Merged asterix_lsm_stabilization upto r1547
git-svn-id: https://asterixdb.googlecode.com/svn/trunk/asterix@1622 eaa15691-b419-025a-1212-ee371bd00084
diff --git a/asterix-app/src/test/resources/optimizerts/queries/inverted-index-join-noeqjoin/ngram-fuzzyeq-jaccard.aql b/asterix-app/src/test/resources/optimizerts/queries/inverted-index-join-noeqjoin/ngram-fuzzyeq-jaccard.aql
new file mode 100644
index 0000000..1a97936
--- /dev/null
+++ b/asterix-app/src/test/resources/optimizerts/queries/inverted-index-join-noeqjoin/ngram-fuzzyeq-jaccard.aql
@@ -0,0 +1,42 @@
+/*
+ * Description : Fuzzy joins two datasets, DBLP and CSX, based on ~= using Jaccard of their titles' 3-gram tokens.
+ * DBLP has a 3-gram index on title, and we expect the join to be transformed into an indexed nested-loop join.
+ * We expect the top-level equi join introduced because of surrogate optimization to be removed, since it is not necessary.
+ * Success : Yes
+ */
+
+drop dataverse test if exists;
+create dataverse test;
+use dataverse test;
+
+create type DBLPType as closed {
+ id: int32,
+ dblpid: string,
+ title: string,
+ authors: string,
+ misc: string
+}
+
+create type CSXType as closed {
+ id: int32,
+ csxid: string,
+ title: string,
+ authors: string,
+ misc: string
+}
+
+create dataset DBLP(DBLPType) primary key id;
+
+create dataset CSX(CSXType) primary key id;
+
+create index ngram_index on DBLP(title) type ngram(3);
+
+write output to nc1:"rttest/inverted-index-join-noeqjoin_ngram-fuzzyeq-jaccard.adm";
+
+set simfunction 'jaccard';
+set simthreshold '0.5f';
+
+for $a in dataset('DBLP')
+for $b in dataset('CSX')
+where gram-tokens($a.title, 3, false) ~= gram-tokens($b.title, 3, false) and $a.id < $b.id
+return {"atitle": $a.title, "btitle": $b.title}
\ No newline at end of file