X-Git-Url: http://www.git.stargrave.org/?a=blobdiff_plain;f=t%2Fsearch.t;h=f026e5093c9176d00f7bd39487d919caff6baa27;hb=d73d783ab2cf14ba28ca63723223d8c85a68cdd5;hp=cf3254169ca3485e689e5498f97df82b2b0dc69f;hpb=601658e48f0e28ca178ef5c4d7eed0e50ef42167;p=public-inbox.git

diff --git a/t/search.t b/t/search.t
index cf325416..f026e509 100644
--- a/t/search.t
+++ b/t/search.t
@@ -21,8 +21,8 @@ ok($@, "exception raised on non-existent DB");
 
 my $rw = PublicInbox::SearchIdx->new($ibx, 1);
 $ibx->with_umask(sub {
-	$rw->_xdb_acquire;
-	$rw->_xdb_release;
+	$rw->idx_acquire;
+	$rw->idx_release;
 });
 $rw = undef;
 my $ro = PublicInbox::Search->new($ibx);
@@ -59,6 +59,14 @@ sub oct_is ($$$) {
 	}
 }
 
+{
+	my $crlf_adjust = \&PublicInbox::SearchIdx::crlf_adjust;
+	is($crlf_adjust->("hi\r\nworld\r\n"), 0, 'no adjustment needed');
+	is($crlf_adjust->("hi\nworld\n"), 2, 'LF-only counts two CR');
+	is($crlf_adjust->("hi\r\nworld\n"), 1, 'CRLF/LF-mix 1 counts 1 CR');
+	is($crlf_adjust->("hi\nworld\r\n"), 1, 'CRLF/LF-mix 2 counts 1 CR');
+}
+
 $ibx->with_umask(sub {
 	my $root = PublicInbox::Eml->new(<<'EOF');
 Date: Fri, 02 Oct 1993 00:00:00 +0000
@@ -153,7 +161,7 @@ are real
 EOF
 	my $ghost_id = $rw->add_message($was_ghost);
 	is($ghost_id, int($ghost_id), "ghost_id is an integer: $ghost_id");
-	my $msgs = $rw->{over}->get_thread('ghost-message@s');
+	my $msgs = $rw->{oidx}->get_thread('ghost-message@s');
 	is(scalar(@$msgs), 2, 'got both messages in ghost thread');
 	foreach (qw(sid tid)) {
 		is($msgs->[0]->{$_}, $msgs->[1]->{$_}, "{$_} match");
@@ -255,13 +263,14 @@ To: list@example.com
 theatre
 fade
 EOF
-	my $res = $rw->query("theatre");
+	$rw_commit->();
+	my $res = $ro->reopen->query("theatre");
 	is(scalar(@$res), 2, "got both matches");
 	if (@$res == 2) {
 		is($res->[0]->{mid}, 'nquote@a', 'non-quoted scores higher');
 		is($res->[1]->{mid}, 'quote@a', 'quoted result still returned');
 	}
-	$res = $rw->query("illusions");
+	$res = $ro->query("illusions");
 	is(scalar(@$res), 1, "got a match for quoted text");
 	is($res->[0]->{mid}, 'quote@a',
 		"quoted result returned if nothing else") if scalar(@$res);
@@ -282,7 +291,8 @@ To: list\@example.com
 LOOP!
 EOF
 	ok($doc_id > 0, "doc_id defined with circular reference");
-	my $smsg = $rw->query('m:circle@a', {limit=>1})->[0];
+	$rw_commit->();
+	my $smsg = $ro->reopen->query('m:circle@a', {limit=>1})->[0];
 	is(defined($smsg), 1, 'found m:circl@a');
 	if (defined $smsg) {
 		is($smsg->{references}, '', "no references created");
@@ -290,11 +300,22 @@ EOF
 	}
 });
 
+{
+	my $msgs = $ro->query('d:19931002..20101002');
+	ok(scalar(@$msgs) > 0, 'got results within range');
+	$msgs = $ro->query('d:20101003..');
+	is(scalar(@$msgs), 0, 'nothing after 20101003');
+	$msgs = $ro->query('d:..19931001');
+	is(scalar(@$msgs), 0, 'nothing before 19931001');
+}
+
 $ibx->with_umask(sub {
 	my $mime = eml_load 't/utf8.eml';
 	my $doc_id = $rw->add_message($mime);
 	ok($doc_id > 0, 'message indexed doc_id with UTF-8');
-	my $msg = $rw->query('m:testmessage@example.com', {limit => 1})->[0];
+	$rw_commit->();
+	my $msg = $ro->reopen->
+		query('m:testmessage@example.com', {limit => 1})->[0];
 	is(defined($msg), 1, 'found testmessage@example.com');
 	if (defined $msg) {
 		is($mime->header('Subject'), $msg->{subject},
@@ -302,19 +323,10 @@ $ibx->with_umask(sub {
 	}
 });
 
-{
-	my $msgs = $ro->query('d:19931002..20101002');
-	ok(scalar(@$msgs) > 0, 'got results within range');
-	$msgs = $ro->query('d:20101003..');
-	is(scalar(@$msgs), 0, 'nothing after 20101003');
-	$msgs = $ro->query('d:..19931001');
-	is(scalar(@$msgs), 0, 'nothing before 19931001');
-}
-
 # names and addresses
 {
 	my $mset = $ro->query('t:list@example.com', {mset => 1});
-	is($mset->size, 6, 'searched To: successfully');
+	is($mset->size, 9, 'searched To: successfully');
 	foreach my $m ($mset->items) {
 		my $smsg = $ro->{over_ro}->get_art($m->get_docid);
 		like($smsg->{to}, qr/\blist\@example\.com\b/, 'to appears');
@@ -324,10 +336,15 @@ $ibx->with_umask(sub {
 		like($bytes, qr/\A[0-9]+\z/, '$bytes stored as digit');
 		ok($bytes > 0, '$bytes is > 0');
 		is($bytes, $smsg->{bytes}, 'bytes Xapian value matches Over');
+
+		$col = PublicInbox::Search::UID();
+		my $uid = PublicInbox::Smsg::get_val($doc, $col);
+		is($uid, $smsg->{num}, 'UID column matches {num}');
+		is($uid, $m->get_docid, 'UID column matches docid');
 	}
 
 	$mset = $ro->query('tc:list@example.com', {mset => 1});
-	is($mset->size, 6, 'searched To+Cc: successfully');
+	is($mset->size, 9, 'searched To+Cc: successfully');
 	foreach my $m ($mset->items) {
 		my $smsg = $ro->{over_ro}->get_art($m->get_docid);
 		my $tocc = join("\n", $smsg->{to}, $smsg->{cc});
@@ -384,7 +401,9 @@ $ibx->with_umask(sub {
 
 $ibx->with_umask(sub {
 	my $amsg = eml_load 't/search-amsg.eml';
-	ok($rw->add_message($amsg), 'added attachment');
+	my $oid = ('0'x40);
+	my $smsg = bless { blob => $oid }, 'PublicInbox::Smsg';
+	ok($rw->add_message($amsg, $smsg), 'added attachment');
 	$rw_commit->();
 	$ro->reopen;
 	my $n = $ro->query('n:attached_fart.txt');
@@ -405,7 +424,8 @@ $ibx->with_umask(sub {
 		$art = $ro->{over_ro}->next_by_mid($mid, \$id, \$prev);
 		ok($art, 'article exists in OVER DB');
 	}
-	$rw->unindex_blob($amsg);
+	$rw->_msgmap_init;
+	$rw->unindex_eml($oid, $amsg);
 	$rw->commit_txn_lazy;
 	SKIP: {
 		skip('$art not defined', 1) unless defined $art;