database/doris
2
0
Fork 0
Code Issues Pull Requests Actions 3 Packages Projects Releases Wiki Activity

[Fix](statistics)Fix analyze sql including key word bug (#27321)

Browse Source 
Fix analyze sql including key word bug. Need use `` to wrap column names.
This commit is contained in:
Jibing-Li
2023-11-21 13:15:37 +08:00
committed by GitHub
parent cee8cc44e2
commit dd2e9f682f
3 changed files with 11 additions and 11 deletions
Download Patch File Download Diff File Expand all files Collapse all files

14
fe/fe-core/src/main/java/org/apache/doris/statistics/BaseAnalysisTask.java
View File

@ -57,9 +57,9 @@ public abstract class BaseAnalysisTask {
+ " NULL AS `part_id`, "
+ " COUNT(1) AS `row_count`, "
+ " NDV(`${colName}`) AS `ndv`, "
+ " COUNT(1) - COUNT(${colName}) AS `null_count`, "
+ " CAST(MIN(${colName}) AS STRING) AS `min`, "
+ " CAST(MAX(${colName}) AS STRING) AS `max`, "
+ " COUNT(1) - COUNT(`${colName}`) AS `null_count`, "
+ " CAST(MIN(`${colName}`) AS STRING) AS `min`, "
+ " CAST(MAX(`${colName}`) AS STRING) AS `max`, "
+ " ${dataSizeFunction} AS `data_size`, "
+ " NOW() AS `update_time` "
+ " FROM `${catalogName}`.`${dbName}`.`${tblName}`";
@ -91,13 +91,13 @@ public abstract class BaseAnalysisTask {
+ "NULL AS `part_id`, "
+ "${rowCount} AS `row_count`, "
+ "${ndvFunction} as `ndv`, "
+ "IFNULL(SUM(IF(`t1`.`column_key` IS NULL, `t1`.count, 0)), 0) * ${scaleFactor} as `null_count`, "
+ "IFNULL(SUM(IF(`t1`.`column_key` IS NULL, `t1`.`count`, 0)), 0) * ${scaleFactor} as `null_count`, "
+ "'${min}' AS `min`, "
+ "'${max}' AS `max`, "
+ "${dataSizeFunction} * ${scaleFactor} AS `data_size`, "
+ "NOW() "
+ "FROM ( "
+ " SELECT t0.`${colName}` as column_key, COUNT(1) as `count` "
+ " SELECT t0.`${colName}` as `column_key`, COUNT(1) as `count` "
+ " FROM "
+ " (SELECT `${colName}` FROM `${catalogName}`.`${dbName}`.`${tblName}` "
+ " ${sampleHints} ${limit}) as `t0` "
@ -260,8 +260,8 @@ public abstract class BaseAnalysisTask {
}
protected String getNdvFunction(String totalRows) {
String sampleRows = "SUM(t1.count)";
String onceCount = "SUM(IF(t1.count = 1, 1, 0))";
String sampleRows = "SUM(`t1`.`count`)";
String onceCount = "SUM(IF(`t1`.`count` = 1, 1, 0))";
String countDistinct = "COUNT(1)";
// DUJ1 estimator: n*d / (n - f1 + f1*n/N)
// f1 is the count of element that appears only once in the sample.

4
fe/fe-core/src/test/java/org/apache/doris/statistics/BaseAnalysisTaskTest.java
View File

@ -55,8 +55,8 @@ public class BaseAnalysisTaskTest {
Assertions.assertEquals("NULL", maxFunction);
String ndvFunction = olapAnalysisTask.getNdvFunction(String.valueOf(100));
Assertions.assertEquals("SUM(t1.count) * COUNT(1) / (SUM(t1.count) - SUM(IF(t1.count = 1, 1, 0)) "
+ "+ SUM(IF(t1.count = 1, 1, 0)) * SUM(t1.count) / 100)", ndvFunction);
Assertions.assertEquals("SUM(`t1`.`count`) * COUNT(1) / (SUM(`t1`.`count`) - SUM(IF(`t1`.`count` = 1, 1, 0)) "
+ "+ SUM(IF(`t1`.`count` = 1, 1, 0)) * SUM(`t1`.`count`) / 100)", ndvFunction);
System.out.println(ndvFunction);
}

4
fe/fe-core/src/test/java/org/apache/doris/statistics/OlapAnalysisTaskTest.java
View File

@ -151,7 +151,7 @@ public class OlapAnalysisTaskTest {
@Mock
public void runQuery(String sql, boolean needEncode) {
Assertions.assertFalse(needEncode);
Assertions.assertEquals("SELECT CONCAT('30001', '-', '-1', '-', 'null') AS `id`, 10001 AS `catalog_id`, 20001 AS `db_id`, 30001 AS `tbl_id`, -1 AS `idx_id`, 'null' AS `col_id`, NULL AS `part_id`, 500 AS `row_count`, SUM(t1.count) * COUNT(1) / (SUM(t1.count) - SUM(IF(t1.count = 1, 1, 0)) + SUM(IF(t1.count = 1, 1, 0)) * SUM(t1.count) / 500) as `ndv`, IFNULL(SUM(IF(`t1`.`column_key` IS NULL, `t1`.count, 0)), 0) * 5.0 as `null_count`, 'MQ==' AS `min`, 'Mg==' AS `max`, SUM(LENGTH(`column_key`) * count) * 5.0 AS `data_size`, NOW() FROM ( SELECT t0.`${colName}` as column_key, COUNT(1) as `count` FROM (SELECT `${colName}` FROM `catalogName`.`${dbName}`.`${tblName}` limit 100) as `t0` GROUP BY `t0`.`${colName}` ) as `t1` ", sql);
Assertions.assertEquals("SELECT CONCAT('30001', '-', '-1', '-', 'null') AS `id`, 10001 AS `catalog_id`, 20001 AS `db_id`, 30001 AS `tbl_id`, -1 AS `idx_id`, 'null' AS `col_id`, NULL AS `part_id`, 500 AS `row_count`, SUM(`t1`.`count`) * COUNT(1) / (SUM(`t1`.`count`) - SUM(IF(`t1`.`count` = 1, 1, 0)) + SUM(IF(`t1`.`count` = 1, 1, 0)) * SUM(`t1`.`count`) / 500) as `ndv`, IFNULL(SUM(IF(`t1`.`column_key` IS NULL, `t1`.`count`, 0)), 0) * 5.0 as `null_count`, 'MQ==' AS `min`, 'Mg==' AS `max`, SUM(LENGTH(`column_key`) * count) * 5.0 AS `data_size`, NOW() FROM ( SELECT t0.`${colName}` as `column_key`, COUNT(1) as `count` FROM (SELECT `${colName}` FROM `catalogName`.`${dbName}`.`${tblName}` limit 100) as `t0` GROUP BY `t0`.`${colName}` ) as `t1` ", sql);
return;
}
};
@ -292,7 +292,7 @@ public class OlapAnalysisTaskTest {
@Mock
public void runQuery(String sql, boolean needEncode) {
Assertions.assertFalse(needEncode);
Assertions.assertEquals("SELECT CONCAT('30001', '-', '-1', '-', 'null') AS `id`, 10001 AS `catalog_id`, 20001 AS `db_id`, 30001 AS `tbl_id`, -1 AS `idx_id`, 'null' AS `col_id`, NULL AS `part_id`, 500 AS `row_count`, SUM(t1.count) * COUNT(1) / (SUM(t1.count) - SUM(IF(t1.count = 1, 1, 0)) + SUM(IF(t1.count = 1, 1, 0)) * SUM(t1.count) / 500) as `ndv`, IFNULL(SUM(IF(`t1`.`column_key` IS NULL, `t1`.count, 0)), 0) * 5.0 as `null_count`, 'MQ==' AS `min`, 'Mg==' AS `max`, SUM(LENGTH(`column_key`) * count) * 5.0 AS `data_size`, NOW() FROM ( SELECT t0.`${colName}` as column_key, COUNT(1) as `count` FROM (SELECT `${colName}` FROM `catalogName`.`${dbName}`.`${tblName}` limit 100) as `t0` GROUP BY `t0`.`${colName}` ) as `t1` ", sql);
Assertions.assertEquals("SELECT CONCAT('30001', '-', '-1', '-', 'null') AS `id`, 10001 AS `catalog_id`, 20001 AS `db_id`, 30001 AS `tbl_id`, -1 AS `idx_id`, 'null' AS `col_id`, NULL AS `part_id`, 500 AS `row_count`, SUM(`t1`.`count`) * COUNT(1) / (SUM(`t1`.`count`) - SUM(IF(`t1`.`count` = 1, 1, 0)) + SUM(IF(`t1`.`count` = 1, 1, 0)) * SUM(`t1`.`count`) / 500) as `ndv`, IFNULL(SUM(IF(`t1`.`column_key` IS NULL, `t1`.`count`, 0)), 0) * 5.0 as `null_count`, 'MQ==' AS `min`, 'Mg==' AS `max`, SUM(LENGTH(`column_key`) * count) * 5.0 AS `data_size`, NOW() FROM ( SELECT t0.`${colName}` as `column_key`, COUNT(1) as `count` FROM (SELECT `${colName}` FROM `catalogName`.`${dbName}`.`${tblName}` limit 100) as `t0` GROUP BY `t0`.`${colName}` ) as `t1` ", sql);
return;
}
};