问题
I have written below cursor in SQL and working file. But I am not able to run the same cursor on snowflake, please help.
DECLARE @CurrentMonth NVARCHAR(100)
DECLARE @CurrentMonth1 NVARCHAR(100)
DECLARE MYDateCURSOR CURSOR
DYNAMIC
FOR
SELECT Collections_COE FROM [CollectionsAgeing_OTCN024_028_029]
OPEN MYDateCURSOR
FETCH LAST FROM MYDateCURSOR INTO @CurrentMonth
CLOSE MYDateCURSOR
DEALLOCATE MYDateCURSOR
--select value from STRING_SPLIT(@CurrentMonth,'-') ;
select @CurrentMonth1=LEFT(@CurrentMonth,4)+cast(cast(RIGHT(@CurrentMonth,2) as int)-1 as varchar(2))
select Date, x1,y1,x1/y1 as DividedValue, round((x1/y1)-1,5) as Actual from(
SELECT X.Past_Due_0 as x1,Y.Past_Due_0 as y1,X.Collections_COE as Date FROM [CollectionsAgeing_OTCN024_028_029] X
CROSS JOIN [CollectionsAgeing_OTCN024_028_029] Y
WHERE X.Collections_COE=@CurrentMonth and y.Collections_COE=@CurrentMonth1
)z
回答1:
Snowflake SQL does not support native SQL cursors. I believe you'll want to leverage a Javascript UDF, UDTF, or Stored Procedure (depending on what you want your output to be). Based on your example above, I believe a Javascript UDTF is what you need: https://docs.snowflake.net/manuals/sql-reference/udf-js-table-functions.html
回答2:
Snowflake SQL does not support native SQL cursors there are options which can be leveraged for your scenario. Please take a look at below links. Also please be informed that Snowflake's real processing power in terms of performance is when data is processed in bulk instead of processing data row by row.
https://community.snowflake.com/s/question/0D50Z00009f7StWSAU/i-have-written-below-cursor-in-sql-and-working-file-but-i-am-not-able-to-run-the-same-cursor-on-snowflake-please-help
https://docs.snowflake.com/en/user-guide/python-connector-example.html
https://docs.snowflake.com/en/sql-reference/udf-js-table-functions.html
来源：https://stackoverflow.com/questions/58482326/how-to-run-cursor-in-snowflake