dovecot-2.2: fts-solr: Avoid sending too large queries to Solr d...

[dovecot at dovecot.org]

details:   http://hg.dovecot.org/dovecot-2.2/rev/1e6f38e8d5d5
changeset: 18877:1e6f38e8d5d5
user:      Timo Sirainen <tss at iki.fi>
date:      Wed Jun 17 14:04:42 2015 +0300
description:
fts-solr: Avoid sending too large queries to Solr due to listing wanted mailboxes.
This is mainly a problem if there are a lot of mailboxes and "All Mails"
virtual mailbox. For now hardcoded to send max 10 mailboxes in the query,
afterwards it just returns matches in all the mailboxes and we'll filter out
the unwanted mailboxes.

diffstat:

 src/plugins/fts-solr/fts-backend-solr.c |  26 +++++++++++++++++++-------
 1 files changed, 19 insertions(+), 7 deletions(-)

diffs (65 lines):

diff -r fec87fb9b071 -r 1e6f38e8d5d5 src/plugins/fts-solr/fts-backend-solr.c
--- a/src/plugins/fts-solr/fts-backend-solr.c	Wed Jun 17 13:55:37 2015 +0300
+++ b/src/plugins/fts-solr/fts-backend-solr.c	Wed Jun 17 14:04:42 2015 +0300
@@ -26,6 +26,8 @@
    header fields as long as they're smaller than this */
 #define SOLR_HEADER_LINE_MAX_TRUNC_SIZE 1024
 
+#define SOLR_QUERY_MAX_MAILBOX_COUNT 10
+
 struct solr_fts_backend {
 	struct fts_backend backend;
 	struct solr_connection *solr_conn;
@@ -831,6 +833,7 @@
 	struct mailbox *box;
 	const char *box_guid;
 	unsigned int i, len;
+	bool search_all_mailboxes;
 
 	/* use a separate filter query for selecting the mailbox. it shouldn't
 	   affect the score and there could be some caching benefits too. */
@@ -841,19 +844,26 @@
 		str_append(str, "%22%22");
 
 	hash_table_create(&mailboxes, default_pool, 0, str_hash, strcmp);
-	str_append(str, "%2B(");
+	for (i = 0; boxes[i] != NULL; i++) ;
+	search_all_mailboxes = i > SOLR_QUERY_MAX_MAILBOX_COUNT;
+	if (!search_all_mailboxes)
+		str_append(str, "%2B(");
 	len = str_len(str);
+
 	for (i = 0; boxes[i] != NULL; i++) {
 		if (fts_mailbox_get_guid(boxes[i], &box_guid) < 0)
 			continue;
 
-		if (str_len(str) != len)
-			str_append(str, "+OR+");
-		str_printfa(str, "box:%s", box_guid);
+		if (!search_all_mailboxes) {
+			if (str_len(str) != len)
+				str_append(str, "+OR+");
+			str_printfa(str, "box:%s", box_guid);
+		}
 		hash_table_insert(mailboxes, t_strdup_noconst(box_guid),
 				  boxes[i]);
 	}
-	str_append_c(str, ')');
+	if (!search_all_mailboxes)
+		str_append_c(str, ')');
 
 	if (solr_connection_select(backend->solr_conn, str_c(str),
 				   result->pool, &solr_results) < 0) {
@@ -865,8 +875,10 @@
 	for (i = 0; solr_results[i] != NULL; i++) {
 		box = hash_table_lookup(mailboxes, solr_results[i]->box_id);
 		if (box == NULL) {
-			i_warning("fts_solr: Lookup returned unexpected mailbox "
-				  "with guid=%s", solr_results[i]->box_id);
+			if (!search_all_mailboxes) {
+				i_warning("fts_solr: Lookup returned unexpected mailbox "
+					  "with guid=%s", solr_results[i]->box_id);
+			}
 			continue;
 		}
 		fts_result = array_append_space(&fts_results);