bug: anyOf sql types not parsed properly #1774
Comments
pnadolny13
added a commit
to MeltanoLabs/target-snowflake
that referenced
this issue
Jun 20, 2023
Closes #62 Test added before fix, failures in https://github.com/MeltanoLabs/target-snowflake/actions/runs/5316144247/jobs/9625351211?pr=63. - adds a tests to alter from ~timestamp to string and~ string to integer. UPDATE: turns out snowflake doesnt allow timestamp to string like I expected so it only alters string to int. - overrides _jsonschema_type_check method to fix bug from meltano/sdk#1774 - overrides `get_table_columns` to cast custom types from sqlalchemy inspect to our custom types with `python_type` implemented --------- Co-authored-by: Ken Payne <[email protected]>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
This json type is returning a VARCHAR sql type when it should be returning a DATETIME:
{"anyOf": [{"type": "string", "format": "date-time"}, {"type": "null"}]}It looks like the bug is in the fact that to_sql_type is calling _jsonschema_type_check to check if its considered a string and its returning false because the anyof logic is looking for the
{"type": "string", "format": "date-time"}in("string",)when really it should be calling itself again for each anyof item.The text was updated successfully, but these errors were encountered: