Windows support (#951)

* supporting Windows and Python 3.12 * Fix FileNotFoundErrors to support windows * Use dirname over rsplit * Fix temp file opening issues * Fix other temp file issues * Fix linting and put unlink in tearDown * Testing workflow failure of python3.7 on windows * Remove test line for python3.7 * Utilize setUp and tearDown for removing tmp files * Update changelog * Update ChangeLog.md --------- Co-authored-by: qiyunzhu <[email protected]> Co-authored-by: Daniel McDonald <[email protected]>
biocore · Apr 15, 2024 · 738d240 · 738d240
1 parent 50b6908
commit 738d240
Show file tree

Hide file tree

Showing 13 changed files with 133 additions and 66 deletions.
diff --git a/.github/workflows/python-package-conda.yml b/.github/workflows/python-package-conda.yml
@@ -9,8 +9,8 @@ on:
     branches: [ master ]
 
 env:
-  latest_python: "3.11"
-  supported_pythons: '["3.7", "3.8", "3.9", "3.10", "3.11"]'
+  latest_python: "3.12"
+  supported_pythons: '["3.7", "3.8", "3.9", "3.10", "3.11", "3.12"]'
   miniforge_version: "22.9.0-2"
   miniforge_variant: "Mambaforge"
 
@@ -34,7 +34,7 @@ jobs:
     needs: conf
     runs-on: "ubuntu-latest"
     steps:
-      - uses: actions/checkout@v3
+      - uses: actions/checkout@v4
       - uses: conda-incubator/setup-miniconda@v2
         with:
           auto-update-conda: true
@@ -56,7 +56,7 @@ jobs:
     needs: ["conf", "lint"]
     runs-on: "ubuntu-latest"
     steps:
-      - uses: actions/checkout@v3
+      - uses: actions/checkout@v4
       - uses: conda-incubator/setup-miniconda@v2
         with:
           auto-update-conda: true
@@ -81,11 +81,11 @@ jobs:
     strategy:
       fail-fast: true
       matrix:
-        os: ["ubuntu-latest", "macos-latest"]
+        os: ["ubuntu-latest", "macos-latest", "windows-latest"]
         python_version: ${{ fromJSON(needs.conf.outputs.supported_pythons) }}
         use_conda: [true, false]
     steps:
-      - uses: actions/checkout@v3
+      - uses: actions/checkout@v4
       - uses: conda-incubator/setup-miniconda@v2
         with:
           auto-update-conda: true
@@ -115,7 +115,7 @@ jobs:
     needs: ["conf", "lint", "doc", "test-all"]
     runs-on: ubuntu-latest
     steps:
-      - uses: actions/checkout@v3
+      - uses: actions/checkout@v4
         # setup-buildx-action uses the git context directly
         # but checklist wants the .git directory
       - name: Set up QEMU

diff --git a/.github/workflows/release.yml b/.github/workflows/release.yml
@@ -7,15 +7,15 @@ jobs:
     name: Build sdist
     runs-on: ubuntu-latest
     steps:
-      - uses: actions/checkout@v3
+      - uses: actions/checkout@v4
 
       - name: Build distribution
         run: |
           export RELEASE_VERSION=${{ github.ref_name }}
           pip install numpy cython
           pipx run build --sdist
 
-      - uses: actions/upload-artifact@v3
+      - uses: actions/upload-artifact@v4
         with:
           name: dist-artifacts
           path: dist/*.tar.gz
@@ -27,13 +27,13 @@ jobs:
     runs-on: ${{ matrix.os }}
     strategy:
       matrix:
-        os: [ubuntu-latest, macos-latest]
-        pyver: ["37", "38", "39", "310", "311"]
+        os: [ubuntu-latest, macos-latest, windows-latest]
+        pyver: ["37", "38", "39", "310", "311", "312"]
 
     steps:
-      - uses: actions/checkout@v3
+      - uses: actions/checkout@v4
       - name: Set up Python
-        uses: actions/setup-python@v2
+        uses: actions/setup-python@v5
         with:
           python-version: 3.9
 
@@ -47,7 +47,7 @@ jobs:
       - name: Build wheels (py ${{ matrix.pyver }}) Linux
         if: matrix.os == 'ubuntu-latest' 
         env:
-          CIBW_ARCHS_LINUX: x86_64
+          CIBW_ARCHS_LINUX: "x86_64 aarch64"
           CIBW_SKIP: "*-musllinux*"
           CIBW_BUILD: "cp${{ matrix.pyver }}-*"
 
@@ -60,10 +60,15 @@ jobs:
           CIBW_BUILD: "cp${{ matrix.pyver }}-*"
 
         uses: pypa/[email protected]
-
-
+
+      - name: Build wheels (py ${{ matrix.pyver }}) Windows
+        if: matrix.os == 'windows-latest'
+        env:
+          CIBW_ARCHS_WINDOWS: "amd64 win32 arm64"
+          CIBW_BUILD: "cp${{ matrix.pyver }}-*"
+
       - name: Upload wheels
-        uses: actions/upload-artifact@v3
+        uses: actions/upload-artifact@v4
         with:
           name: dist-artifacts
           path: ./wheelhouse/*.whl

diff --git a/ChangeLog.md b/ChangeLog.md
@@ -6,6 +6,7 @@ biom 2.1.15-dev
 
 Performance improvements:
 
+* Add Windows support. PR[#951](https://github.com/biocore/biom-format/pull/951) revises codebase to be Windows compatible and adds this support to the CI testing matrix.
 * Add NumPy 2.0 support. PR [#950](https://github.com/biocore/biom-format/pull/950) ensures code compatibility with NumPy 2.0. This support is yet to be added to the CI testing matrix.
 * Revise `Table._fast_merge` to use COO directly. For very large tables, this reduces runtime by ~50x and memory by ~5x. See PR [#913](https://github.com/biocore/biom-format/pull/933).
 * Drastically reduce the memory needs of subsampling when sums are large. Also adds 64-bit support. See PR [#935](https://github.com/biocore/biom-format/pull/935).

diff --git a/biom/tests/test_cli/test_add_metadata.py b/biom/tests/test_cli/test_add_metadata.py
@@ -9,6 +9,7 @@
 # -----------------------------------------------------------------------------
 
 import tempfile
+import os
 from unittest import TestCase, main
 
 import biom
@@ -20,13 +21,17 @@ class TestAddMetadata(TestCase):
     def setUp(self):
         """Set up data for use in unit tests."""
         self.cmd = _add_metadata
-        with tempfile.NamedTemporaryFile('w') as fh:
+        with tempfile.NamedTemporaryFile('w', delete=False) as fh:
             fh.write(biom1)
             fh.flush()
             self.biom_table1 = biom.load_table(fh.name)
+            self.temporary_fh_name = fh.name
         self.sample_md_lines1 = sample_md1.split('\n')
         self.obs_md_lines1 = obs_md1.split('\n')
 
+    def tearDown(self):
+        os.unlink(self.temporary_fh_name)
+
     def test_add_sample_metadata_no_casting(self):
         """Correctly adds sample metadata without casting it."""
         # Add a subset of sample metadata to a table that doesn't have any

diff --git a/biom/tests/test_cli/test_subset_table.py b/biom/tests/test_cli/test_subset_table.py
@@ -55,9 +55,10 @@ def test_invalid_input(self):
     def test_subset_samples_hdf5(self):
         """Correctly subsets samples in a hdf5 table"""
         cwd = os.getcwd()
-        if '/' in __file__:
-            os.chdir(__file__.rsplit('/', 1)[0])
-        obs = _subset_table(hdf5_biom='test_data/test.biom', axis='sample',
+        if os.path.sep in __file__:
+            os.chdir(os.path.dirname(__file__))
+        obs = _subset_table(hdf5_biom=os.path.join('test_data', 'test.biom'),
+                            axis='sample',
                             ids=['Sample1', 'Sample2', 'Sample3'],
                             json_table_str=None)
         os.chdir(cwd)
@@ -71,9 +72,9 @@ def test_subset_samples_hdf5(self):
     def test_subset_observations_hdf5(self):
         """Correctly subsets samples in a hdf5 table"""
         cwd = os.getcwd()
-        if '/' in __file__:
-            os.chdir(__file__.rsplit('/', 1)[0])
-        obs = _subset_table(hdf5_biom='test_data/test.biom',
+        if os.path.sep in __file__:
+            os.chdir(os.path.dirname(__file__))
+        obs = _subset_table(hdf5_biom=os.path.join('test_data', 'test.biom'),
                             axis='observation',
                             ids=['GG_OTU_1', 'GG_OTU_3', 'GG_OTU_5'],
                             json_table_str=None)

diff --git a/biom/tests/test_cli/test_summarize_table.py b/biom/tests/test_cli/test_summarize_table.py
@@ -12,16 +12,21 @@
 from biom.parse import load_table
 
 import tempfile
+import os
 from unittest import TestCase, main
 
 
 class TestSummarizeTable(TestCase):
 
     def setUp(self):
-        with tempfile.NamedTemporaryFile(mode='w') as fh:
+        with tempfile.NamedTemporaryFile(mode='w', delete=False) as fh:
             fh.write(biom1)
             fh.flush()
             self.biom1 = load_table(fh.name)
+            self.temporary_fh_name = fh.name
+
+    def tearDown(self):
+        os.unlink(self.temporary_fh_name)
 
     def test_default(self):
         """ TableSummarizer functions as expected

diff --git a/biom/tests/test_cli/test_table_converter.py b/biom/tests/test_cli/test_table_converter.py
@@ -8,6 +8,7 @@
 # The full license is in the file COPYING.txt, distributed with this software.
 # -----------------------------------------------------------------------------
 
+import os
 from os.path import abspath, dirname, join
 import tempfile
 
@@ -28,16 +29,18 @@ def setUp(self):
         self.cmd = _convert
         self.output_filepath = tempfile.NamedTemporaryFile().name
 
-        with tempfile.NamedTemporaryFile('w') as fh:
+        with tempfile.NamedTemporaryFile('w', delete=False) as fh:
             fh.write(biom1)
             fh.flush()
             self.biom_table1 = load_table(fh.name)
+            self.temporary_fh_table_name = fh.name
 
         self.biom_lines1 = biom1.split('\n')
-        with tempfile.NamedTemporaryFile('w') as fh:
+        with tempfile.NamedTemporaryFile('w', delete=False) as fh:
             fh.write(classic1)
             fh.flush()
             self.classic_biom1 = load_table(fh.name)
+            self.temporary_fh_classic_name = fh.name
 
         self.sample_md1 = MetadataMap.from_file(sample_md1.split('\n'))
 
@@ -47,6 +50,10 @@ def setUp(self):
         self.json_collapsed_samples = join(test_data_dir,
                                            'json_sample_collapsed.biom')
 
+    def tearDown(self):
+        os.unlink(self.temporary_fh_classic_name)
+        os.unlink(self.temporary_fh_table_name)
+
     def test_classic_to_biom(self):
         """Correctly converts classic to biom."""
         self.cmd(table=self.classic_biom1,

diff --git a/biom/tests/test_cli/test_table_normalizer.py b/biom/tests/test_cli/test_table_normalizer.py
@@ -24,9 +24,9 @@ def setUp(self):
         self.cmd = _normalize_table
 
         cwd = os.getcwd()
-        if '/' in __file__:
-            os.chdir(__file__.rsplit('/', 1)[0])
-        self.table = biom.load_table('test_data/test.json')
+        if os.path.sep in __file__:
+            os.chdir(os.path.dirname(__file__))
+        self.table = biom.load_table(os.path.join('test_data', 'test.json'))
         os.chdir(cwd)
 
     def test_bad_inputs(self):

diff --git a/biom/tests/test_cli/test_validate_table.py b/biom/tests/test_cli/test_validate_table.py
@@ -39,7 +39,8 @@ def setUp(self):
         self.to_remove = []
 
         cur_path = os.path.split(os.path.abspath(__file__))[0]
-        examples_path = os.path.join(cur_path.rsplit('/', 3)[0], 'examples')
+        examples_path = os.path.join(cur_path.rsplit(os.path.sep, 3)[0],
+                                     'examples')
         self.hdf5_file_valid = os.path.join(examples_path,
                                             'min_sparse_otu_table_hdf5.biom')
         self.hdf5_file_valid_md = os.path.join(examples_path,

diff --git a/biom/tests/test_parse.py b/biom/tests/test_parse.py
@@ -46,14 +46,19 @@ def setUp(self):
         self.legacy_otu_table1 = legacy_otu_table1
         self.otu_table1 = otu_table1
         self.otu_table1_floats = otu_table1_floats
-        self.files_to_remove = []
+        self.to_remove = []
         self.biom_minimal_sparse = biom_minimal_sparse
 
         self.classic_otu_table1_w_tax = classic_otu_table1_w_tax.split('\n')
         self.classic_otu_table1_no_tax = classic_otu_table1_no_tax.split('\n')
         self.classic_table_with_complex_metadata = \
             classic_table_with_complex_metadata.split('\n')
 
+    def tearDown(self):
+        if self.to_remove:
+            for f in self.to_remove:
+                os.remove(f)
+
     def test_from_tsv_bug_854(self):
         data = StringIO('#FeatureID\tSample1')
         exp = Table([], [], ['Sample1'])
@@ -281,38 +286,40 @@ def test_parse_adjacency_table_no_header(self):
     def test_parse_biom_table_hdf5(self):
         """Make sure we can parse a HDF5 table through the same loader"""
         cwd = os.getcwd()
-        if '/' in __file__[1:]:
-            os.chdir(__file__.rsplit('/', 1)[0])
-        Table.from_hdf5(h5py.File('test_data/test.biom', 'r'))
+        if os.path.sep in __file__[1:]:
+            os.chdir(os.path.dirname(__file__))
+        Table.from_hdf5(h5py.File(os.path.join('test_data', 'test.biom'),
+                                  'r'))
         os.chdir(cwd)
 
     def test_save_table_filepath(self):
         t = Table(np.array([[0, 1, 2], [3, 4, 5]]), ['a', 'b'],
                   ['c', 'd', 'e'])
-        with NamedTemporaryFile() as tmpfile:
+        with NamedTemporaryFile(delete=False) as tmpfile:
             save_table(t, tmpfile.name)
             obs = load_table(tmpfile.name)
             self.assertEqual(obs, t)
+        self.to_remove.append(tmpfile.name)
 
     def test_load_table_filepath(self):
         cwd = os.getcwd()
-        if '/' in __file__[1:]:
-            os.chdir(__file__.rsplit('/', 1)[0])
-        load_table('test_data/test.biom')
+        if os.path.sep in __file__[1:]:
+            os.chdir(os.path.dirname(__file__))
+        load_table(os.path.join('test_data', 'test.biom'))
         os.chdir(cwd)
 
     def test_load_table_inmemory(self):
         cwd = os.getcwd()
-        if '/' in __file__[1:]:
-            os.chdir(__file__.rsplit('/', 1)[0])
-        load_table(h5py.File('test_data/test.biom', 'r'))
+        if os.path.sep in __file__[1:]:
+            os.chdir(os.path.dirname(__file__))
+        load_table(h5py.File(os.path.join('test_data', 'test.biom'), 'r'))
         os.chdir(cwd)
 
     def test_load_table_inmemory_json(self):
         cwd = os.getcwd()
-        if '/' in __file__[1:]:
-            os.chdir(__file__.rsplit('/', 1)[0])
-        load_table(open('test_data/test.json'))
+        if os.path.sep in __file__[1:]:
+            os.chdir(os.path.dirname(__file__))
+        load_table(open(os.path.join('test_data', 'test.json')))
         os.chdir(cwd)
 
     def test_load_table_inmemory_stringio(self):
@@ -350,10 +357,11 @@ def test_parse_biom_table_with_hdf5(self):
         """tests for parse_biom_table when we have h5py"""
         # We will round-trip the HDF5 file to several different formats, and
         # make sure we can recover the same table using parse_biom_table
-        if '/' in __file__[1:]:
-            os.chdir(__file__.rsplit('/', 1)[0])
+        if os.path.sep in __file__[1:]:
+            os.chdir(os.path.dirname(__file__))
 
-        t = parse_biom_table(h5py.File('test_data/test.biom', 'r'))
+        t = parse_biom_table(h5py.File(os.path.join('test_data', 'test.biom'),
+                                       'r'))
 
         # These things are not round-trippable using the general-purpose
         # parse_biom_table function