Added/Updated tests\bugs\core_1254_test.py: Reimplemented in order to have same results regardless on OptimizeForFirstRows = true / false

tests/bugs/core_1254_test.py

@@ -2,52 +2,103 @@
 
 """
 ID:          issue-1678
-ISSUE:       1678
+ISSUE:       https://github.com/FirebirdSQL/firebird/issues/1678
 TITLE:       Problem with DISTINCT and insensitive collations
-DESCRIPTION:
+DESCRIPTION: See https://github.com/FirebirdSQL/firebird/issues/2965
 JIRA:        CORE-1254
 FBTEST:      bugs.core_1254
+NOTES:
+    1. Confirmed problem on 2.1.3.18185
+       Both queries: 'select ... group by ...' and 'select distinct ...' issued six rows:
+        GROUP_ID     QUESTION 
+        ======== ============ 
+        a                   1 
+        a                   2 
+        a                   3 
+        A                   1 
+        A                   2 
+        A                   3 
+        (instead of expected three rows with 'a' or 'A' in the 1st column).
+        The only correct result issued when index was used.
+
+     2. Values in 1st column can vary if OptimizeForFirstRows = true (FB 5.x+).
+        Because of this, we have to check only COUNT of letters in this column
+        that are unique being compared using case SENSITIVE collation.
+        In all cases (for queries and with/without index) this count must be 1.
 """
 
 import pytest
 from firebird.qa import *
 
-init_script = """CREATE TABLE TEST
-(GROUP_ID VARCHAR(1) CHARACTER SET UTF8 COLLATE UNICODE_CI,
-QUESTION INTEGER,
-SCORE INTEGER);
-COMMIT;
-INSERT INTO TEST (GROUP_ID,QUESTION,SCORE) VALUES ('a',1,1);
-INSERT INTO TEST (GROUP_ID,QUESTION,SCORE) VALUES ('a',2,1);
-INSERT INTO TEST (GROUP_ID,QUESTION,SCORE) VALUES ('a',3,1);
-INSERT INTO TEST (GROUP_ID,QUESTION,SCORE) VALUES ('A',1,1);
-INSERT INTO TEST (GROUP_ID,QUESTION,SCORE) VALUES ('A',2,1);
-INSERT INTO TEST (GROUP_ID,QUESTION,SCORE) VALUES ('A',3,1);
-COMMIT;
-
+init_script = """
+    create table test(
+        group_id varchar(1) character set utf8 collate unicode_ci,
+        question integer,
+        score integer
+    );
+    commit;
+    insert into test (group_id,question,score) values ('a',1,11);
+    insert into test (group_id,question,score) values ('a',3,13);
+    insert into test (group_id,question,score) values ('A',1,14);
+    insert into test (group_id,question,score) values ('a',2,12);
+    insert into test (group_id,question,score) values ('A',2,15);
+    insert into test (group_id,question,score) values ('A',3,16);
+    commit;
+    -- See https://github.com/FirebirdSQL/firebird/issues/2965#issue-866882047
+    -- GROUP BY will use an index on multi-byte or insensitive collation only
+    -- when this index is: 1) UNIQUE and 2) ASCENDING.
+    create UNIQUE index test_gr_que_score on test(group_id, question, score);
+    commit;
 """
 
+
 db = db_factory(charset='UTF8', init=init_script)
 
-test_script = """SELECT GROUP_ID, QUESTION, SUM(SCORE) FROM TEST GROUP BY 1,2;
-SELECT DISTINCT GROUP_ID, QUESTION FROM TEST;"""
+test_script = """
+    --set explain on;
+    --set plan on;
+    set list on;
+    alter index test_gr_que_score inactive;
+    commit;
+    
+    select count(
+        distinct cast( group_id as varchar(1)
+                       -- Check count of unique values in 1st column using
+                       -- case SENSITIVE collation:
+                       -- #########################
+                       character set ascii
+                     )
+                ) as "case_SENSITIVE_distinct_gr_1"
+    from (
+        select group_id, question from test group by 1,2
+    );
+
+    select count( distinct cast(group_id as varchar(1) character set ascii)) as "case_SENSITIVE_distinct_gr_2"
+    from (
+        select distinct group_id, question from test
+    );
+
+    alter index test_gr_que_score active;
+    commit;
+
+    select count( distinct cast(group_id as varchar(1) character set ascii)) as "case_SENSITIVE_distinct_gr_3"
+    from (
+        select group_id, question from test group by 1,2
+    );
+
+    select count( distinct cast(group_id as varchar(1) character set ascii)) as "case_SENSITIVE_distinct_gr_4"
+    from (
+        select distinct group_id, question from test
+    );
+"""
 
 act = isql_act('db', test_script)
 
 expected_stdout = """
-GROUP_ID     QUESTION                   SUM
-======== ============ =====================
-a                   1                     2
-a                   2                     2
-a                   3                     2
-
-
-GROUP_ID     QUESTION
-======== ============
-a                   1
-a                   2
-a                   3
-
+    case_SENSITIVE_distinct_gr_1    1
+    case_SENSITIVE_distinct_gr_2    1
+    case_SENSITIVE_distinct_gr_3    1
+    case_SENSITIVE_distinct_gr_4    1
 """
 
 @pytest.mark.version('>=3')