extindex: support `--gc' to remove dead inboxes

[public-inbox.git] / lib / PublicInbox / SearchIdx.pm

diff --git a/lib/PublicInbox/SearchIdx.pm b/lib/PublicInbox/SearchIdx.pm
index d06c159b8d48bdc052555104ddbbf614cba2d33b..c18c7c367c8f76beeaa011af2ee4165ad58beb0b 100644 (file)
--- 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;