Keep non-FSF copyrights in GPL license text files

[JustinWonjaePark]

• Bug Fixes
  • Improved GPL license-text parsing by excluding only copyright entries associated with the Free Software Foundation.
  • Preserved other valid copyright information instead of removing all copyright entries.

[coderabbitai]

📝 Walkthrough

Walkthrough

The change adds a shared helper that removes copyright entries containing “free software foundation” from GPL-family license text. Both ScanCode parser paths use the helper and retain other copyright entries. Tests cover GPL-family and non-applicable inputs.

Changes

GPL copyright filtering

Layer / File(s)	Summary
Copyright filter and parser integration src/fosslight_source/_parsing_scancode_file_item.py, tests/test_tox.py	The code uses a compiled case-insensitive GPL pattern and shared FSF marker. Both parser paths apply selective filtering. Tests cover GPL-family, mixed, non-GPL, non-license-text, and empty-license inputs.

Estimated code review effort: 2 (Simple) | ~10 minutes

Suggested labels: bug fix

Suggested reviewers: soimkim

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.
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.
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title clearly and concisely describes the main change: preserving non-FSF copyrights in GPL license text files.

✨ 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 fix/fsf_copyright_remove

---

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.}

[coderabbitai]

Actionable comments posted: 1

🧹 Nitpick comments (1)

src/fosslight_source/_parsing_scancode_file_item.py (1)

233-235: 🗄️ Data Integrity & Integration | 🔵 Trivial | ⚡ Quick win

Add regression coverage for both parser paths.

The supplied test at tests/test_tox.py:113-137 verifies only the GPL-family gate. Add cases for mixed-case Free Software Foundation text and another copyright entry. Assert that both parsing_scancode_32_earlier and parsing_scancode_32_later remove only the FSF entry.

Also applies to: 730-732

🤖 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 `@src/fosslight_source/_parsing_scancode_file_item.py` around lines 233 - 235,
Add regression tests covering both parsing_scancode_32_earlier and
parsing_scancode_32_later with mixed-case “Free Software Foundation” text plus
another copyright entry; assert each parser removes only the FSF entry while
preserving the other entry, without changing the existing GPL-family coverage.

🤖 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 `@src/fosslight_source/_parsing_scancode_file_item.py`:
- Around line 63-69: In exclude_free_software_foundation_copyrights, rename the
list-comprehension variable copyright to a non-shadowing name such as
copyright_entry and update its references within the comprehension, preserving
the existing filtering behavior.

---

Nitpick comments:
In `@src/fosslight_source/_parsing_scancode_file_item.py`:
- Around line 233-235: Add regression tests covering both
parsing_scancode_32_earlier and parsing_scancode_32_later with mixed-case “Free
Software Foundation” text plus another copyright entry; assert each parser
removes only the FSF entry while preserving the other entry, without changing
the existing GPL-family coverage.

🪄 Autofix

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: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Pro Plus

Run ID: df5652c6-45ea-41ae-b6b8-436d62bd049c

📥 Commits

Reviewing files that changed from the base of the PR and between e4c9c3a and 8149292.

📒 Files selected for processing (1)

• src/fosslight_source/_parsing_scancode_file_item.py

[coderabbitai]

Actionable comments posted: 1

🤖 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 `@src/fosslight_source/_parsing_scancode_file_item.py`:
- Around line 216-218: Update the legacy parser’s license-item aggregation
around result_item.is_license_text so it preserves whether any matched lic_item
is license text instead of overwriting the flag per item. Use this aggregate
value both for result_item.is_license_text and the
filter_fsf_copyright_from_gpl_license_text call, and add a regression test
covering multiple license items in differing orders.

🪄 Autofix

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: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Pro Plus

Run ID: cd6b2a76-a679-4b32-b2c4-bea5282f5dfa

📥 Commits

Reviewing files that changed from the base of the PR and between 8149292 and cd30983.

📒 Files selected for processing (2)

• src/fosslight_source/_parsing_scancode_file_item.py
• tests/test_tox.py