Google bigquery BigQuery-仅按分隔符拆分一次

Google bigquery BigQuery-仅按分隔符拆分一次,google-bigquery,Google Bigquery,有没有一种方法可以只被分隔符拆分一次?我的数据可能在多个索引处有分隔符。我希望能把一个场分解成两个独立的场 例如,当使用句点作为分隔符时,我希望字符串how.now.brown.cow拆分为两个字段:[how,now.brown.cow] SPLIT({field},'delimiter')[SAFE_OFFSET(0)]可以很好地获得第一部分,但是我的数据中可能有不相等的数组长度,因此连接其他索引时遇到问题。下面是针对BigQuery标准SQL的 #standardSQL WITH `proj

有没有一种方法可以只被分隔符拆分一次?我的数据可能在多个索引处有分隔符。我希望能把一个场分解成两个独立的场

例如,当使用句点作为分隔符时,我希望字符串how.now.brown.cow拆分为两个字段:[how,now.brown.cow]


SPLIT({field},'delimiter')[SAFE_OFFSET(0)]可以很好地获得第一部分,但是我的数据中可能有不相等的数组长度,因此连接其他索引时遇到问题。

下面是针对BigQuery标准SQL的

#standardSQL
WITH `project.dataset.table` AS (
  SELECT 'how.now.brown.cow' col UNION ALL
  SELECT 'how'
)
SELECT col, 
  SPLIT(col, '.')[OFFSET(0)] AS first_item,
  ( SELECT STRING_AGG(item, '.' ORDER BY OFFSET)
    FROM UNNEST(SPLIT(col, '.')) item WITH OFFSET 
    WHERE OFFSET > 0
  ) AS rest_of_items
FROM `project.dataset.table`  
有输出

Row col                 first_item  rest_of_items    
1   how.now.brown.cow   how         now.brown.cow    
2   how                 how         null     
注意:以上只是其中一种方法。看起来有很多方法可以达到相同的效果-例如

#standardSQL
WITH `project.dataset.table` AS (
  SELECT 'how.now.brown.cow' col UNION ALL
  SELECT 'how'
)
SELECT col, 
  REGEXP_EXTRACT(col, r'^([^.]*)\.?') AS first_item,
  REGEXP_EXTRACT(col, r'^[^.]*\.?(.*)$') AS rest_of_items
FROM `project.dataset.table`
有输出

Row col                 first_item  rest_of_items    
1   how.now.brown.cow   how         now.brown.cow    
2   how                 how      

下面是BigQuery标准SQL

#standardSQL
WITH `project.dataset.table` AS (
  SELECT 'how.now.brown.cow' col UNION ALL
  SELECT 'how'
)
SELECT col, 
  SPLIT(col, '.')[OFFSET(0)] AS first_item,
  ( SELECT STRING_AGG(item, '.' ORDER BY OFFSET)
    FROM UNNEST(SPLIT(col, '.')) item WITH OFFSET 
    WHERE OFFSET > 0
  ) AS rest_of_items
FROM `project.dataset.table`  
有输出

Row col                 first_item  rest_of_items    
1   how.now.brown.cow   how         now.brown.cow    
2   how                 how         null     
注意:以上只是其中一种方法。看起来有很多方法可以达到相同的效果-例如

#standardSQL
WITH `project.dataset.table` AS (
  SELECT 'how.now.brown.cow' col UNION ALL
  SELECT 'how'
)
SELECT col, 
  REGEXP_EXTRACT(col, r'^([^.]*)\.?') AS first_item,
  REGEXP_EXTRACT(col, r'^[^.]*\.?(.*)$') AS rest_of_items
FROM `project.dataset.table`
有输出

Row col                 first_item  rest_of_items    
1   how.now.brown.cow   how         now.brown.cow    
2   how                 how