[test](mtmv) Add inject statistic when mv rewrite regression test to …

…make sure rewrite result stable (apache#43785)

The result of successful rewriting by the cbo optimizer depends on the
statistics.
The priority of the optimizer consumption statistics in descending order
is
1. the injected statistics
2. the statistics reported by be
3. and the statistics analyzed actively.

When the pipeline runs the case, the statistics reported by be may not
be timely. Therefore, the outcome that leads to the cbo optimizer's
successful selection of overwrites is not very certain, so the
statistics are currently injected manually in the test cases

regression-test/data/mv_p0/ssb/q_1_1/q_1_1.out

@@ -6,10 +6,13 @@
 19930101	1	1	1	1	1	1	1	1	1	1	100	1	1	1	2023-06-09	shipmode	name	address	city	nation	AMERICA	phone	mktsegment	name	address	city	nation	AMERICA	phone	name	MFGR#1	category	brand	color	type	4	container
 19930101	1	1	1	1	1	1	1	1	1	1	100	1	1	1	2023-06-09	shipmode	name	address	city	nation	AMERICA	phone	mktsegment	name	address	city	nation	AMERICA	phone	name	MFGR#1	category	brand	color	type	4	container
 19930101	2	2	2	2	2	2	2	2	2	2	2	2	2	2	2023-06-09	shipmode	name	address	city	nation	region	phone	mktsegment	name	address	city	nation	region	phone	name	mfgr	category	brand	color	type	4	container
+19930101	2	2	2	2	2	2	2	2	2	2	2	2	2	2	2023-06-09	shipmode	name	address	city	nation	region	phone	mktsegment	name	address	city	nation	region	phone	name	mfgr	category	brand	color	type	4	container
+19930101	2	2	2	2	2	2	2	2	2	2	2	2	2	2	2023-06-09	shipmode	name	address	city	nation	region	phone	mktsegment	name	address	city	nation	region	phone	name	mfgr	category	brand	color	type	4	container
+19930101	2	2	2	2	2	2	2	2	2	2	2	2	2	2	2023-06-09	shipmode	name	address	city	nation	region	phone	mktsegment	name	address	city	nation	region	phone	name	mfgr	category	brand	color	type	4	container
 
 -- !select_mv --
-4
+16
 
 -- !select --
-4
+16
 

regression-test/framework/src/main/groovy/org/apache/doris/regression/suite/Suite.groovy

@@ -1617,9 +1617,8 @@ class Suite implements GroovyInterceptable {
             check { result ->
                 boolean success = true;
                 for (String mv_name : mv_names) {
-                    success = success && result.contains("${mv_name} chose")
+                    Assert.assertEquals(true, result.contains("${mv_name} chose"))
                 }
-                Assert.assertEquals(true, success)
             }
         }
     }

regression-test/suites/mv_p0/agg_have_dup_base/agg_have_dup_base.groovy

@@ -56,6 +56,8 @@ suite ("agg_have_dup_base") {
     qt_select_mv "select unix_timestamp(k1) tmp,sum(k2) from d_table group by tmp order by tmp;"
 
     sql """set enable_stats=true;"""
+    sql """alter table d_table modify column k4 set stats ('row_count'='5');"""
+
     mv_rewrite_success("select k1,sum(k2),max(k2) from d_table group by k1;", "k12s3m")
 
     mv_rewrite_success("select k1,sum(k2) from d_table group by k1;", "k12s3m")

regression-test/suites/mv_p0/agg_state/test_agg_state_max_by.groovy

@@ -69,6 +69,7 @@ suite ("test_agg_state_max_by") {
     qt_select_star "select * from d_table order by 1,2;"
     mv_rewrite_success("select k1,max_by(k2,k3) from d_table group by k1 order by 1,2;", "k1mb")
     sql """set enable_stats=true;"""
+    sql """alter table d_table modify column k4 set stats ('row_count'='8');"""
     mv_rewrite_success("select k1,max_by(k2,k3) from d_table group by k1 order by 1,2;", "k1mb")
     qt_select_mv "select k1,max_by(k2,k3) from d_table group by k1 order by 1,2;"
 
@@ -101,6 +102,7 @@ suite ("test_agg_state_max_by") {
     qt_select_star "select * from d_table order by 1,2;"
 
     sql """set enable_stats=true;"""
+    sql """alter table d_table modify column k4 set stats ('row_count'='8');"""
     sql "analyze table d_table with sync;"
     sql """set enable_stats=false;"""
 
@@ -114,6 +116,7 @@ suite ("test_agg_state_max_by") {
     qt_select_mv "select k1,max_by(k2,abs(k3)) from d_table group by k1 order by 1,2;"
 
     sql """set enable_stats=true;"""
+    sql """alter table d_table modify column k4 set stats ('row_count'='8');"""
     mv_rewrite_success("select k1,max_by(k2+k3,abs(k3)) from d_table group by k1 order by 1,2;", "k1mbcp1")
     mv_rewrite_success("select k1,max_by(k2+k3,k3) from d_table group by k1 order by 1,2;", "k1mbcp2")
     mv_rewrite_success("select k1,max_by(k2,abs(k3)) from d_table group by k1 order by 1,2;", "k1mbcp3")

regression-test/suites/mv_p0/case_ignore/case_ignore.groovy

@@ -37,6 +37,8 @@ suite ("case_ignore") {
     sql "insert into d_table select 2,2,2,'b';"
     sql "insert into d_table select 3,-3,null,'c';"
 
+    sql """alter table d_table modify column k4 set stats ('row_count'='4');"""
+
     createMV ("create materialized view k12a as select K1,abs(K2) from d_table;")
 
     sql "insert into d_table select -4,-4,-4,'d';"
@@ -53,6 +55,7 @@ suite ("case_ignore") {
     qt_select_mv "select K1,abs(K2) from d_table order by K1;"
 
     sql """set enable_stats=true;"""
+    sql """alter table d_table modify column k4 set stats ('row_count'='8');"""
     mv_rewrite_success("select k1,abs(k2) from d_table order by k1;", "k12a")
     mv_rewrite_success("select K1,abs(K2) from d_table order by K1;", "k12a")
 

regression-test/suites/mv_p0/count_star/count_star.groovy

@@ -64,6 +64,8 @@ suite ("count_star") {
     qt_select_mv "select count(*) from d_table where k3=1;"
 
     sql """set enable_stats=true;"""
+
+    sql """alter table d_table modify column k4 set stats ('row_count'='8');"""
     mv_rewrite_success("select k1,k4,count(*) from d_table group by k1,k4;", "kstar")
     mv_rewrite_success("select k1,k4,count(*) from d_table where k1=1 group by k1,k4;", "kstar")
     mv_rewrite_fail("select k1,k4,count(*) from d_table where k3=1 group by k1,k4;", "kstar")

regression-test/suites/mv_p0/dis_26495/dis_26495.groovy

@@ -28,6 +28,8 @@ suite ("dis_26495") {
 
     sql """insert into doris_test values (1,2,max_by_state(1,2));"""
 
+    sql """alter table doris_test modify column agg_st_1 set stats ('row_count'='1');"""
+
     streamLoad {
         table "doris_test"
         set 'column_separator', ','

regression-test/suites/mv_p0/k1ap2spa/k1ap2spa.groovy

@@ -52,6 +52,8 @@ suite ("k1ap2spa") {
 
     sql """set enable_stats=true;"""
 
+    sql """alter table d_table modify column k1 set stats ('row_count'='5');"""
+
     mv_rewrite_success("select abs(k1)+1 t,sum(abs(k2+1)) from d_table group by t order by t;", "k1ap2spa")
 
 }

regression-test/suites/mv_p0/k1s2m3/k1s2m3.groovy

@@ -36,6 +36,7 @@ suite ("k1s2m3") {
     sql "insert into d_table select 2,2,2,'b';"
     sql "insert into d_table select 3,-3,null,'c';"
 
+    sql """alter table d_table modify column k1 set stats ('row_count'='6');"""
     createMV("create materialized view k1s2m3 as select k1,sum(k2*k3) from d_table group by k1;")
 
     sql "insert into d_table select -4,-4,-4,'d';"

regression-test/suites/mv_p0/k1s2m3_auto_inc/k1s2m3_auto_inc.groovy

@@ -57,5 +57,6 @@ suite ("k1s2m3_auto_inc") {
     qt_select_mv "select k3,sum(abs(k2+1)) from d_table group by k3 order by 1;"
 
     sql """set enable_stats=true;"""
+    sql """alter table d_table modify column k1 set stats ('row_count'='2');"""
     mv_rewrite_success("select k3,sum(abs(k2+1)) from d_table group by k3 order by 1;", "k3ap2spa")
 }

regression-test/suites/mv_p0/multi_agg_with_same_slot/multi_agg_with_same_slot.groovy

@@ -64,6 +64,7 @@ suite ("multi_agg_with_same_slot") {
 
     sql """set enable_stats=true;"""
 
+    sql """alter table d_table modify column k1 set stats ('row_count'='5');"""
     mv_rewrite_success("select k1,k2,avg(k3),max(k3) from d_table group by k1,k2 order by 1,2;", "kmv")
     mv_rewrite_success("select k1,k2,avg(k3)+max(k3) from d_table group by k1,k2 order by 1,2;", "kmv")
     mv_rewrite_success("select k1,k2,avg(k3)+max(k3) from d_table group by grouping sets((k1),(k1,k2),()) order by 1,2;", "kmv")

regression-test/suites/mv_p0/multi_slot_k123p/multi_slot_k123p.groovy

@@ -66,5 +66,6 @@ suite ("multi_slot_k123p") {
     qt_select_mv "select k1,version() from d_table order by k1;"
 
     sql """set enable_stats=true;"""
+    sql """alter table d_table modify column k1 set stats ('row_count'='5');"""
     mv_rewrite_success("select k1,k2+k3 from d_table order by k1;", "k123p")
 }

regression-test/suites/mv_p0/multi_slot_k1a2p2ap3p/multi_slot_k1a2p2ap3p.groovy

@@ -51,6 +51,7 @@ suite ("multi_slot_k1a2p2ap3p") {
     qt_select_mv "select abs(k1)+k2+1,abs(k2+2)+k3+3 from d_table order by abs(k1)+k2+1,abs(k2+2)+k3+3;"
 
     sql """set enable_stats=true;"""
+    sql """alter table d_table modify column k1 set stats ('row_count'='7');"""
     mv_rewrite_success("select abs(k1)+k2+1,abs(k2+2)+k3+3 from d_table order by abs(k1)+k2+1,abs(k2+2)+k3+3", "k1a2p2ap3p")
 
 }

regression-test/suites/mv_p0/multi_slot_k1a2p2ap3ps/multi_slot_k1a2p2ap3ps.groovy

@@ -61,6 +61,7 @@ suite ("multi_slot_k1a2p2ap3ps") {
     qt_select_base "select abs(k1)+k2+1,sum(abs(k2+2)+k3+3) from d_table group by abs(k1)+k2 order by abs(k1)+k2;"
 
     sql """set enable_stats=true;"""
+    sql """alter table d_table modify column k1 set stats ('row_count'='4');"""
     mv_rewrite_success("select abs(k1)+k2+1,sum(abs(k2+2)+k3+3) from d_table group by abs(k1)+k2+1 order by abs(k1)+k2+1", "k1a2p2ap3ps")
 
     mv_rewrite_fail("select abs(k1)+k2+1,sum(abs(k2+2)+k3+3) from d_table group by abs(k1)+k2 order by abs(k1)+k2", "k1a2p2ap3ps")

regression-test/suites/mv_p0/multi_slot_k1p2ap3p/multi_slot_k1p2ap3p.groovy

@@ -50,5 +50,6 @@ suite ("multi_slot_k1p2ap3p") {
     qt_select_mv "select k1+1,abs(k2+2)+k3+3 from d_table order by k1+1;"
 
     sql """set enable_stats=true;"""
+    sql """alter table d_table modify column k1 set stats ('row_count'='4');"""
     mv_rewrite_success("select k1+1,abs(k2+2)+k3+3 from d_table order by k1+1;", "k1p2ap3p")
 }

regression-test/suites/mv_p0/multi_slot_k1p2ap3ps/multi_slot_k1p2ap3ps.groovy

@@ -49,5 +49,6 @@ suite ("multi_slot_k1p2ap3ps") {
     qt_select_mv "select k1+1,sum(abs(k2+2)+k3+3) from d_table group by k1+1 order by k1+1;"
 
     sql """set enable_stats=true;"""
+    sql """alter table d_table modify column k1 set stats ('row_count'='5');"""
     mv_rewrite_success("select k1+1,sum(abs(k2+2)+k3+3) from d_table group by k1+1 order by k1+1;", "k1p2ap3ps")
 }

regression-test/suites/mv_p0/mv_with_view/mv_with_view.groovy

@@ -36,28 +36,25 @@ suite ("mv_with_view") {
     sql """insert into d_table select 1,1,1,'a';"""
     sql """insert into d_table select 2,2,2,'b';"""
 
-    createMV("create materialized view k132 as select k1,k3,k2 from d_table;")
+    createMV("create materialized view k312 as select k3,k1,k2 from d_table;")
 
     sql """insert into d_table select 3,-3,null,'c';"""
 
-    explain {
-        sql("select * from d_table order by k1;")
-        contains "(d_table)"
-    }
+    sql "analyze table d_table with sync;"
+    sql """set enable_stats=false;"""
+
+    mv_rewrite_fail("select * from d_table order by k1;", "k312")
     qt_select_star "select * from d_table order by k1;"
 
     sql """
-        drop view if exists v_k132;
+        drop view if exists v_k312;
     """
 
     sql """
-        create view v_k132 as select k1,k3,k2 from d_table where k1 = 1;
+        create view v_k312 as select k1,k3,k2 from d_table where k3 = 1;
     """
-    explain {
-        sql("select * from v_k132 order by k1;")
-        contains "(k132)"
-    }
-    qt_select_mv "select * from v_k132 order by k1;"
+    mv_rewrite_success("select * from v_k312 order by k1;", "k312")
+    qt_select_mv "select * from v_k312 order by k1;"
 
     sql """
         drop view if exists v_k124;
@@ -66,9 +63,6 @@ suite ("mv_with_view") {
     sql """
         create view v_k124 as select k1,k2,k4 from d_table where k1 = 1;
     """
-    explain {
-        sql("select * from v_k124 order by k1;")
-        contains "(d_table)"
-    }
+    mv_rewrite_fail("select * from v_k124 order by k1;", "k312")
     qt_select_mv "select * from v_k124 order by k1;"
 }

regression-test/suites/mv_p0/null_insert/null_insert.groovy

@@ -76,6 +76,7 @@ suite ("null_insert") {
                     GROUP BY date,vid,os,ver,ip_country;"""
 
     sql """set enable_stats=true;"""
+    sql """alter table test modify column date set stats ('row_count'='3');"""
     mv_rewrite_success("""SELECT date, vid, os, ver, ip_country, hll_union(hll_hash(uid))
                 FROM test
                 GROUP BY date,vid,os,ver,ip_country;""", "mv_test")

regression-test/suites/mv_p0/routine_load_hll/routine_load_hll.groovy

@@ -58,5 +58,6 @@ suite ("routine_load_hll") {
     qt_select_mv "select time_stamp, hll_union_agg(device_id) from test group by time_stamp order by 1;"
 
     sql """set enable_stats=true;"""
+    sql """alter table test modify column event_id set stats ('row_count'='2');"""
     mv_rewrite_success("select time_stamp, hll_union_agg(device_id) from test group by time_stamp order by 1;", "m_view")
 }

regression-test/suites/mv_p0/ssb/multiple_ssb_between/multiple_ssb_between.groovy

@@ -65,14 +65,6 @@ suite ("multiple_ssb_between") {
         ) ENGINE=OLAP
         DUPLICATE KEY(`LO_ORDERDATE`, `LO_ORDERKEY`)
         COMMENT "OLAP"
-        PARTITION BY RANGE(`LO_ORDERDATE`)
-        (PARTITION p1992 VALUES [("-2147483648"), ("19930101")),
-        PARTITION p1993 VALUES [("19930101"), ("19940101")),
-        PARTITION p1994 VALUES [("19940101"), ("19950101")),
-        PARTITION p1995 VALUES [("19950101"), ("19960101")),
-        PARTITION p1996 VALUES [("19960101"), ("19970101")),
-        PARTITION p1997 VALUES [("19970101"), ("19980101")),
-        PARTITION p1998 VALUES [("19980101"), ("19990101")))
         DISTRIBUTED BY HASH(`LO_ORDERKEY`) BUCKETS 48
         PROPERTIES (
         "replication_num" = "1",
@@ -153,17 +145,22 @@ suite ("multiple_ssb_between") {
     sql """INSERT INTO lineorder_flat (LO_ORDERDATE, LO_ORDERKEY, LO_LINENUMBER, LO_CUSTKEY, LO_PARTKEY, LO_SUPPKEY, LO_ORDERPRIORITY, LO_SHIPPRIORITY, LO_QUANTITY, LO_EXTENDEDPRICE, LO_ORDTOTALPRICE, LO_DISCOUNT, LO_REVENUE, LO_SUPPLYCOST, LO_TAX, LO_COMMITDATE, LO_SHIPMODE, C_NAME, C_ADDRESS, C_CITY, C_NATION, C_REGION, C_PHONE, C_MKTSEGMENT, S_NAME, S_ADDRESS, S_CITY, S_NATION, S_REGION, S_PHONE, P_NAME, P_MFGR, P_CATEGORY, P_BRAND, P_COLOR,P_TYPE,P_SIZE,P_CONTAINER) VALUES (1 , 1 , 1 , 1 , 1 , 1 , '1' , 1 , 1 , 1 , 1 , 1 , 1 , 1 , 1 , '2023-06-09' , 'shipmode' , 'name' , 'address' , 'city' , 'nation' , 'AMERICA' , 'phone' , 'mktsegment' , 'name' , 'address' , 'city' , 'nation' , 'AMERICA' ,'phone', 'name', 'MFGR#1', 'category', 'brand', 'color', 'type', 4 ,'container');"""
 
     qt_select_star "select * from lineorder_flat order by 1,2, P_MFGR;"
+
+    sql "analyze table lineorder_flat with sync;"
+    sql """set enable_stats=true;"""
+
+    sql """alter table lineorder_flat modify column LO_ORDERDATE set stats ('row_count'='8');"""
 
-    explain {
-        sql("""SELECT SUM(LO_EXTENDEDPRICE * LO_DISCOUNT) AS revenue
+    mv_rewrite_success("""SELECT SUM(LO_EXTENDEDPRICE * LO_DISCOUNT) AS revenue
                 FROM lineorder_flat
                 WHERE
                     LO_ORDERDATE >= 19930101
                     AND LO_ORDERDATE <= 19931231
                     AND LO_DISCOUNT BETWEEN 1 AND 3
-                    AND LO_QUANTITY < 25;""")
-        contains "(lineorder_q_1_1)"
-    }
+                    AND LO_QUANTITY < 25;""",
+                    "lineorder_q_1_1"
+    )
+
     qt_select_q_1_1 """SELECT SUM(LO_EXTENDEDPRICE * LO_DISCOUNT) AS revenue
                 FROM lineorder_flat
                 WHERE
@@ -172,16 +169,16 @@ suite ("multiple_ssb_between") {
                     AND LO_DISCOUNT BETWEEN 1 AND 3
                     AND LO_QUANTITY < 25;"""
 
-    explain {
-        sql("""SELECT
+    mv_rewrite_success("""SELECT
                 SUM(LO_REVENUE), (LO_ORDERDATE DIV 10000) AS YEAR,
                 P_BRAND
             FROM lineorder_flat
             WHERE P_CATEGORY = 'MFGR#12' AND S_REGION = 'AMERICA'
             GROUP BY (LO_ORDERDATE DIV 10000), P_BRAND
-            ORDER BY YEAR, P_BRAND;""")
-        contains "(lineorder_q_2_1)"
-    }
+            ORDER BY YEAR, P_BRAND;""",
+        "lineorder_q_2_1"
+    )
+
     qt_select_q_2_1 """SELECT
                     SUM(LO_REVENUE), (LO_ORDERDATE DIV 10000) AS YEAR,
                     P_BRAND
@@ -190,8 +187,7 @@ suite ("multiple_ssb_between") {
                 GROUP BY YEAR, P_BRAND
                 ORDER BY YEAR, P_BRAND;"""
 
-    explain {
-        sql("""SELECT
+    mv_rewrite_success("""SELECT
                 C_NATION,
                 S_NATION, (LO_ORDERDATE DIV 10000) AS YEAR,
                 SUM(LO_REVENUE) AS revenue
@@ -202,9 +198,9 @@ suite ("multiple_ssb_between") {
                 AND LO_ORDERDATE >= 19920101
                 AND LO_ORDERDATE <= 19971231
             GROUP BY C_NATION, S_NATION, YEAR
-            ORDER BY YEAR ASC, revenue DESC;""")
-        contains "(lineorder_q_3_1)"
-    }
+            ORDER BY YEAR ASC, revenue DESC;""",
+        "lineorder_q_3_1")
+
     qt_select_q_3_1 """SELECT
                         C_NATION,
                         S_NATION, (LO_ORDERDATE DIV 10000) AS YEAR,
@@ -218,8 +214,7 @@ suite ("multiple_ssb_between") {
                     GROUP BY C_NATION, S_NATION, YEAR
                     ORDER BY YEAR ASC, revenue DESC;"""
 
-    explain {
-        sql("""SELECT (LO_ORDERDATE DIV 10000) AS YEAR,
+    mv_rewrite_success("""SELECT (LO_ORDERDATE DIV 10000) AS YEAR,
                 C_NATION,
                 SUM(LO_REVENUE - LO_SUPPLYCOST) AS profit
                 FROM lineorder_flat
@@ -228,9 +223,9 @@ suite ("multiple_ssb_between") {
                 AND S_REGION = 'AMERICA'
                 AND P_MFGR IN ('MFGR#1', 'MFGR#2')
                 GROUP BY YEAR, C_NATION
-                ORDER BY YEAR ASC, C_NATION ASC;""")
-        contains "(lineorder_q_4_1)"
-    }
+                ORDER BY YEAR ASC, C_NATION ASC;""",
+                "lineorder_q_4_1")
+
     qt_select_q_4_1 """SELECT (LO_ORDERDATE DIV 10000) AS YEAR,
                 C_NATION,
                 SUM(LO_REVENUE - LO_SUPPLYCOST) AS profit