Question

我想使用“ WHERE xxx IN（'aaa'，'bbb'）”执行以下标准SQL。

但是发生了错误。可以在STANDARD SQL中使用'IN'吗？

标准SQL

SELECT commit, author.name, committer.name, committer.time_sec,
committer.tz_offset, committer.date.seconds , subject, message,
repo_name
FROM `bigquery-public-data.github_repos.commits`
WHERE repo_name IN 
('tensorflow/tensorflow', 'facebook/react')

错误

No matching signature for operator IN for argument types ARRAY<STRING> and {STRING} at [5:17]

下面的旧版SQL似乎可以执行。

旧版SQL

SELECT commit, author.name, committer.name, committer.time_sec,
committer.tz_offset, committer.date.seconds , subject, message,
repo_name
FROM [bigquery-public-data:github_repos.commits] 
WHERE repo_name IN 
('tensorflow/tensorflow', 'facebook/react')

Answer 1

您无法在Standard SQL中真正做到这一点。但是在本文How to use the UNNEST function in BigQuery to analyze event parameters in Analytics中，他们对此进行了很好的解释。

我认为这是您获得所需结果的方式：

SELECT commit, author.name as author_name, committer.name as committer_name, committer.time_sec,
committer.tz_offset, committer.date.seconds , subject, message,
repo_name
FROM `bigquery-public-data.github_repos.commits`
CROSS JOIN UNNEST(repo_name) as repo_names_unnested
WHERE repo_names_unnested IN ('tensorflow/tensorflow', 'facebook/react')

请注意，您不能同时拥有author.name和committer.name，因为二者将同时显示为name。因此，我将它们更改为author_name和committer_name。

我还认为您实际上正在寻找的是将repo_name替换为repo_names_unnested的结果，因此请尝试在SELECT子句中也替换它。

Answer 2

以下是用于BigQuery标准SQL

如果您需要保留原始记录并仅在repo_name中输出带有'tensorflow / tensorflow'，'facebook / react'的记录，则为

#standardSQL
SELECT commit, author.name AS author_name, committer.name AS committer_name, committer.time_sec,
committer.tz_offset, committer.date.seconds , subject, message,
repo_name
FROM `bigquery-public-data.github_repos.commits`
WHERE EXISTS (
  SELECT 1 FROM UNNEST(repo_name) name WHERE name IN ('tensorflow/tensorflow', 'facebook/react') 
)

Answer 3

字段“ repo_name”是REPEATED（传统SQL）或ARRAY（标准SQL），IN运算符不支持标准SQL。要将数组转换为一组行，您可以使用UNNEST operator。

对于参数类型ARRAY <STRING>，运算符IN没有匹配的签名

3 个答案: