使用 Google BigQuery 的逗号作为带有 IN 子句的 UNION ALL

Question

我正在尝试执行以下查询：

SELECT 
  author, link_id, COUNT(link_id) as cnt
FROM 
  [fh-bigquery:reddit_comments.2015_12],
  [fh-bigquery:reddit_comments.2015_11]
WHERE link_id IN (
  SELECT posts.name
  FROM [fh-bigquery:reddit_posts.full_corpus_201512] AS posts
  WHERE posts.subreddit = 'politics'
  ORDER BY posts.created_utc DESC
  LIMIT 300
)
GROUP BY author, link_id
ORDER BY author

我在执行时收到此错误消息：JOIN（包括半连接）和 UNION ALL（逗号、日期范围）不能组合在单个 SELECT 语句中。将 UNION ALL 移动到内部查询或将 JOIN 移动到外部查询。

删除其中一个 comments 表工作正常，但我似乎无法弄清楚 BigQuery 的 Comma as UNION ALL 是如何工作的。我试图将联合移动到内部查询，但我仍然遇到同样的错误。

Answer 1

错误是因为我误解了将 UNION ALL 移动到内部查询。为了解决错误，我不得不将两个表放在一个基本的 select * from ... 中。工作查询如下：

SELECT 
  author, link_id, COUNT(link_id) as cnt
FROM (
  SELECT * 
  FROM
    [fh-bigquery:reddit_comments.2015_12],
    [fh-bigquery:reddit_comments.2015_11]
  )
WHERE link_id IN (
  SELECT posts.name
  FROM [fh-bigquery:reddit_posts.full_corpus_201512] AS posts
  WHERE posts.subreddit = 'politics'
  ORDER BY posts.created_utc DESC
  LIMIT 300
)
GROUP BY author, link_id
ORDER BY author

使用 Google BigQuery 的逗号作为带有 IN 子句的 UNION ALL

Using Google BigQuery's Comma as UNION ALL with IN clause

sql

google-bigquery