fix: silently skip duplicate TestFiles.add() instead of raising#2144
Merged
aseembits93 merged 8 commits intomainfrom May 8, 2026
Merged
fix: silently skip duplicate TestFiles.add() instead of raising#2144aseembits93 merged 8 commits intomainfrom
aseembits93 merged 8 commits intomainfrom
Conversation
This was referenced May 7, 2026
KRRT7
force-pushed
the
fix/test-files-silent-dedup
branch
from
46f63d2 to
72b75d6
Compare
KRRT7
force-pushed
the
cf-fix-concurrency-measurement
branch
from
51fb98a to
9399561
Compare
KRRT7
force-pushed
the
fix/test-files-silent-dedup
branch
from
72b75d6 to
5208de8
Compare
The previous implementation used `if test_file not in self.test_files` which performs O(n) equality comparison across all TestFile objects. Replace with a `_seen_paths` set keyed on `instrumented_behavior_file_path` for O(1) deduplication lookups.
Java test discovery adds the same instrumented_behavior_file_path with different fields. The old Pydantic __eq__ treated them as different; the set-based dedup correctly identifies them as the same test file. Since get_test_type_by_instrumented_file_path() returns on first path match anyway, duplicates by path are dead weight. Silent skip (first write wins) is both correct and O(1).
- Remove .as_posix() from --junitxml and --rcfile subprocess arguments; use native path format (str(path)) which works on all platforms. - Replace get_cross_platform_subprocess_run_args() in execute_test_subprocess with plain subprocess.run(capture_output=True, close_fds=False). The Windows-specific CREATE_NEW_PROCESS_GROUP and stdin=DEVNULL flags caused pytest to sporadically fail to write its junitxml output file, resulting in flaky test_pickle_patcher failures. Matches the working implementation in codeflash-python. - Fix 10 pre-existing mypy errors in support.py (bare generics, re-export issues, missing attribute guards).
When the test subprocess exits non-zero and produces no JUnit XML, log the return code and stdout/stderr at WARNING level so the root cause is visible in CI logs. Previously this was a generic "No test results found" message that made Windows CI flakes impossible to diagnose. Also fixes pre-existing mypy strict errors in parse_xml.py: - Add return type to _parse_func - Type CompletedProcess[str] (subprocess uses text=True) - Parameterize generic types (tuple, re.Match) - Remove dead .decode() branches (stdout is already str)
KRRT7
force-pushed
the
fix/test-files-silent-dedup
branch
from
c58d349 to
50f5070
Compare
fix: use native path format for pytest subprocess args on Windows
aseembits93
reviewed
May 8, 2026
| with custom_addopts(): | ||
| run_args = get_cross_platform_subprocess_run_args( | ||
| cwd=cwd, env=env, timeout=timeout, check=False, text=True, capture_output=True | ||
| return subprocess.run( |
aseembits93
reviewed
May 8, 2026
|
|
||
| class TestFiles(BaseModel): | ||
| test_files: list[TestFile] | ||
| _seen_paths: set[Path] = PrivateAttr(default_factory=set) |
aseembits93
previously approved these changes
May 8, 2026
fix: surface subprocess errors when pytest XML is missing
aseembits93
approved these changes
May 8, 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
Changes
TestFiles.add()duplicate handling from raisingValueErrorto silently skipping (first-write-wins). Fixes Java e2e test failures where test discovery adds the sameinstrumented_behavior_file_pathwith different metadata fields.Stack
Part of #2132 linear stack. Replaces merged-by-accident #2133.
Changes
codeflash/models/models.py: Use set for O(1) membership check, silently return on duplicate pathtests/test_test_files_add.py: Tests for dedup behaviorTest plan
test_add_duplicate_is_noopverifies silent skiptest_add_unique_test_filepassestest_add_many_files_performancepasses