Windows support

.github/workflows/ci.yml

@@ -38,31 +38,39 @@ jobs:
         shell: bash -el {0}
     strategy:
       matrix:
-        os: [macos-14, ubuntu-latest]
+        os: [macos-14, ubuntu-latest, windows-latest]
         python-version: ["3.11", "3.12", "3.13"]
         exclude:
-          # Just run macos tests on one Python version
+          # Just run macos and windows tests on one Python version
           - os: macos-14
             python-version: "3.12"
           - os: macos-14
             python-version: "3.13"
+          - os: windows-latest
+            python-version: "3.12"
+          - os: windows-latest
+            python-version: "3.13"
     steps:
       - uses: actions/checkout@v4.2.2
         with:
           fetch-depth: 0
 
       - name: Set up Miniconda with Python ${{ matrix.python-version }}
+        if: runner.os != 'Windows'
         uses: conda-incubator/setup-miniconda@v3
         with:
           auto-update-conda: true
           python-version: ${{ matrix.python-version }}
           channels: conda-forge,bioconda
 
       - name: Install conda test tools
+        if: runner.os != 'Windows'
         run: conda install bcftools plink
 
       - name: Install uv
         uses: astral-sh/setup-uv@v6
+        with:
+          python-version: ${{ matrix.python-version }}
 
       - name: Install test dependencies
         run: |

pyproject.toml

@@ -31,6 +31,7 @@ classifiers = [
   "Operating System :: POSIX :: Linux",
   "Operating System :: MacOS",
   "Operating System :: MacOS :: MacOS X",
+  "Operating System :: Microsoft :: Windows",
   "Intended Audience :: Science/Research",
   "Programming Language :: Python",
   "Programming Language :: Python :: 3",
@@ -50,7 +51,7 @@ vcztools = "vcztools.cli:vcztools_main"
 [dependency-groups]
 test = [
   "bio2zarr",
-  "cyvcf2",
+  "cyvcf2 ; sys_platform != 'win32'",
   "obstore",
   "pytest",
   "pytest-cov",

tests/test_bcftools_validation.py

@@ -1,5 +1,6 @@
 import pathlib
 import subprocess
+import sys
 
 import click.testing as ct
 import pytest
@@ -8,6 +9,10 @@
 
 from .utils import assert_vcfs_close, vcz_path_cache
 
+pytestmark = pytest.mark.skipif(
+    sys.platform == "win32", reason="Not supported on Windows"
+)
+
 
 def run_bcftools(args: str, expect_error=False) -> tuple[str, str]:
     """

tests/test_cli.py

@@ -1,4 +1,5 @@
 import pathlib
+import sys
 from unittest import mock
 
 import click.testing as ct
@@ -9,6 +10,10 @@
 from tests.utils import vcz_path_cache
 from vcztools import provenance
 
+pytestmark = pytest.mark.skipif(
+    sys.platform == "win32", reason="Not supported on Windows"
+)
+
 
 @pytest.fixture
 def vcz_path():

tests/test_cpython_interface.py

@@ -1,3 +1,5 @@
+import sys
+
 import numpy as np
 import pytest
 
@@ -78,6 +80,7 @@ def example_encoder(num_variants=1, num_samples=0, add_info=True):
     return encoder
 
 
+@pytest.mark.skipif(sys.platform == "win32", reason="Not implemented on Windows")
 class TestPrintState:
     def test_nomimal_case(self, tmp_path):
         encoder = example_encoder()
@@ -440,7 +443,18 @@ def test_array_bad_alignment(self):
 class TestUninitialised:
     @pytest.mark.parametrize(
         "name",
-        ["add_info_field", "add_gt_field", "add_format_field", "print_state", "encode"],
+        [
+            "add_info_field",
+            "add_gt_field",
+            "add_format_field",
+            pytest.param(
+                "print_state",
+                marks=pytest.mark.skipif(
+                    sys.platform == "win32", reason="Not implemented on Windows"
+                ),
+            ),
+            "encode",
+        ],
     )
     def test_methods(self, name):
         cls = _vcztools.VcfEncoder

tests/test_filter.py

@@ -1,4 +1,5 @@
 import pathlib
+import sys
 
 import numpy as np
 import numpy.testing as nt
@@ -9,6 +10,10 @@
 from tests.utils import vcz_path_cache
 from vcztools import filter as filter_mod
 
+pytestmark = pytest.mark.skipif(
+    sys.platform == "win32", reason="Not supported on Windows"
+)
+
 
 class TestFilterExpressionParser:
     @pytest.fixture

tests/test_plink_validation.py

@@ -1,6 +1,7 @@
 import os
 import pathlib
 import subprocess
+import sys
 
 import click.testing as ct
 import pytest
@@ -9,6 +10,10 @@
 
 from . import utils
 
+pytestmark = pytest.mark.skipif(
+    sys.platform == "win32", reason="Not supported on Windows"
+)
+
 
 def assert_files_identical(path1, path2):
     """

tests/test_query.py

@@ -1,5 +1,6 @@
 import pathlib
 import re
+import sys
 from io import StringIO
 
 import numpy as np
@@ -17,6 +18,10 @@
 )
 from vcztools.retrieval import variant_chunk_iter
 
+pytestmark = pytest.mark.skipif(
+    sys.platform == "win32", reason="Not supported on Windows"
+)
+
 
 def test_list_samples(tmp_path):
     vcf_path = pathlib.Path("tests/data/vcf") / "sample.vcf.gz"

tests/test_retrieval.py

@@ -1,4 +1,5 @@
 import pathlib
+import sys
 
 import numpy.testing as nt
 import pytest
@@ -9,6 +10,10 @@
 
 from .utils import vcz_path_cache
 
+pytestmark = pytest.mark.skipif(
+    sys.platform == "win32", reason="Not supported on Windows"
+)
+
 
 def test_variant_chunk_iter():
     original = pathlib.Path("tests/data/vcf") / "sample.vcf.gz"

tests/test_stats.py

@@ -1,4 +1,5 @@
 import pathlib
+import sys
 from io import StringIO
 
 import pytest
@@ -9,6 +10,10 @@
 
 from .utils import vcz_path_cache
 
+pytestmark = pytest.mark.skipif(
+    sys.platform == "win32", reason="Not supported on Windows"
+)
+
 
 def test_nrecords():
     original = pathlib.Path("tests/data/vcf") / "sample.vcf.gz"

tests/test_tskit_data.py

@@ -3,6 +3,8 @@
 with various outputs.
 """
 
+import sys
+
 import bio2zarr.plink as p2z
 import bio2zarr.tskit as ts2z
 import bio2zarr.vcf as v2z
@@ -205,6 +207,7 @@ def fx_simple_ts(tmp_path):
 # handled)
 
 
+@pytest.mark.skipif(sys.platform == "win32", reason="Not supported on Windows")
 class TestVcfRoundTrip:
     def assert_bio2zarr_rt(self, tmp_path, tskit_vcz):
         vcf_path = tmp_path / "out.vcf"

tests/test_vcf_roundtrip.py

@@ -1,4 +1,5 @@
 import pathlib
+import sys
 
 import pytest
 
@@ -7,6 +8,10 @@
 
 from .utils import assert_vcfs_close
 
+pytestmark = pytest.mark.skipif(
+    sys.platform == "win32", reason="Not supported on Windows"
+)
+
 
 @pytest.mark.parametrize(
     "vcf_file",

tests/test_vcf_writer.py

@@ -7,14 +7,16 @@
 import pytest
 import zarr
 from bio2zarr import vcf
-from cyvcf2 import VCF
 from numpy.testing import assert_array_equal
 
 from vcztools.constants import INT_FILL, INT_MISSING
 from vcztools.vcf_writer import _compute_info_fields, c_chunk_to_vcf, write_vcf
 
 from .utils import assert_vcfs_close, to_vcz_icechunk, vcz_path_cache
 
+cyvcf2 = pytest.importorskip("cyvcf2")
+VCF = cyvcf2.VCF
+
 
 @pytest.mark.parametrize("output_is_path", [True, False])
 @pytest.mark.parametrize("zarr_backend_storage", [None, "fsspec", "obstore"])

tests/utils.py

@@ -5,7 +5,6 @@
 from pathlib import Path
 from typing import Any
 
-import cyvcf2
 import numpy as np
 import xarray as xr
 import zarr
@@ -30,8 +29,10 @@ def load_dataset(
 
 
 @contextmanager
-def open_vcf(path) -> Iterator[cyvcf2.VCF]:
+def open_vcf(path) -> Iterator:
     """A context manager for opening a VCF file."""
+    import cyvcf2  # noqa: PLC0415
+
     vcf = cyvcf2.VCF(path)
     try:
         yield vcf

vcztools/_vcztoolsmodule.c

@@ -33,6 +33,13 @@ handle_library_error(int err)
     }
 }
 
+/* The dup/fdopen/fclose pattern used by make_file does not work on Windows
+ * because _dup shares the underlying OS handle rather than duplicating it.
+ * When fclose closes the duped FILE*, it invalidates the original handle,
+ * causing an access violation when Python later closes its file object.
+ * Since print_state is only a debugging tool, we simply exclude it on Windows. */
+#ifndef _WIN32
+
 static FILE *
 make_file(PyObject *fileobj, const char *mode)
 {
@@ -60,6 +67,8 @@ make_file(PyObject *fileobj, const char *mode)
     return ret;
 }
 
+#endif /* !_WIN32 */
+
 /*===================================================================
  * VcfEncoder
  *===================================================================
@@ -474,6 +483,7 @@ VcfEncoder_encode(VcfEncoder *self, PyObject *args)
     return ret;
 }
 
+#ifndef _WIN32
 static PyObject *
 VcfEncoder_print_state(VcfEncoder *self, PyObject *args)
 {
@@ -500,6 +510,8 @@ VcfEncoder_print_state(VcfEncoder *self, PyObject *args)
     return ret;
 }
 
+#endif /* !_WIN32 */
+
 /* Return a copy of the dictionary of arrays providing the memory backing.
  * Note that we return copy of the Dictionary here so that the arrays themselves
  * can't be removed from it.
@@ -524,10 +536,12 @@ static PyGetSetDef VcfEncoder_getsetters[] = {
 };
 
 static PyMethodDef VcfEncoder_methods[] = {
+#ifndef _WIN32
     { .ml_name = "print_state",
         .ml_meth = (PyCFunction) VcfEncoder_print_state,
         .ml_flags = METH_VARARGS,
         .ml_doc = "Debug method to print out the low-level state" },
+#endif
     { .ml_name = "add_info_field",
         .ml_meth = (PyCFunction) VcfEncoder_add_info_field,
         .ml_flags = METH_VARARGS,