refactor: update plugin to follow Python best practices#9
refactor: update plugin to follow Python best practices#9
Conversation
- Modernize pyproject.toml with full project configuration - Add build-system specification - Configure ruff for linting and formatting - Add mypy configuration - Bump Python requirement to >=3.9 - Add pre-commit configuration - Include ruff-format and ruff-lint hooks - Add prettier for markdown/yaml formatting - Include standard hooks (trailing whitespace, merge conflict, etc.) - Add GitHub Actions linting workflow - Separate lint.yaml for ruff and prettier checks - Update linux.yaml with modern action versions and Python 3.9-3.12 - Refactor Python code - Add comprehensive type hints throughout - Use f-strings consistently - Replace deprecated pkg_resources with importlib.metadata - Remove OrderedDict (use dict in Python 3.7+) - Use contextlib.suppress instead of try-except-pass - Improve docstrings and code organization - Add CLAUDE.md for AI assistant guidance - Simplify setup.py to minimal pyproject.toml wrapper
Replace separate ruff and prettier jobs with the official pre-commit action. This ensures CI matches local pre-commit results exactly.
Remove redundant options that match defaults: - target-version (inferred from requires-python) - format options (all defaults) - isort section (not needed) - excessive lint rules Keep only essential settings: - line-length = 120 - select core rules (E, F, W, I, UP, SIM) - ignore E501 (handled by formatter)
Remove E501 ignore rule - fix the actual long lines instead of ignoring them.
- Fix end-of-file newlines in XML files - Fix trailing whitespace in HTML reports - Normalize line endings (CRLF to LF) - Apply prettier formatting to CLAUDE.md
- Remove setup.py (pyproject.toml handles everything) - Add publish.yaml workflow triggered on GitHub releases - Uses trusted publishing with OIDC (no API tokens needed)
Revert changes to test data XML files and add exclude patterns to prevent pre-commit hooks from modifying test data and example files.
Revert docs/example HTML files and remove html from prettier types.
- Use --strict mode to catch issues - Verify HTML report is generated for each test - Check that SAV module appears in the log file - Output to separate directories per sequencer type
|
@matthdsm - I made a start on modernising the codebase of this plugin a little. However, the CI jobs show that it's not finding the test data. Any help would be much appreciated! Was it definitely working on master? Not sure if I broke it here or if it was already broken. |
|
Hi @ewels, I've also already done some work (and added new test data) in the dev branch. It's been quite a while, so I don't remember everything very well. As I said in #6 and mentioned in #5, picking up the input files is quite flaky... |
|
That's the second time in a few weeks that I've done that, not checking to see if there's a I'll see if there's anything to salvage on this PR, but will probably just close it out. |
|
No worries 😅 chances are your PR has more usable stuff than the dev branch |
|
Closing in favour of #10 |
3 participants
Summary
Modernizes the MultiQC_SAV plugin to follow current Python best practices, aligned with the xenium-extra MultiQC plugin and MultiQC module guidelines.
Changes
Project Configuration
pyproject.tomlwith full project metadata, dependencies, and entry pointssetup.py(no longer needed with modern packaging)Pre-commit & Linting
.pre-commit-config.yamlwith ruff, prettier, and standard hooks.prettierrc.jsand.prettierignorefor consistent formattingtest_data/anddocs/example/from file-modifying hooksGitHub Actions
lint.yamlworkflow using pre-commit actionpublish.yamlworkflow for PyPI releases (trusted publishing)linux.yamlwith modern action versions and Python 3.9-3.12 matrixCode Refactoring
pkg_resourceswithimportlib.metadataOrderedDict(use dict in Python 3.7+)contextlib.suppressinstead of try-except-passDocumentation
CLAUDE.mdfor AI assistant guidanceTest plan
pre-commit run --all-filesruff check . && ruff format --check .pip install -e . && multiqc -m SAV test_data/MiSeq