bazel-contrib / bazel-contrib/rules_python

Empty coverage reports for Python 3.10.15 and 3.10.16

Offen
#2,844 1 Kommentar 0 Reaktionen 0 zugewiesene Personen Auf GitHub ansehen
help wanted type: toolchain
Vorherrschende Sprache
Starlark
Sterne
688
Forks
721
Ø Merge
15 Std. 7 Min.
Gemergte PRs (30 T.)
76

Beschreibung

# 🐞 bug report

### Affected Rule

rules_python

### Is this a regression?

No

### Description

Running `bazel coverage` does not produce coverage reports for all versions of Python. The coverage.dat files are empty when using Python version 3.10.15 and 3.10.16. Other versions of Python 3.10 do seem to produce valid coverage reports.

I did not test Python 3.11 or 3.12.

## 🔬 Minimal Reproduction

I added tests on a fork of main: [python-3-10-16-coverage](https://github.com/keesvanhespen/rules_python/tree/python-3-10-16-coverage).

Running

`bazel coverage --action_env=COVERAGE_GCOV_OPTIONS=-b --combined_report=lcov --strategy=CoverageReport=local -- //test_code/...
`

from the `tests/coverage` directory returns a filled coverage.dat file in bazel-out for Python 3.10.14, but it is empty for 3.10.16.

## 🌍 Your Environment

**Operating System:**

  

MacOS Sequoia ARM 64

**Output of `bazel version`:**

  

Aspect CLI version: 2025.06.52+25667cae2
INFO: Invocation ID: 58fd115a-8c31-42c8-a32f-e11056aa355a
Build label: 8.2.1
Build target: @@//src/main/java/com/google/devtools/build/lib/bazel:BazelServer
Build time: Thu Apr 17 18:31:45 2025 (1744914705)
Build timestamp: 1744914705
Build timestamp as int: 1744914705

**Rules_python version:**

  

1.3.0

**Anything else relevant?**

Beitragsleitfaden

Beitragsleitfaden öffnen

Rechercherichtung

Start in tests/coverage and compare the reported bazel coverage command and coverage.dat output for Python 3.10.14 versus 3.10.15 and 3.10.16, using the linked python-3-10-16-coverage branch as the reproduction. Done means coverage.dat is populated for the affected Python versions and the coverage tests pass.

Vom Indexierungsmodell aus dem Issue-Text verfasst.

Bewertung

Tech-Stack
python
Bereich
build-system, testing
Issue-Typ
Bug
Schwierigkeit
4/5
Geschätzter Aufwand
3-5 Tage
Aktivitätsstatus
Veraltet
Klarheit
Größtenteils klar
Anfängerfreundlichkeit
35/100

Neue Issues direkt in Ihr Postfach

Eine kurze Übersicht über anfängerfreundliche GitHub-Issues.