Google bigquery BigQuery中sql函数的执行顺序是什么

我需要对REGEXP\u REPLACE找到的字符串元素进行转换，但是函数SUBSTR看起来像是在REGEXP\u REPLACE之前执行的

我无法找到

替换
的任何限制

WITH tabA AS
  (SELECT "[\"5cc623dd-41f5-42d9-9637-a169af42e2b1\",\"6cc623dd-41f5-42d9-9637-a169af42e2b1\"]" as myids
  union all
  SELECT "[\"5cc623dd-41f5-42d9-9637-a169af42e2b1\"]" as myids
  )

SELECT
  myids,
  REGEXP_REPLACE(myids,"\"(.{36})\"",SUBSTR("\\1",0,8)) as ccc,
  SUBSTR(myids,0,8) as ddd
FROM tabA;  

我希望只输出每个正则表达式的前8个字符，而不是全部36个字符

预计将看到：

5cc623dd,6cc623dd
5cc623dd

在查询中，SUBSTR（…）是函数REGEXP_REPLACE（）的输入的一部分，因此必须在调用REGEXP_REPLACE（）之前对其求值

必须将SUBSTR等效项作为正则表达式的一部分，如下面的示例所示

WITH tabA AS
  (SELECT "[\"5cc623dd-41f5-42d9-9637-a169af42e2b1\",\"6cc623dd-41f5-42d9-9637-a169af42e2b1\"]" as myids
  union all
  SELECT "[\"5cc623dd-41f5-42d9-9637-a169af42e2b1\"]" as myids
  )

SELECT
  myids,
  REGEXP_REPLACE(myids,"\"(.{8}).{28}\"","\\1") as ccc
FROM tabA; 

输出：

+---------------------------------------------------------------------------------+---------------------+
|                                      myids                                      |         ccc         |
+---------------------------------------------------------------------------------+---------------------+
| ["5cc623dd-41f5-42d9-9637-a169af42e2b1","6cc623dd-41f5-42d9-9637-a169af42e2b1"] | [5cc623dd,6cc623dd] |
| ["5cc623dd-41f5-42d9-9637-a169af42e2b1"]                                        | [5cc623dd]          |
+---------------------------------------------------------------------------------+---------------------+

在查询中，SUBSTR（…）是函数REGEXP_REPLACE（）的输入的一部分，因此必须在调用REGEXP_REPLACE（）之前对其求值

必须将SUBSTR等效项作为正则表达式的一部分，如下面的示例所示

WITH tabA AS
  (SELECT "[\"5cc623dd-41f5-42d9-9637-a169af42e2b1\",\"6cc623dd-41f5-42d9-9637-a169af42e2b1\"]" as myids
  union all
  SELECT "[\"5cc623dd-41f5-42d9-9637-a169af42e2b1\"]" as myids
  )

SELECT
  myids,
  REGEXP_REPLACE(myids,"\"(.{8}).{28}\"","\\1") as ccc
FROM tabA; 

输出：

+---------------------------------------------------------------------------------+---------------------+
|                                      myids                                      |         ccc         |
+---------------------------------------------------------------------------------+---------------------+
| ["5cc623dd-41f5-42d9-9637-a169af42e2b1","6cc623dd-41f5-42d9-9637-a169af42e2b1"] | [5cc623dd,6cc623dd] |
| ["5cc623dd-41f5-42d9-9637-a169af42e2b1"]                                        | [5cc623dd]          |
+---------------------------------------------------------------------------------+---------------------+

下面是BigQuery标准SQL

#standardSQL
WITH tabA AS (
  SELECT "[\"5cc623dd-41f5-42d9-9637-a169af42e2b1\",\"6cc623dd-41f5-42d9-9637-a169af42e2b1\"]" AS myids UNION ALL
  SELECT "[\"5cc623dd-41f5-42d9-9637-a169af42e2b1\"]" AS myids
)
SELECT myids, 
  ( SELECT STRING_AGG(id)
    FROM UNNEST(REGEXP_EXTRACT_ALL(myids, r'"(.{8})-')) id
  ) ids  
FROM tabA  

结果

Row myids                                                                           ids  
1   ["5cc623dd-41f5-42d9-9637-a169af42e2b1","6cc623dd-41f5-42d9-9637-a169af42e2b1"] 5cc623dd,6cc623dd    
2   ["5cc623dd-41f5-42d9-9637-a169af42e2b1"]                                        5cc623dd    

下面是BigQuery标准SQL

#standardSQL
WITH tabA AS (
  SELECT "[\"5cc623dd-41f5-42d9-9637-a169af42e2b1\",\"6cc623dd-41f5-42d9-9637-a169af42e2b1\"]" AS myids UNION ALL
  SELECT "[\"5cc623dd-41f5-42d9-9637-a169af42e2b1\"]" AS myids
)
SELECT myids, 
  ( SELECT STRING_AGG(id)
    FROM UNNEST(REGEXP_EXTRACT_ALL(myids, r'"(.{8})-')) id
  ) ids  
FROM tabA  

结果

Row myids                                                                           ids  
1   ["5cc623dd-41f5-42d9-9637-a169af42e2b1","6cc623dd-41f5-42d9-9637-a169af42e2b1"] 5cc623dd,6cc623dd    
2   ["5cc623dd-41f5-42d9-9637-a169af42e2b1"]                                        5cc623dd    

您可能希望提供预期输出的示例（基于显示的输入）您可能希望提供预期输出的示例（基于显示的输入）来回答我关于“执行顺序”的问题，而下面的@MikhailBerlyant answer提供了执行更复杂事情的选项-例如，应用SHA256（）等函数.这就是我关于“执行顺序”的问题的答案，而下面的@MikhailBerlyant answer提供了执行更复杂操作的选项-例如，应用SHA256（）之类的函数。