Unable to convert key: num when using 'partition_on' argument

[austinvdubs]

Hello, I was having an issue with adding the 'partition_on' argument to the query. I outlined it more clearly on Stack Overflow here:
https://stackoverflow.com/questions/73792804/connectorx-unable-to-convert-key-num-when-using-partition-on-argument
(I don't have enough reputation on Stack Overflow to create a new tag for connectorx, so I'm cross posting it here for visibility)

I also submitted another question on StackOverflow about a potential bug in handling cx.read_sql() on a database table that is being concurrently updated (https://stackoverflow.com/questions/73794989/connectorx-cx-read-sql-returns-dataframe-of-all-zeros-rows-intermittently-or)

[wangxiaoying]

Hi @austinvdubs , it looks like the error is caused by partitioning the query on the quoteTimeInLong column. May I ask what's the data type of quoteTimeInLong? It will be helpful to reproduce the error locally.

[austinvdubs]

Hello! Thanks for getting back to me. quoteTimeInLong is an INTEGER data type. I have also tried it on TIMESTAMP and REAL data types. I'm using a local sqlite database

[wangxiaoying]

That's weird, we have a CI test which partition on an integer column and it passes, the DDL can be found here. Can you confirm that the error remains in the newest version? If it is, can you provide a toy example (including DDL, inserting sample data and the query) that throws the error so we can reproduce the error?