IN语句中的常量类型不一致时，查询结果会非预期。
b列为VARCHAR类型：
starrocks> select * from t;
+------+------+
| a | b |
+------+------+
| 0 | 1 |
| 1 | abc |
+------+------+
2 rows in set (0.02 sec)
当IN语句中全为字符串时，结果正确：
starrocks> select * from t where b in ('1', 'abc');
+------+------+
| a | b |
+------+------+
| 0 | 1 |
| 1 | abc |
+------+------+
2 rows in set (0.02 sec)
当类型不一致，出现INT和VARCHAR的组合时，只查出了数字
starrocks> select * from t where b in (1, 'abc');
+------+------+
| a | b |
+------+------+
| 0 | 1 |
+------+------+
1 row in set (0.01 sec)
查看执行计划和代码，发现这种情况，会都转换成DOUBLE类型：
starrocks> explain select * from t where b in(1, 'abc');
+-------------------------------------------------------------------------------------------+
| Explain String |
+-------------------------------------------------------------------------------------------+
| PLAN FRAGMENT 0 |
| OUTPUT EXPRS:1: a | 2: b |
| PARTITION: UNPARTITIONED |
| |
| RESULT SINK |
| |
| 1:EXCHANGE |
| |
| PLAN FRAGMENT 1 |
| OUTPUT EXPRS: |
| PARTITION: RANDOM |
| |
| STREAM DATA SINK |
| EXCHANGE ID: 01 |
| UNPARTITIONED |
| |
| 0:OlapScanNode |
| TABLE: t |
| PREAGGREGATION: ON |
| PREDICATES: CAST(2: b AS DOUBLE) IN (1.0, CAST('abc' AS DOUBLE)) |
| partitions=1/1 |
| rollup: t |
| tabletRatio=16/16 |
| tabletList=109098,109100,109102,109104,109106,109108,109110,109112,109114,109116 ... |
| cardinality=1 |
| avgRowSize=5.6 |
| numNodes=0 |
+-------------------------------------------------------------------------------------------+
27 rows in set (0.01 sec)
比较能让人接受的方式，应该是都转换成字符串，比如使用 ‘=’ 的场景：
starrocks> explain select * from t where b = 1;
+-------------------------------------------------------------------------------------------+
| Explain String |
+-------------------------------------------------------------------------------------------+
| PLAN FRAGMENT 0 |
| OUTPUT EXPRS:1: a | 2: b |
| PARTITION: UNPARTITIONED |
| |
| RESULT SINK |
| |
| 1:EXCHANGE |
| |
| PLAN FRAGMENT 1 |
| OUTPUT EXPRS: |
| PARTITION: RANDOM |
| |
| STREAM DATA SINK |
| EXCHANGE ID: 01 |
| UNPARTITIONED |
| |
| 0:OlapScanNode |
| TABLE: t |
| PREAGGREGATION: ON |
| PREDICATES: 2: b = '1' |
| partitions=1/1 |
| rollup: t |
| tabletRatio=16/16 |
| tabletList=109098,109100,109102,109104,109106,109108,109110,109112,109114,109116 ... |
| cardinality=1 |
| avgRowSize=5.6 |
| numNodes=0 |
+-------------------------------------------------------------------------------------------+
27 rows in set (0.01 sec)
帮看看这个场景有没有必要修改一下？