Question

我有两个表，我试图加入bigquery。我正在加入他们的序列号，但是对于其中一个表，序列号以5个零开头而另一个表没有。如何使用截断的值编写此连接？

我可以在联接中执行此操作，还是必须先执行此操作并将其另存为新表？

SELECT * FROM [dataworks-356fa:FirebaseArchive.PT_Results] AS Model_Results
JOIN [dataworks-356fa:FirebaseArchive.Master_List_PT] as Master_List on Serial=Serial_Number

具有5个零的表是PT_Results。

例如...... PT_Results serial＃= 0000028106 Master_list_PT serial＃= 28106

Answer 1

尝试将两列都转换为int：

SELECT *
FROM
(
    SELECT
        SUBSTR(CAST(CAST(Serial AS INT) AS STRING), 1, 5) AS Serial,
        Latest_Use,
        Total_Hours,
        Devices_Connected
    FROM [dataworks-356fa:FirebaseArchive.PT_Results]
) AS Model_Results
JOIN
(
    SELECT
       SUBSTR(CAST(CAST(Serial_Number AS INT) AS STRING), 1, 5) AS Serial_Number,
       Status,
       Model_Number
    FROM [dataworks-356fa:FirebaseArchive.Master_List_PT]
) AS Master_List
    ON Model_Results.Serial = Master_List.Serial_Number

将Serial转换为整数应该从比较的角度删除前导零。请注意，我使用派生表来执行转换，以便ON子句仅涉及纯列而不涉及函数。

<强>更新

如果你想比较两个串行列之后的前五位数已经剥离了前导零，你可以尝试双重投射：

SUBSTR(CAST(CAST(Serial AS INT) AS STRING), 1, 5)

在Bigquery中

1 个答案: