add golden test, optional_wd resolves to full path

.github/workflows/pytest-golden.yml

@@ -0,0 +1,29 @@
+name: Golden Test
+
+on:
+  pull_request:
+    branches:
+      - main
+
+jobs:
+  pytest:
+    runs-on: ubuntu-latest
+
+    strategy:
+      matrix:
+        python_version: ["3.12"]
+
+    steps:
+      - uses: actions/checkout@v4
+
+      - uses: mamba-org/setup-micromamba@v1
+        with:
+          environment-file: environment.yml
+          create-args: >-
+            python=${{ matrix.python_version }}
+
+      - name: Pytest in conda environment
+        shell: bash -l {0}
+        run: |
+          python -m pip install --no-deps .
+          pytest -m golden

.github/workflows/pytest-integration.yml

@@ -26,4 +26,4 @@ jobs:
         shell: bash -l {0}
         run: |
           python -m pip install --no-deps .
-          pytest --cov=burst2safe -m integration
+          pytest -m integration

pyproject.toml

@@ -51,8 +51,8 @@ develop = [
 
 [tool.pytest.ini_options]
 minversion = "6.0"
-addopts = '-ra -q -m "not integration"'
-markers = ["integration"]
+addopts = '-ra -q -m "not integration and not golden"'
+markers = ["integration", "golden"]
 testpaths = ["tests"]
 
 [tool.black]

src/burst2safe/measurement.py

@@ -79,6 +79,15 @@ def get_burst_byte_offsets(self):
         byte_offsets = [offsets[self.burst_length * i] for i in range(len(self.burst_infos))]
         return byte_offsets
 
+    def get_time_tag(self) -> str:
+        """Get the current time as a time tag.
+        This is a separate method to allow for easy mocking in tests.
+
+        Returns:
+            The time tag as a string
+        """
+        return datetime.strftime(datetime.now(), '%Y:%m:%d %H:%M:%S')
+
     def add_metadata(self, dataset: gdal.Dataset):
         """Add metadata to an existing GDAL dataset.
 
@@ -90,7 +99,7 @@ def add_metadata(self, dataset: gdal.Dataset):
         srs.ImportFromEPSG(4326)
         dataset.SetGCPs(gdal_gcps, srs.ExportToWkt())
 
-        dataset.SetMetadataItem('TIFFTAG_DATETIME', datetime.strftime(datetime.now(), '%Y:%m:%d %H:%M:%S'))
+        dataset.SetMetadataItem('TIFFTAG_DATETIME', self.get_time_tag())
         # TODO make sure A/B is being set correctly.
         dataset.SetMetadataItem('TIFFTAG_IMAGEDESCRIPTION', 'Sentinel-1A IW SLC L1')
         dataset.SetMetadataItem('TIFFTAG_SOFTWARE', f'Sentinel-1 IPF {self.version}')

src/burst2safe/utils.py

@@ -156,7 +156,7 @@ def optional_wd(wd: Optional[Path | str] = None) -> Path:
     """
     if wd is None:
         wd = Path.cwd()
-    return Path(wd)
+    return Path(wd).resolve()
 
 
 def calculate_crc16(file_path: Path) -> str:

tests/test_golden.py

@@ -0,0 +1,77 @@
+import filecmp
+from pathlib import Path
+from unittest.mock import patch
+from zipfile import ZipFile
+
+import pytest
+import requests
+from shapely import box
+
+from burst2safe.burst2safe import burst2safe
+
+
+def bit_for_bit(reference: Path, secondary: Path):
+    filecmp.clear_cache()
+    return filecmp.cmp(reference, secondary, shallow=False)
+
+
+def prep_golden_safe(tmp_path: Path):
+    golden_url = 'https://ffwilliams2-shenanigans.s3.us-west-2.amazonaws.com/burst2safe/S1A_IW_SLC__1SSV_20240103T015108_20240103T015112_051936_064669_51EE.zip'
+
+    golden_dir = tmp_path / 'golden'
+    golden_dir.mkdir(exist_ok=True)
+
+    golden_zip = golden_dir / Path(golden_url).name
+    with requests.get(golden_url, stream=True) as r:
+        r.raise_for_status()
+        with open(golden_zip, 'wb') as f:
+            for chunk in r.iter_content(chunk_size=2**22):
+                f.write(chunk)
+
+    with ZipFile(golden_zip, 'r') as z:
+        z.extractall(golden_dir)
+
+    golden_safe = golden_zip.with_suffix('.SAFE')
+    return golden_safe
+
+
+@pytest.mark.golden()
+def test_golden(tmp_path):
+    safe_name = 'S1A_IW_SLC__1SSV_20240103T015108_20240103T015112_051936_064669_51EE.SAFE'
+    golden_safe = (tmp_path / 'golden' / safe_name).resolve()
+    new_safe = (tmp_path / safe_name).resolve()
+
+    prep_golden_safe(tmp_path)
+
+    with patch('burst2safe.measurement.Measurement.get_time_tag') as mock_get_time_tag:
+        mock_get_time_tag.return_value = '2024:01:01 00:00:00'
+        burst2safe(
+            granules=[],
+            orbit=51936,
+            footprint=box(*[-117.3, 35.5, -117.2, 35.6]),
+            polarizations=['VV', 'VH'],
+            keep_files=True,
+            work_dir=tmp_path,
+        )
+
+    golden_files = sorted([x.resolve() for x in golden_safe.rglob('*')])
+    new_files = sorted([x.resolve() for x in new_safe.rglob('*')])
+
+    golden_set = set([f.relative_to(golden_safe) for f in golden_files])
+    new_set = set([f.relative_to(new_safe) for f in new_files])
+
+    only_in_golden = [str(x) for x in golden_set - new_set]
+    assert not only_in_golden
+
+    only_in_new = [str(x) for x in new_set - golden_set]
+    assert not only_in_new
+
+    differing_files = []
+    for golden_file, new_file in zip(golden_files, new_files):
+        if golden_file.is_dir():
+            continue
+
+        if not bit_for_bit(golden_file, new_file):
+            differing_files.append(new_file.relative_to(new_safe))
+    differing_files = [str(x) for x in differing_files]
+    assert not differing_files

tests/test_utils.py

@@ -96,7 +96,7 @@ def test_optional_wd():
     existing_dir = 'working'
     wd = utils.optional_wd(existing_dir)
     assert isinstance(wd, Path)
-    assert wd == Path(existing_dir)
+    assert wd == Path(existing_dir).resolve()
 
 
 def test_calculate_crc16(tmp_path, test_data_dir):