Always regenerate projects index so variant-flag toggles refresh links

[cboos]

Summary

• Drops the any_cache_updated gate on projects-index regeneration and the now-unused renderer.is_outdated(index_path) check.
• Index is now always regenerated from the in-memory project_summaries aggregation (one template pass + one file write).

Why

The previous gate only fired when source JSONL files changed. Toggling a variant flag like --compact / --no-timestamps / --detail between runs produces new per-project filenames (e.g. combined_transcripts.compact.md) but doesn't touch the cache, so the index kept linking to the previous run's filenames. The bug also bit --expand-paths --combined no (per-session bullet links) and the symmetric toggle-back case (going back to the bare variant when the file already exists on disk).

A first attempt tracked output writes via a new any_output_written flag — that fixed the forward toggle but left the backward toggle stale. Always regenerating closes both directions and dispenses with the per-call freshness logic; the saved work on a true no-op run is negligible vs the per-project cache work that already ran.

Test plan

• test/test_output_paths.py::TestProjectsIndexVariantRefresh — three new regression scenarios:
  • forward toggle (default → --compact)
  • --expand-paths --combined no + --compact
  • backward toggle (--compact → default, file already on disk)
• test/test_html_regeneration.py::test_projects_index_regeneration_on_jsonl_change — updated to pin the always-regenerate contract and assert the removed "skipping regeneration" log line stays gone.
• just ci clean (1791 unit + benchmark; ruff / pyright / ty all green).

Summary by CodeRabbit

• Bug Fixes

  • Projects index is now regenerated on every run so it always reflects current project summaries and variant-specific output links.

• Tests

  • Added and updated tests to ensure the index is rewritten on no-op runs, that timestamps advance, and that links refresh when output variant options toggle.

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: defaults

Review profile: CHILL

Plan: Pro

Run ID: 00bbeffc-05f0-44ab-8cab-da8d024fd0b9

📥 Commits

Reviewing files that changed from the base of the PR and between b682db2 and 16ed262.

📒 Files selected for processing (2)

• test/test_html_regeneration.py
• test/test_integration_realistic.py

✅ Files skipped from review due to trivial changes (1)

• test/test_integration_realistic.py

---

📝 Walkthrough

Walkthrough

Index generation is refactored to always run unconditionally, removing cache-update tracking that previously gated regeneration. Tests confirm the index rewrites on every run and that links refresh when output format variants toggle across consecutive invocations.

Changes

Unconditional Index Regeneration

Layer / File(s)	Summary
Index generation refactoring claude_code_log/converter.py	Removes any_cache_updated tracking variable and changes index generation from conditional (checking staleness, date filters, and cache state) to unconditional. Index rendering now conditionally enables expand_paths_tree only for HTML/Markdown formats and always writes with errors="replace".
Index regeneration test validation test/test_html_regeneration.py, test/test_integration_realistic.py	Updates test_projects_index_regeneration_on_jsonl_change to assert the index regenerates on every run (no-op reruns included) and that JSONL edits appear in the refreshed output. Switches modification-time checks from second to nanosecond precision for more reliable timing assertions; updates an integration test docstring to the always-regenerate wording.

Variant Refresh Regression Tests

Layer / File(s)	Summary
Test setup and helper test/test_output_paths.py	Imports process_projects_hierarchy and introduces _build_one_project_projects_dir() helper to create minimal on-disk projects structures for variant-refresh test cases.
Variant refresh test cases test/test_output_paths.py	Adds TestProjectsIndexVariantRefresh class with three test cases verifying that toggling --compact flag, combining --combined no with expand_paths, and reverting to defaults all trigger index regeneration with links pointing to the correct variant-suffixed (or unsuffixed) filenames.

Estimated code review effort

🎯 3 (Moderate) | ⏱️ ~20 minutes

Poem

🐰 I nibble bytes beneath the moon,
I hop where indexes renew,
Flags flip, links twine and change their name,
Each run I write — no skip, no shame.
Hooray for fresh and tidy view!

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title clearly and specifically describes the main change: unconditional index regeneration to fix variant-flag toggle issues that previously left stale links.
Docstring Coverage	✅ Passed	Docstring coverage is 90.00% which is sufficient. The required threshold is 80.00%.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

📝 Generate docstrings

• Create stacked PR
• Commit on current branch

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch dev/filtered-output-follow-up

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[coderabbitai]

🧹 Nitpick comments (1)

test/test_html_regeneration.py (1)

133-202: 💤 Low value

Test correctly validates always-regenerate contract.

The switch to st_mtime_ns for nanosecond precision is more reliable for timing assertions, and the test properly verifies both the no-op always-regen behavior and regeneration after JSONL changes.

Optional enhancement: verify content changes

The test could strengthen the third-run validation by confirming the new JSONL content actually appears in the regenerated index:

 post_change_mtime_ns = index_file.stat().st_mtime_ns
 process_projects_hierarchy(projects_dir)
 assert index_file.stat().st_mtime_ns > post_change_mtime_ns
+# Verify the new content appears in the index
+new_content = index_file.read_text(encoding="utf-8")
+assert "updated content for index regeneration test" in new_content

This would fully validate the docstring's claim that "JSONL edits are picked up."

🤖 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 `@test/test_html_regeneration.py` around lines 133 - 202, The test verifies
always-regenerate and JSONL-change detection but doesn't assert that the new
JSONL line actually appears in the regenerated index; after calling
process_projects_hierarchy(projects_dir) in the third run, read index_file (from
variable index_file) and assert the new_message (or a distinctive substring like
the summary text) is present to prove the content was incorporated; update
test_projects_index_regeneration_on_jsonl_change to open and read
index_file.read_text(encoding="utf-8") after the final
process_projects_hierarchy call and assert the expected JSON snippet or summary
string is in the content.

🤖 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.

Nitpick comments:
In `@test/test_html_regeneration.py`:
- Around line 133-202: The test verifies always-regenerate and JSONL-change
detection but doesn't assert that the new JSONL line actually appears in the
regenerated index; after calling process_projects_hierarchy(projects_dir) in the
third run, read index_file (from variable index_file) and assert the new_message
(or a distinctive substring like the summary text) is present to prove the
content was incorporated; update
test_projects_index_regeneration_on_jsonl_change to open and read
index_file.read_text(encoding="utf-8") after the final
process_projects_hierarchy call and assert the expected JSON snippet or summary
string is in the content.

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: defaults

Review profile: CHILL

Plan: Pro

Run ID: e8a9e2d5-67af-4a36-9ce1-257022597d91

📥 Commits

Reviewing files that changed from the base of the PR and between 68e5bfb and b682db2.

📒 Files selected for processing (3)

• claude_code_log/converter.py
• test/test_html_regeneration.py
• test/test_output_paths.py