Uh oh!
There was an error while loading. Please reload this page.
chore: remove workarounds for BQ Storage issue with small result sets - #133
Conversation
| job = self._make_one(self.JOB_ID, self.QUERY, client) | ||
| tbl = job.to_arrow() | ||
| tbl = job.to_arrow(create_bqstorage_client=False) |
There was a problem hiding this comment.
Before this PR, a BigQuery Storage client was automatically created (a default) behind the scenes and actually hit the backend, but only to trigger a 401 Unauthorized error and fall back to the tabledata.list API that uses the (mocked) BigQuery client.
These create_bqstorage_client=False additions ensure that tests behave as before, i.e. using the (mocked) BigQuery client, but without a futile attempt with a real BQ Storage client.
Some issue started occurring under Python 2.7 all of a sudden: Seems like the Edit: Looking at the release notes, it seems that they dropped Python 2.7 support in version 4.1 already, but forgot to tag it as such. We need to pin Edit 2: Will be fixed in #135. |
shollyman
left a comment
There was a problem hiding this comment.
Thanks for cleaning this up.
Uh oh!
There was an error while loading. Please reload this page.
Uh oh!
There was an error while loading. Please reload this page.
| job = self._make_one(self.JOB_ID, self.QUERY, client) | ||
| tbl = job.to_arrow() | ||
| tbl = job.to_arrow(create_bqstorage_client=False) |
Closes#106.
This PR removes workarounds the BQ Storage bug with small anonymous result sets that has been fixed recently, rendering those workarounds obsolete.
PR checklist