X-Git-Url: http://www.git.stargrave.org/?a=blobdiff_plain;f=lib%2FPublicInbox%2FSearch.pm;h=ca389e320b2706edacbbd578b33dacc65812dece;hb=7503aeb540af5afd5cb1b554b3c29f35f5fc918d;hp=eac11bd44f905bd854af6e80cc4de104d26da93a;hpb=feabfb1809b911fc97538282234c8b1f087ddb6a;p=public-inbox.git

diff --git a/lib/PublicInbox/Search.pm b/lib/PublicInbox/Search.pm
index eac11bd4..ca389e32 100644
--- a/lib/PublicInbox/Search.pm
+++ b/lib/PublicInbox/Search.pm
@@ -8,16 +8,14 @@ use strict;
 use warnings;
 
 # 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 constant TS => 0;  # Received: header in Unix time
+use constant YYYYMMDD => 1; # for searching in the WWW UI
+use constant NUM => 2; # NNTP article number
 
 use Search::Xapian qw/:standard/;
 use PublicInbox::SearchMsg;
 use PublicInbox::MIME;
-use PublicInbox::MID qw/mid_clean id_compress/;
+use PublicInbox::MID qw/id_compress/;
 
 # This is English-only, everything else is non-standard and may be confused as
 # a prefix common in patch emails
@@ -56,9 +54,10 @@ my %bool_pfx_internal = (
 );
 
 my %bool_pfx_external = (
-	mid => 'XMID', # Message-ID (full/exact)
+	mid => 'Q', # Message-ID (full/exact), this is mostly uniQue
 );
 
+my $non_quoted_body = 'XNQ XDFN XDFA XDFB XDFHH XDFCTX XDFPRE XDFPOST';
 my %prob_prefix = (
 	# for mairix compatibility
 	s => 'S',
@@ -69,12 +68,12 @@ my %prob_prefix = (
 	c => 'XCC',
 	tcf => 'XTO XCC A',
 	a => 'XTO XCC A',
-	b => 'XNQ XQUOT',
-	bs => 'XNQ XQUOT S',
+	b => $non_quoted_body . ' XQUOT',
+	bs => $non_quoted_body . ' XQUOT S',
 	n => 'XFN',
 
 	q => 'XQUOT',
-	nq => 'XNQ',
+	nq => $non_quoted_body,
 	dfn => 'XDFN',
 	dfa => 'XDFA',
 	dfb => 'XDFB',
@@ -85,7 +84,7 @@ my %prob_prefix = (
 	dfblob => 'XDFPRE XDFPOST',
 
 	# default:
-	'' => 'XM S A XNQ XQUOT XFN',
+	'' => 'XM S A XQUOT XFN ' . $non_quoted_body,
 );
 
 # not documenting m: and mid: for now, the using the URLs works w/o Xapian
@@ -124,7 +123,10 @@ sub xdir {
 	if ($self->{version} == 1) {
 		"$self->{mainrepo}/public-inbox/xapian" . SCHEMA_VERSION;
 	} else {
-		"$self->{mainrepo}/xap" . SCHEMA_VERSION;
+		my $dir = "$self->{mainrepo}/xap" . SCHEMA_VERSION;
+		my $part = $self->{partition};
+		defined $part or die "partition not given";
+		$dir .= "/$part";
 	}
 }
 
@@ -141,11 +143,36 @@ sub new {
 		altid => $altid,
 		version => $version,
 	}, $class;
-	$self->{xdb} = Search::Xapian::Database->new($self->xdir);
+	if ($version >= 2) {
+		my $dir = "$self->{mainrepo}/xap" . SCHEMA_VERSION;
+		my $xdb;
+		my $parts = 0;
+		foreach my $part (<$dir/*>) {
+			-d $part && $part =~ m!/\d+\z! or next;
+			$parts++;
+			my $sub = Search::Xapian::Database->new($part);
+			if ($xdb) {
+				$xdb->add_database($sub);
+			} else {
+				$xdb = $sub;
+			}
+		}
+		$self->{xdb} = $xdb;
+		$self->{skel} = Search::Xapian::Database->new("$dir/skel");
+	} else {
+		$self->{xdb} = Search::Xapian::Database->new($self->xdir);
+	}
 	$self;
 }
 
-sub reopen { $_[0]->{xdb}->reopen }
+sub reopen {
+	my ($self) = @_;
+	$self->{xdb}->reopen;
+	if (my $skel = $self->{skel}) {
+		$skel->reopen;
+	}
+	$self; # make chaining easier
+}
 
 # read-only
 sub query {
@@ -163,9 +190,8 @@ sub query {
 
 sub get_thread {
 	my ($self, $mid, $opts) = @_;
-	my $smsg = eval { $self->lookup_message($mid) };
-
-	return { total => 0, msgs => [] } unless $smsg;
+	my $smsg = first_smsg_by_mid($self, $mid) or
+			return { total => 0, msgs => [] };
 	my $qtid = Search::Xapian::Query->new('G' . $smsg->thread_id);
 	my $path = $smsg->path;
 	if (defined $path && $path ne '') {
@@ -179,24 +205,27 @@ sub get_thread {
 	# always sort threads by timestamp, this makes life easier
 	# for the threading algorithm (in SearchThread.pm)
 	$opts->{asc} = 1;
-
+	$opts->{enquire} = enquire_skel($self);
 	_do_enquire($self, $qtid, $opts);
 }
 
 sub retry_reopen {
 	my ($self, $cb) = @_;
 	my $ret;
-	for (1..10) {
+	for my $i (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') {
+			warn "reopen try #$i on $@\n";
 			reopen($self);
 		} else {
+			warn "ref: ", ref($@), "\n";
 			die;
 		}
 	}
+	die "Too many Xapian database modifications in progress\n";
 }
 
 sub _do_enquire {
@@ -206,7 +235,7 @@ sub _do_enquire {
 
 sub _enquire_once {
 	my ($self, $query, $opts) = @_;
-	my $enquire = $self->enquire;
+	my $enquire = $opts->{enquire} || enquire($self);
 	if (defined $query) {
 		$query = Search::Xapian::Query->new(OP_AND,$query,$mail_query);
 	} else {
@@ -288,67 +317,87 @@ sub num_range_processor {
 sub query_xover {
 	my ($self, $beg, $end, $offset) = @_;
 	my $qp = Search::Xapian::QueryParser->new;
-	$qp->set_database($self->{xdb});
+	$qp->set_database($self->{skel} || $self->{xdb});
 	$qp->add_valuerangeprocessor($self->num_range_processor);
 	my $query = $qp->parse_query("$beg..$end", QP_FLAGS);
 
-	_do_enquire($self, $query, {num => 1, limit => 200, offset => $offset});
+	my $opts = {
+		enquire => enquire_skel($self),
+		num => 1,
+		limit => 200,
+		offset => $offset,
+	};
+	_do_enquire($self, $query, $opts);
 }
 
-sub lookup_message {
-	my ($self, $mid) = @_;
-	$mid = mid_clean($mid);
+sub query_ts {
+	my ($self, $ts, $opts) = @_;
+	my $qp = $self->{qp_ts} ||= eval {
+		my $q = Search::Xapian::QueryParser->new;
+		$q->set_database($self->{skel} || $self->{xdb});
+		$q->add_valuerangeprocessor(
+			Search::Xapian::NumberValueRangeProcessor->new(TS));
+		$q
+	};
+	my $query = $qp->parse_query($ts, QP_FLAGS);
+	$opts->{enquire} = enquire_skel($self);
+	_do_enquire($self, $query, $opts);
+}
 
-	my $doc_id = $self->find_first_doc_id('XMID' . $mid);
+sub first_smsg_by_mid {
+	my ($self, $mid) = @_;
 	my $smsg;
-	if (defined $doc_id) {
-		# raises on error:
-		my $doc = $self->{xdb}->get_document($doc_id);
-		$smsg = PublicInbox::SearchMsg->wrap($doc, $mid);
-		$smsg->{doc_id} = $doc_id;
-	}
+	retry_reopen($self, sub {
+		each_smsg_by_mid($self, $mid, sub { $smsg = $_[0]; undef });
+	});
 	$smsg;
 }
 
-sub lookup_mail { # no ghosts!
-	my ($self, $mid) = @_;
+sub lookup_article {
+	my ($self, $num) = @_;
+	my $term = 'XNUM'.$num;
+	my $db = $self->{skel} || $self->{xdb};
 	retry_reopen($self, sub {
-		my $smsg = lookup_message($self, $mid) or return;
+		my $head = $db->postlist_begin($term);
+		my $tail = $db->postlist_end($term);
+		return if $head->equal($tail);
+		my $doc_id = $head->get_docid;
+		return unless defined $doc_id;
+		$head->inc;
+		if ($head->nequal($tail)) {
+			my $loc= $self->{mainrepo} .
+				($self->{skel} ? 'skel' : 'xdb');
+			warn "article #$num is not unique in $loc\n";
+		}
+		# raises on error:
+		my $doc = $db->get_document($doc_id);
+		my $smsg = PublicInbox::SearchMsg->wrap($doc);
+		$smsg->{doc_id} = $doc_id;
 		$smsg->load_expand;
 	});
 }
 
-sub find_unique_doc_id {
-	my ($self, $termval) = @_;
-
-	my ($begin, $end) = $self->find_doc_ids($termval);
-
-	return undef if $begin->equal($end); # not found
-
-	my $rv = $begin->get_docid;
-
-	# sanity check
-	$begin->inc;
-	$begin->equal($end) or die "Term '$termval' is not unique\n";
-	$rv;
-}
-
-# returns begin and end PostingIterator
-sub find_doc_ids {
-	my ($self, $termval) = @_;
-	my $db = $self->{xdb};
-
-	($db->postlist_begin($termval), $db->postlist_end($termval));
-}
-
-sub find_first_doc_id {
-	my ($self, $termval) = @_;
-
-	my ($begin, $end) = $self->find_doc_ids($termval);
-
-	return undef if $begin->equal($end); # not found
-
-	$begin->get_docid;
+sub each_smsg_by_mid {
+	my ($self, $mid, $cb) = @_;
+	# XXX retry_reopen isn't necessary for V2Writable, but the PSGI
+	# interface will need it...
+	my $db = $self->{skel} || $self->{xdb};
+	my $term = 'Q' . $mid;
+	my $head = $db->postlist_begin($term);
+	my $tail = $db->postlist_end($term);
+	if ($head == $tail) {
+		$db->reopen;
+		$head = $db->postlist_begin($term);
+		$tail = $db->postlist_end($term);
+	}
+	return ($head, $tail, $db) if wantarray;
+	for (; $head->nequal($tail); $head->inc) {
+		my $doc_id = $head->get_docid;
+		my $doc = $db->get_document($doc_id);
+		my $smsg = PublicInbox::SearchMsg->wrap($doc, $mid);
+		$smsg->{doc_id} = $doc_id;
+		$cb->($smsg) or return;
+	}
 }
 
 # normalize subjects so they are suitable as pathnames for URLs
@@ -375,6 +424,15 @@ sub enquire {
 	$self->{enquire} ||= Search::Xapian::Enquire->new($self->{xdb});
 }
 
+sub enquire_skel {
+	my ($self) = @_;
+	if (my $skel = $self->{skel}) {
+		$self->{enquire_skel} ||= Search::Xapian::Enquire->new($skel);
+	} else {
+		enquire($self);
+	}
+}
+
 sub help {
 	my ($self) = @_;
 	$self->qp; # parse altids