You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Checking whether a database exists for snowflake connection strings currently does not work.
The issue seems to be that the query SELECT 1 is evaluated by the Snowflake query engine without the need for a concrete database existence; thus the result is 1 even though the referenced database might not exist.
The solution would be to just add another case for Snowflake and validate the existence with a Snowflake-specific query.
Another option would be to check whether the snowflake case is covered by another database engine and then handle those together.
➡️ Unfortunatenly, for postgresql and mysql this does not seem to be the case. Other open implementations, e.g. #414, do not seem to handle this as well.
Should I open a quick PR for this? Or is there maybe a cleaner solution?
The text was updated successfully, but these errors were encountered:
sqlalchemy-utils doesn't currently support Snowflake databases. I just checked and found that Snowflake offers no free tier for open source projects, so there isn't a way to test sqlalchemy-utils locally or in CI and begin supporting this.
For this reason, I don't see a path to add Snowflake support. You will need to re-implement database_exists() in your own code, unfortunately.
Checking whether a database exists for snowflake connection strings currently does not work.
The issue seems to be that the query
SELECT 1
is evaluated by the Snowflake query engine without the need for a concrete database existence; thus the result is1
even though the referenced database might not exist.The solution would be to just add another case for Snowflake and validate the existence with a Snowflake-specific query.
Another option would be to check whether the snowflake case is covered by another database engine and then handle those together.
➡️ Unfortunatenly, for
postgresql
andmysql
this does not seem to be the case. Other open implementations, e.g. #414, do not seem to handle this as well.Should I open a quick PR for this? Or is there maybe a cleaner solution?
The text was updated successfully, but these errors were encountered: