Closed BenWu closed 2 months ago
sql.diff
⚠️ Only part of the diff is displayed.
sql.diff
⚠️ Only part of the diff is displayed.
temp udfs turn a query into a script when running in a session
Is that a BigQuery bug?
temp udfs turn a query into a script when running in a session
Is that a BigQuery bug?
I thought so at first but I think it's intended because temp udfs are session-scoped https://cloud.google.com/bigquery/docs/sessions-write-queries#use_temporary_functions_in_sessions
sql.diff
⚠️ Only part of the diff is displayed.
sql.diff
⚠️ Only part of the diff is displayed.
a little late on this but is there a reason to keep these udfs as temporary vs adding them to our udf deployment?
Do you mean generally or specifically for the events stream udfs? Generally, temp udfs do something specific to a single query so it doesn't make sense to share them. I think that applies to the events stream udfs because I wrote them with the data types in those columns i mind. I couldn't come up with one that works generally like JSON.parse
which is why I wrote two. I think it would be confusing to put them in the udf datasets because the error out if not being used in these specific cases
fixes DENG-3905
temp udfs turn a query into a script when running in a session (i.e.
--billing-project
is set) and scripts can't have a destination table. This creates the temp udfs in a separate query.Also reverts the workaround in #5666 for readability.
Checklist for reviewer:
<username>:<branch>
of the fork as parameter. The parameter will also show up in the logs of themanual-trigger-required-for-fork
CI task together with more detailed instructions.For modifications to schemas in restricted namespaces (see
CODEOWNERS
):┆Issue is synchronized with this Jira Task