Port vcfaz to Python 3 and remove legacy py27 directory

[Copilot]

Completes the Python 3 (≥3.10) port of the vcfaz package, migrating from the legacy Python 2.7 conda package layout to a src/ layout, and removes the legacy directory entirely.

Python 3 compatibility fixes

src/vcfaz/model.py

• class _AltRecord(object, metaclass=ABCMeta) replaces __metaclass__ = ABCMeta
• basestring → str
• Removed __cmp__ (Python 2 only)
• Dropped legacy try/except ImportError fallbacks for Counter/OrderedDict

src/vcfaz/parser.py

• All intra-package imports converted to explicit relative imports (from .model import ...)
• itertools.izip → zip
• dict.has_key() → in operator
• dict.iteritems()/iterkeys()/itervalues() → .items()/.keys()/.values()
• def next(self) renamed to def __next__(self) with next = __next__ alias for backwards compat
• Gzip handling simplified — always wraps with codecs.getreader on Python 3
• Invalid regex escape sequences fixed to raw strings

src/vcfaz/sample_filter.py

• Relative import: from .parser import Reader, Writer
• xrange → range

src/vcfaz/utils.py

• readers[i].next() → next(readers[i])

src/vcfaz/__init__.py

• All imports converted to explicit relative imports; exposes Reader, Writer, VCFReader, VCFWriter, Filter, SampleFilter, RESERVED_INFO, RESERVED_FORMAT

Removed

• pyvcfaz-0.6.8-py27_0/ — entire legacy Python 2.7 directory deleted

Tests

Added tests/test_import.py with smoke tests covering imports and basic in-memory VCF parsing:

# Quick non-installed smoke test
PYTHONPATH=src python -c "import vcfaz; print(vcfaz.__file__)"

# Or run the full smoke test suite
PYTHONPATH=src python -m pytest tests/test_import.py -v

Original prompt

Goal

Create a PR in biofold/Py3VCF that completes the Python 3 (>=3.10) port of the legacy vcfaz package and removes the legacy Python 2 directory.

Background / Context

The repository currently contains a legacy, Python 2.7-era copy of the vcfaz package under:

• pyvcfaz-0.6.8-py27_0/lib/python2.7/site-packages/vcfaz/

We want to:

1. Port the package to a modern Python 3 codebase using a src/ layout.
2. Place the Python 3 port under:

• src/vcfaz/

3. Delete the legacy directory:

• pyvcfaz-0.6.8-py27_0/

Requirements

A. Add Python 3 package sources

Create the following files under src/vcfaz/ with Python 3-compatible code:

• src/vcfaz/__init__.py
• src/vcfaz/parser.py
• src/vcfaz/model.py
• src/vcfaz/utils.py
• src/vcfaz/filters.py
• src/vcfaz/sample_filter.py

Key compatibility fixes include (non-exhaustive):

• Convert legacy intra-package imports to explicit relative imports (e.g. from .model import ...).
• Replace Python 2-only APIs:
  • itertools.izip -> zip
  • dict.iteritems()/iterkeys()/itervalues() -> .items()/.keys()/.values()
  • xrange -> range
  • basestring -> str
  • has_key -> in
  • implement iterator protocol (__iter__ + __next__) for Reader.
• Ensure gzipped VCF handling works correctly on Python 3 (decode bytes to text via codecs reader).
• Maintain backwards-compatible aliases where present (e.g. VCFReader, VCFWriter) and export public API from src/vcfaz/__init__.py.

B. Remove legacy directory

Delete the entire directory:

• pyvcfaz-0.6.8-py27_0/

C. Smoke tests / verification

Add minimal smoke tests (can be lightweight and not require external data) or at least include instructions in PR description. Preferably add tests/test_import.py that verifies:

• import vcfaz works
• from vcfaz.parser import Reader works

If tests are added, ensure they pass in CI or at least locally.

D. PR details

• Base branch: main
• Create a feature branch (suggested name: python3-port if available; otherwise python3-port-2).
• PR title suggestion: Port vcfaz to Python 3 and remove legacy py27 directory
• PR should include a clear summary and how to run a quick non-installed smoke test:
  • PYTHONPATH=src python -c "import vcfaz; print(vcfaz.__file__)"

Reference legacy sources

Legacy files to port (current commit context):

• pyvcfaz-0.6.8-py27_0/lib/python2.7/site-packages/vcfaz/__init__.py
• pyvcfaz-0.6.8-py27_0/lib/python2.7/site-packages/vcfaz/parser.py
• pyvcfaz-0.6.8-py27_0/lib/python2.7/site-packages/vcfaz/model.py
• pyvcfaz-0.6.8-py27_0/lib/python2.7/site-packages/vcfaz/utils.py
• pyvcfaz-0.6.8-py27_0/lib/python2.7/site-packages/vcfaz/filters.py
• pyvcfaz-0.6.8-py27_0/lib/python2.7/site-packages/vcfaz/sample_filter.py

Notes

• Keep changes focused on the vcfaz package port and removal of the legacy directory.
• If other packaging files (e.g. pyproject.toml) are required for tests/imports, add them as needed, but do not over-scope.

This pull request was created from Copilot chat.

---

🔒 GitHub Advanced Security automatically protects Copilot coding agent pull requests. You can protect all pull requests by enabling Advanced Security for your repositories. Learn more about Advanced Security.

[ecapriotti]

accepted