Skip to content

fix(tests): two live tests read the problem the api now sends - #133

Merged
olavgg merged 1 commit into
mainfrom
fix/tests-follow-problem-errors
Sep 18, 2026
Merged

olavgg merged 1 commit into
mainfrom
fix/tests-follow-problem-errors

Conversation

@olavgg

@olavgg olavgg commented Sep 17, 2026

Copy link
Copy Markdown
Contributor

The api's errors series changed two refusals these tests still pinned to their old shape, and both fail against current datahub-platform main:

  • A delete that would strand a node is a 409 would-strand (api commit 7fbbb565), not a 400. AGENTS.md already says so. The test now checks the status and the problem type, and finds the stranded node in blockedBy instead of in the message.
  • Datapoints sent to a series that does not exist get a 404 not-found problem that lists the skipped series in missing (api build: add a script that builds the installable Python packages #98). The old wording, "Could not find following timeseries", is gone. The test checks the problem type and that missing names the series.

Both now read the problem's type and members rather than its prose, as problem_integration does, so rewording a detail cannot break them. The binary-path test is unchanged: the SDK raises that 404 itself.

The api's errors series changed two refusals these tests still pinned
to their old shape, and both fail against current datahub-platform
main:

- A delete that would strand a node is a 409 would-strand (api commit
  7fbbb565), not a 400. AGENTS.md already says so. The test now checks
  the status and the problem type, and finds the stranded node in
  blockedBy instead of in the message.
- Datapoints sent to a series that does not exist get a 404 not-found
  problem that lists the skipped series in `missing` (api #98). The old
  wording, "Could not find following timeseries", is gone. The test
  checks the problem type and that `missing` names the series.

Both now read the problem's type and members rather than its prose,
as problem_integration does, so rewording a detail cannot break them.
The binary-path test is unchanged: the SDK raises that 404 itself.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Signed-off-by: Olav Groenaas Gjerde <olav@intellistream.ai>
@olavgg
olavgg requested a review from JosteinGj September 17, 2026 14:01
@JosteinGj JosteinGj closed this Sep 18, 2026
@JosteinGj JosteinGj reopened this Sep 18, 2026
@olavgg
olavgg merged commit 7a56ec1 into main Sep 18, 2026
31 checks passed
@olavgg
olavgg deleted the fix/tests-follow-problem-errors branch September 18, 2026 08:56
JosteinGj added a commit that referenced this pull request Sep 18, 2026
#133 moved the two Rust tests off the pre-unification error shapes but
left their Python twins matching the string "Could not find following
timeseries", which the api no longer sends.

Both now assert the `not-found` problem type and find the absent series
in the `missing` extension, the way the Rust test does. The 404 is a
partial success: data-points for the other series in the batch were
inserted, and `missing` names the ones that were skipped.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Signed-off-by: jgjesdal <jostein@intellistream.ai>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants