Add exception handler for ACL resource exhaustion case

[pshapiro4broad]

Jira ticket: https://broadworkbench.atlassian.net/browse/DT-1097

Addresses

If a user hits a BigQuery quota when creating a snapshot, the error returned doesn't include information about the underlying problem or how to fix it. This change adds that information to the flight's error response.

Summary of changes

In the case where this problem occurs, a GoogleResourceException is thrown with a wrapped cause of BigQueryException with the message Too many authorized entities in this dataset. The maximum number of authorized views, routines, and datasets combined is 2500.. The code checks for this case and returns a custom exception. When the user later retrieves the flight result, they'll get a 400 response and see the custom error message.

$ curl -i -X 'GET' \
  'http://localhost:8080/api/repository/v1/jobs/KenKJ8q8RfmHQmn-8pFlTw/result' \
  -H 'accept: application/json' \
  -H 'Authorization: Bearer <TOKEN>'
HTTP/1.1 400 
X-Request-ID: axRrgBO3
Content-Type: application/json
Content-Length: 222
Date: Wed, 15 Jan 2025 21:45:39 GMT
Connection: close

{"message":"Too many authorized entities in this dataset. The maximum number of authorized views, routines, and datasets combined is 2500. Resolve this by deleting snapshots or creating a second dataset.","errorDetail":[]}
$

Testing Strategy

• run integration tests locally and force the error by modifying the code to throw an exception

[snf2ye]

LGTM - It makes sense that we need to do this per snapshot create type and I like the shared utility. Glad to take another look after tests are added!

[rushtong]

Looks reasonable to me 👍🏽

[fboulnois]

once tests pass, lgtm 👍

[rushtong]

Looks reasonable 👍🏽

[fboulnois]

Looks reasonable 👍

[snf2ye]

One thought, but otherwise LGTM!
As we talked about in standup, if there a lower overhead way to add a unit test, I think that would be good, but not necessary to merge in my opinion.

[sonarqubecloud]

Quality Gate passed

Issues
1 New issue
0 Accepted issues

Measures
0 Security Hotspots
66.7% Coverage on New Code
0.0% Duplication on New Code

See analysis details on SonarQube Cloud