test(jsonl): add jsonl_parser schema coverage and harden null usage#22
Open
clean6378-max-it wants to merge 1 commit intomasterfrom
Open
test(jsonl): add jsonl_parser schema coverage and harden null usage#22clean6378-max-it wants to merge 1 commit intomasterfrom
clean6378-max-it wants to merge 1 commit intomasterfrom
Conversation
Add tests/test_jsonl_parser.py covering _parse_tool_result dispatch, content helpers, user/assistant/system paths, file activity, parse_session integration, quick_session_info, and malformed entries. Treat non-dict message.usage in _process_assistant as empty to avoid crashes on null usage.
📝 WalkthroughWalkthroughA comprehensive test suite for the JSONL parser module is added with coverage of tool result parsing, content handling, message processing, file tracking, and session parsing. A defensive guard is added to ensure usage data is always a dict in assistant message processing. ChangesParser Test Coverage & Defensive Guard
Estimated code review effort🎯 3 (Moderate) | ⏱️ ~25 minutes Possibly related issues
Poem
🚥 Pre-merge checks | ✅ 5✅ Passed checks (5 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. ✨ Finishing Touches📝 Generate docstrings
🧪 Generate unit tests (beta)
Comment |
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 1
Caution
Some comments are outside the diff and can’t be posted inline due to platform limitations.
⚠️ Outside diff range comments (1)
utils/jsonl_parser.py (1)
140-142:⚠️ Potential issue | 🟠 Major | ⚡ Quick winGuard
messagebefore dereferencing in both user/assistant processors.If an entry contains
"message": null, both processors call.get(...)onNoneand abortparse_sessionfor a recoverable malformed line. Apply the same dict-guard pattern here too.💡 Suggested fix
def _process_user(entry: dict, messages: list, metadata: dict): @@ - msg = entry.get("message", {}) + msg = entry.get("message", {}) + if not isinstance(msg, dict): + msg = {} content = msg.get("content", []) @@ def _process_assistant(entry: dict, messages: list, metadata: dict): @@ - msg = entry.get("message", {}) + msg = entry.get("message", {}) + if not isinstance(msg, dict): + msg = {} model = msg.get("model", "")Also applies to: 173-174
🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the rest with a brief reason, keep changes minimal, and validate. In `@utils/jsonl_parser.py` around lines 140 - 142, The processors in utils/jsonl_parser.py currently assume entry["message"] is a dict and call entry.get("message", {}) without guarding against message being None; update both user and assistant processing blocks to first coerce message to a dict (e.g., msg = entry.get("message") or {} ) before calling msg.get("content", []) and passing to _extract_text, and apply the same pattern where similar code exists (the other processor and the parse_session usage) so a null "message" no longer raises when dereferenced.
🤖 Prompt for all review comments with AI agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
Inline comments:
In `@tests/test_jsonl_parser.py`:
- Around line 718-730: Add two regression tests in tests/test_jsonl_parser.py
alongside test_missing_usage_dict_no_crash: one that passes a record with
"message": null and another that passes a record with a non-dict usage (e.g.,
"usage": "invalid" or a number) to ensure parse_session handles these malformed
shapes without crashing and still sets metadata.total_input_tokens to 0;
reference the existing test function name test_missing_usage_dict_no_crash and
the parser entrypoint parse_session to locate where to add the new tests and
mirror the try/finally cleanup pattern used there.
---
Outside diff comments:
In `@utils/jsonl_parser.py`:
- Around line 140-142: The processors in utils/jsonl_parser.py currently assume
entry["message"] is a dict and call entry.get("message", {}) without guarding
against message being None; update both user and assistant processing blocks to
first coerce message to a dict (e.g., msg = entry.get("message") or {} ) before
calling msg.get("content", []) and passing to _extract_text, and apply the same
pattern where similar code exists (the other processor and the parse_session
usage) so a null "message" no longer raises when dereferenced.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: defaults
Review profile: CHILL
Plan: Pro
Run ID: 4b102107-1dd1-4277-936f-9231732d0eb4
📒 Files selected for processing (2)
tests/test_jsonl_parser.pyutils/jsonl_parser.py
Comment on lines
+718
to
+730
| def test_missing_usage_dict_no_crash(self): | ||
| path = _write_jsonl([ | ||
| { | ||
| "type": "assistant", | ||
| "timestamp": "2026-01-01T00:00:00Z", | ||
| "message": {"model": "m", "content": [], "usage": None}, | ||
| }, | ||
| ]) | ||
| try: | ||
| s = parse_session(path) | ||
| assert s["metadata"]["total_input_tokens"] == 0 | ||
| finally: | ||
| os.unlink(path) |
There was a problem hiding this comment.
Add explicit message: null and non-dict usage regression tests.
Coverage currently locks in usage=None and missing message, but not "message": null or non-null non-dict usage values. Those malformed shapes are common in JSONL and should be pinned.
🧪 Suggested test additions
class TestParseSession:
@@
def test_missing_usage_dict_no_crash(self):
@@
s = parse_session(path)
assert s["metadata"]["total_input_tokens"] == 0
finally:
os.unlink(path)
+
+ `@pytest.mark.parametrize`("bad_usage", ["oops", [], 123])
+ def test_non_dict_usage_no_crash(self, bad_usage):
+ path = _write_jsonl([
+ {
+ "type": "assistant",
+ "timestamp": "2026-01-01T00:00:00Z",
+ "message": {"model": "m", "content": [], "usage": bad_usage},
+ },
+ ])
+ try:
+ s = parse_session(path)
+ assert s["metadata"]["total_input_tokens"] == 0
+ finally:
+ os.unlink(path)
@@
class TestMalformedPartialEntries:
@@
def test_assistant_missing_message_key(self):
@@
assert s["messages"][0]["role"] == "assistant"
finally:
os.unlink(path)
+
+ def test_assistant_message_null_no_crash(self):
+ path = _write_jsonl([
+ {"type": "assistant", "timestamp": "2026-01-01T00:00:00Z", "message": None},
+ ])
+ try:
+ s = parse_session(path)
+ assert len(s["messages"]) == 1
+ assert s["messages"][0]["role"] == "assistant"
+ finally:
+ os.unlink(path)Also applies to: 797-807
🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
In `@tests/test_jsonl_parser.py` around lines 718 - 730, Add two regression tests
in tests/test_jsonl_parser.py alongside test_missing_usage_dict_no_crash: one
that passes a record with "message": null and another that passes a record with
a non-dict usage (e.g., "usage": "invalid" or a number) to ensure parse_session
handles these malformed shapes without crashing and still sets
metadata.total_input_tokens to 0; reference the existing test function name
test_missing_usage_dict_no_crash and the parser entrypoint parse_session to
locate where to add the new tests and mirror the try/finally cleanup pattern
used there.
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
1 participant
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.
Closes #21
Summary
Adds comprehensive pytest coverage for
utils/jsonl_parser.py(CCC1) and hardens assistant parsing whenmessage.usageis JSONnullor otherwise not a plain object.Motivation
jsonl_parser.pyis the structural core for untrusted, schema-evolving Claude Code JSONL. Direct tests lock in_parse_tool_resultdispatch, helper behavior, and integration paths so silent regressions from schema drift are caught early.Changes
tests/test_jsonl_parser.pycovering:_parse_tool_resultfor all major result shapes and fallbacks_normalize_content,_extract_text,_extract_images_infer_title,_strip_system_tags_process_user,_process_assistant,_process_system,_track_file_activityparse_sessionintegration (empty files, unknown types, sidechains, snapshots, wall time, malformed lines)quick_session_infofor small vs large sessions_process_assistanttreats non-dictusageas{}sousage: nulldoes not crash.Testing