Fix Report.save_report hang on actively-running projects#181
Open
Fix Report.save_report hang on actively-running projects#181
Conversation
save_report's polling loop re-called request every iteration, hitting the report-creation endpoint each time. On a project that is still receiving task responses, every new response invalidates the server-side cached report, so each iteration of the loop kicks off a brand-new generation job and the report never reaches READY before the next response invalidates it. Result: save_report hangs until poll_time expires. Switch the polling to mirror how a refresh-style consumer would do it: call request once to start (or reuse) a generation job, capture the returned job_id, then poll check_status against that specific job. The job will COMPLETED (or ERROR) regardless of incoming responses, so the loop terminates. Also adds a poll_interval kwarg (default 2s, matching previous behavior) and tests for the READY-immediate, polling, error, and timeout paths. Bumps version to 1.5.22.
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 016fcc3f50
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
Per Codex review: check_status's IN_PROGRESS response (HTTP 202) is
documented as `{status: "IN_PROGRESS"}` with no job_id, so reading
response.job_id on every loop iteration would AttributeError on the
second poll for any report that takes more than poll_interval to
generate.
Capture the job_id once from the initial CREATING response and reuse
it across polls. Update only when the server returns RETRYING (which
does carry a new job_id, indicating a fresh underlying job).
The previous test masked the bug by giving the mocked IN_PROGRESS
response a job_id attribute it wouldn't have in reality. Fix the
mock and add a test for the RETRYING job_id switch.
Pre-existing leak: gzip.open(tmp_file.name, 'r').read() returned the buffer but never closed the underlying file. Wrap the gzip.open in a with block so the handle is closed deterministically rather than relying on garbage collection.
NamedTemporaryFile can't be reopened by path on Windows while the original handle is open. Switch from gzip.open(tmp_file.name) to gzip.GzipFile(fileobj=tmp_file) so we reuse the existing handle. Pre-existing issue, but I'm rewriting this block.
timbauman
approved these changes
May 6, 2026
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Report.save_reportcan hang on projects that are still receiving task responses: each polling iteration re-callsrequest, which re-enters the report-generation flow on the server. Every incoming response invalidates the cached report, so each iteration kicks off a brand-new generation job and the report never reachesREADYbefore the next response invalidates it.This PR switches the polling to mirror a refresh-style consumer: call
requestonce to start (or reuse) a generation job, capture the returnedjob_id, and then pollcheck_statusagainst that specific job. The job will eventuallyCOMPLETED(orERROR) regardless of incoming responses, so the loop terminates.Changes
surge/reports.py—save_reportnow callsrequestonce and pollscheck_status(project_id, job_id). Newpoll_intervalkwarg (default 2s, matching previous behavior). Cleaned up the download path slightly (removed the inner-response-shadowing in theurlopenblock; preserved the str-vs-IOfilepathhandling).tests/test_reports.py— adds tests for READY-immediate, CREATING→IN_PROGRESS→COMPLETED, error status, and timeout. Existing empty-project test preserved.setup.py— bump 1.5.21 → 1.5.22.Behavior preserved
download_json(which wrapssave_report) inherits the fix automatically.SurgeRequestErroris unchanged.Test plan
pytest— 40 passedyapf -d -r -e '.cci_pycache/**' .clean