 0) { + warn "<$mid> missing for removal from overview\n"; + } + } + return unless need_xapian($self); + my $db = $self->{xdb}; + my $nr = 0; eval { batch_do($self, 'Q' . $mid, sub { my ($ids) = @_; $db->delete_document($_) for @$ids; - $over->delete_articles($ids) if $over; - $called = 1; + $nr = scalar @$ids; }); }; if ($@) { - warn "failed to remove message <$mid>: $@\n"; - } elsif (!$called) { - warn "cannot remove non-existent <$mid>\n"; + warn "failed to remove <$mid> from Xapian: $@\n"; + } elsif ($nr == 0) { + warn "<$mid> missing for removal from Xapian\n"; } } # MID is a hint in V2 sub remove_by_oid { my ($self, $oid, $mid) = @_; - my $db = $self->{xdb}; $self->{over}->remove_oid($oid, $mid) if $self->{over}; + return unless need_xapian($self); + my $db = $self->{xdb}; + # XXX careful, we cannot use batch_do here since we conditionally # delete documents based on other factors, so we cannot call # find_doc_ids twice. @@ -517,17 +537,19 @@ sub do_cat_mail { $@ ? undef : $mime; } +# called by public-inbox-index sub index_sync { my ($self, $opts) = @_; + delete $self->{lock_path} if $opts->{-skip_lock}; $self->{-inbox}->with_umask(sub { $self->_index_sync($opts) }) } -sub batch_adjust ($$$$) { - my ($max, $bytes, $batch_cb, $latest) = @_; +sub batch_adjust ($$$$$) { + my ($max, $bytes, $batch_cb, $latest, $nr) = @_; $$max -= $bytes; if ($$max <= 0) { $$max = BATCH_BYTES; - $batch_cb->($latest); + $batch_cb->($nr, $latest); } } @@ -546,6 +568,7 @@ sub read_log { my %D; my $line; my $newest; + my $nr = 0; while (defined($line = <$log>)) { if ($line =~ /$addmsg/o) { my $blob = $1; @@ -557,7 +580,7 @@ sub read_log { next; } my $mime = do_cat_mail($git, $blob, \$bytes) or next; - batch_adjust(\$max, $bytes, $batch_cb, $latest); + batch_adjust(\$max, $bytes, $batch_cb, $latest, ++$nr); $add_cb->($self, $mime, $bytes, $blob); } elsif ($line =~ /$delmsg/o) { my $blob = $1; @@ -572,7 +595,7 @@ sub read_log { my $mime = do_cat_mail($git, $blob, \$bytes) or next; $del_cb->($self, $mime); } - $batch_cb->($latest, $newest); + $batch_cb->($nr, $latest, $newest); } sub _msgmap_init { @@ -585,7 +608,7 @@ sub _msgmap_init { } sub _git_log { - my ($self, $range) = @_; + my ($self, $opts, $range) = @_; my $git = $self->{git}; if (index($range, '..') < 0) { @@ -602,12 +625,17 @@ sub _git_log { # Count the new files so they can be added newest to oldest # and still have numbers increasing from oldest to newest my $fcount = 0; + my $pr = $opts->{-progress}; + $pr->("counting changes\n\t$range ... ") if $pr; # can't use 'rev-list --count' if we use --diff-filter my $fh = $git->popen(qw(log --pretty=tformat:%h --no-notes --no-color --no-renames --diff-filter=AM), $range); ++$fcount while <$fh>; + close $fh; my $high = $self->{mm}->num_highwater; + $pr->("$fcount\n") if $pr; # continue previous line + $self->{ntodo} = $fcount; if (index($range, '..') < 0) { if ($high && $high == $fcount) { @@ -648,50 +676,64 @@ sub need_update ($$$) { ($n eq '' || $n > 0); } +# The last git commit we indexed with Xapian or SQLite (msgmap) +# This needs to account for cases where Xapian or SQLite is +# out-of-date with respect to the other. +sub _last_x_commit { + my ($self, $mm) = @_; + my $lm = $mm->last_commit || ''; + my $lx = ''; + if (need_xapian($self)) { + $lx = $self->{xdb}->get_metadata('last_commit') || ''; + } else { + $lx = $lm; + } + # Use last_commit from msgmap if it is older or unset + if (!$lm || ($lx && $lm && is_ancestor($self->{git}, $lm, $lx))) { + $lx = $lm; + } + $lx; +} + +sub reindex_from ($$) { + my ($reindex, $last_commit) = @_; + return $last_commit unless $reindex; + ref($reindex) eq 'HASH' ? $reindex->{from} : ''; +} + # indexes all unindexed messages (v1 only) sub _index_sync { my ($self, $opts) = @_; my $tip = $opts->{ref} || 'HEAD'; - my $reindex = $opts->{reindex}; - my ($mkey, $last_commit, $lx, $xlog); + my ($last_commit, $lx, $xlog); my $git = $self->{git}; $git->batch_prepare; + my $pr = $opts->{-progress}; my $xdb = $self->begin_txn_lazy; my $mm = _msgmap_init($self); do { $xlog = undef; - $mkey = 'last_commit'; - $last_commit = $xdb->get_metadata('last_commit'); - $lx = $last_commit; - if ($reindex) { - $lx = ''; - $mkey = undef if $last_commit ne ''; - } - - # use last_commit from msgmap if it is older or unset - my $lm = $mm->last_commit || ''; - if (!$lm || ($lm && $lx && is_ancestor($git, $lm, $lx))) { - $lx = $lm; - } + $last_commit = _last_x_commit($self, $mm); + $lx = reindex_from($opts->{reindex}, $last_commit); $self->{over}->rollback_lazy; $self->{over}->disconnect; $git->cleanup; delete $self->{txn}; - $xdb->cancel_transaction; + $xdb->cancel_transaction if $xdb; $xdb = _xdb_release($self); # ensure we leak no FDs to "git log" with Xapian <= 1.2 my $range = $lx eq '' ? $tip : "$lx..$tip"; - $xlog = _git_log($self, $range); + $xlog = _git_log($self, $opts, $range); $xdb = $self->begin_txn_lazy; - } while ($xdb->get_metadata('last_commit') ne $last_commit); + } while (_last_x_commit($self, $mm) ne $last_commit); my $dbh = $mm->{dbh} if $mm; my $cb = sub { - my ($commit, $newest) = @_; + my ($nr, $commit, $newest) = @_; if ($dbh) { if ($newest) { my $cur = $mm->last_commit || ''; @@ -701,16 +743,17 @@ sub _index_sync { } $dbh->commit; } - if ($mkey && $newest && $self->{indexlevel} =~ $xapianlevels) { - my $cur = $xdb->get_metadata($mkey); + if ($newest && need_xapian($self)) { + my $cur = $xdb->get_metadata('last_commit'); if (need_update($self, $cur, $newest)) { - $xdb->set_metadata($mkey, $newest); + $xdb->set_metadata('last_commit', $newest); } } $self->commit_txn_lazy; $git->cleanup; $xdb = _xdb_release($self); # let another process do some work... < + $pr->("indexed $nr/$self->{ntodo}\n") if $pr && $nr; if (!$newest) { $xdb = $self->begin_txn_lazy; $dbh->begin_work if $dbh; @@ -754,7 +797,7 @@ sub remote_close { sub remote_remove { my ($self, $oid, $mid) = @_; if (my $w = $self->{w}) { - # triggers remove_by_oid in a partition + # triggers remove_by_oid in a shard print $w "D $oid $mid\n" or die "failed to write remove $!"; } else { $self->begin_txn_lazy; @@ -769,7 +812,7 @@ sub begin_txn_lazy { $self->{-inbox}->with_umask(sub { my $xdb = $self->{xdb} || $self->_xdb_acquire; $self->{over}->begin_lazy if $self->{over}; - $xdb->begin_transaction; + $xdb->begin_transaction if $xdb; $self->{txn} = 1; $xdb; }); @@ -779,14 +822,27 @@ sub commit_txn_lazy { my ($self) = @_; delete $self->{txn} or return; $self->{-inbox}->with_umask(sub { - $self->{xdb}->commit_transaction; + if (my $xdb = $self->{xdb}) { + + # store 'indexlevel=medium' in v2 shard=0 and + # v1 (only one shard) + # This metadata is read by Admin::detect_indexlevel: + if (!$self->{shard} # undef or 0, not >0 + && $self->{indexlevel} eq 'medium') { + $xdb->set_metadata('indexlevel', 'medium'); + } + + $xdb->commit_transaction; + } $self->{over}->commit_lazy if $self->{over}; }); } sub worker_done { my ($self) = @_; - die "$$ $0 xdb not released\n" if $self->{xdb}; + if (need_xapian($self)) { + die "$$ $0 xdb not released\n" if $self->{xdb}; + } die "$$ $0 still in transaction\n" if $self->{txn}; }