geometamaker 0.2.0__tar.gz → 0.2.1__tar.gz

{geometamaker-0.2.0 → geometamaker-0.2.1}/.github/workflows/test.yml

@@ -22,7 +22,7 @@ jobs:
     strategy:
       fail-fast: false
       matrix:
-        python-version: ["3.9", "3.10", "3.11", "3.12", "3.13"]
+        python-version: ["3.10", "3.11", "3.12", "3.13", "3.14"]
         os: [ubuntu-latest, windows-latest, macos-latest]
 
     steps:

{geometamaker-0.2.0 → geometamaker-0.2.1}/HISTORY.rst

@@ -1,6 +1,16 @@
 Release History
 ===============
 
+0.2.1 (2026-02-02)
+------------------
+* Make sure that the option to compute band statistics will include
+  STATISTICS_VALID_PERCENT. https://github.com/natcap/geometamaker/issues/106
+* Improve the error message if frictionless raises an exception while
+  trying to detect a filetype.
+  https://github.com/natcap/geometamaker/issues/107
+* Fixed bug where non-ascii characters in filepaths raised an exception
+  in ``describe``. https://github.com/natcap/geometamaker/issues/112
+
 0.2.0 (2025-07-22)
 ------------------
 * Allow CLI to ``describe`` remote datasets.

{geometamaker-0.2.0 → geometamaker-0.2.1}/PKG-INFO

@@ -1,8 +1,8 @@
 Metadata-Version: 2.4
 Name: geometamaker
-Version: 0.2.0
+Version: 0.2.1
 Summary: metadata creation for geospatial data
-Maintainer: Natural Capital Project Software Team
+Maintainer: Natural Capital Alliance Software Team
 License: 
                                          Apache License
                                    Version 2.0, January 2004
@@ -215,11 +215,11 @@ Classifier: Operating System :: MacOS :: MacOS X
 Classifier: Operating System :: Microsoft
 Classifier: Operating System :: POSIX
 Classifier: Programming Language :: Python :: 3
-Classifier: Programming Language :: Python :: 3.9
 Classifier: Programming Language :: Python :: 3.10
 Classifier: Programming Language :: Python :: 3.11
 Classifier: Programming Language :: Python :: 3.12
 Classifier: Programming Language :: Python :: 3.13
+Classifier: Programming Language :: Python :: 3.14
 Classifier: License :: OSI Approved :: Apache Software License
 Classifier: Topic :: Scientific/Engineering :: GIS
 Requires-Python: >=3.9

{geometamaker-0.2.0 → geometamaker-0.2.1}/pyproject.toml

@@ -5,7 +5,7 @@ readme = "README.md"
 requires-python = ">=3.9"
 license = {file = "LICENSE.txt"}
 maintainers = [
-    {name = "Natural Capital Project Software Team"}
+    {name = "Natural Capital Alliance Software Team"}
 ]
 keywords = ["gis", "geospatial", "metadata"]
 classifiers = [
@@ -17,11 +17,11 @@ classifiers = [
     "Operating System :: Microsoft",
     "Operating System :: POSIX",
     "Programming Language :: Python :: 3",
-    "Programming Language :: Python :: 3.9",
     "Programming Language :: Python :: 3.10",
     "Programming Language :: Python :: 3.11",
     "Programming Language :: Python :: 3.12",
     "Programming Language :: Python :: 3.13",
+    "Programming Language :: Python :: 3.14",
     "License :: OSI Approved :: Apache Software License",
     "Topic :: Scientific/Engineering :: GIS"
 ]

{geometamaker-0.2.0 → geometamaker-0.2.1}/src/geometamaker/geometamaker.py

@@ -39,6 +39,12 @@ PROTOCOLS = [
 DT_FMT = '%Y-%m-%d %H:%M:%S %Z'
 
 
+def _gdal_progress_callback(complete, message, data):
+    percentage = complete * 100
+    if (percentage > 0) & (percentage % 5 == 0):
+        LOGGER.info(f'{message} {percentage}%')
+
+
 # TODO: In the future we can remove these exception managers in favor of the
 # builtin gdal.ExceptionMgr. It was released in 3.7.0 and debugged in 3.9.1.
 # https://github.com/OSGeo/gdal/blob/v3.9.3/NEWS.md#gdalogr-391-release-notes
@@ -203,7 +209,10 @@ def detect_file_type(filepath, scheme):
     # We'll likely want a different data model for multi-dimensional arrays.
     # Frictionless supports a wide range of formats. The quickest way to
     # determine if a file is recognized as a table or archive is to call list.
-    info = frictionless.list(filepath)[0]
+    try:
+        info = frictionless.list(filepath)[0]
+    except frictionless.FrictionlessException:
+        raise RuntimeError(f'Cannot detect file type of "{filepath}"')
     if info.type == 'table':
         return 'table'
     # Frictionless doesn't recognize .tgz compression (but does recognize .tar.gz)
@@ -259,7 +268,7 @@ def describe_file(source_dataset_path, scheme):
     hash_func = hashlib.new('sha256')
     hash_func.update(
         f'{description["bytes"]}{description["last_modified"]}\
-        {description["path"]}'.encode('ascii'))
+        {description["path"]}'.encode('utf-8'))
     description['uid'] = f'sizetimestamp:{hash_func.hexdigest()}'
 
     # We don't have a use for including these attributes in our metadata:
@@ -388,18 +397,24 @@ def describe_raster(source_dataset_path, scheme, **kwargs):
     for i in range(info['n_bands']):
         b = i + 1
         band = raster.GetRasterBand(b)
+        band_gdal_metadata = band.GetMetadata()
         if compute_stats:
             try:
-                # 0=do not approximate stats, 1=calculate if they don't exist
-                # If exact stats exist they will be retrieved without
-                # computing them, otherwise, this forces computation.
-                # https://github.com/OSGeo/gdal/blob/master/gcore/gdalrasterband.cpp
-                _ = band.GetStatistics(0, 1)
+                if not 'STATISTICS_VALID_PERCENT' in band_gdal_metadata:
+                    # Sometimes some stats exist, but not all. If this one doesn't,
+                    # it's important enough that we want to force computation.
+                    _ = band.ComputeStatistics(0, callback=_gdal_progress_callback)
+                else:
+                    # 0=do not approximate stats, 1=calculate if they don't exist
+                    # If exact stats exist they will be retrieved without
+                    # computing them, otherwise, this forces computation.
+                    # https://github.com/OSGeo/gdal/blob/master/gcore/gdalrasterband.cpp
+                    _ = band.GetStatistics(0, 1)
+                band_gdal_metadata = band.GetMetadata()
             except RuntimeError as e:
                 LOGGER.warning(
                     f'Could not compute statistics for band {b} of '
                     f'{source_dataset_path}: {e}')
-        band_gdal_metadata = band.GetMetadata()
 
         bands.append(models.BandSchema(
             index=b,

{geometamaker-0.2.0 → geometamaker-0.2.1}/src/geometamaker.egg-info/PKG-INFO

@@ -1,8 +1,8 @@
 Metadata-Version: 2.4
 Name: geometamaker
-Version: 0.2.0
+Version: 0.2.1
 Summary: metadata creation for geospatial data
-Maintainer: Natural Capital Project Software Team
+Maintainer: Natural Capital Alliance Software Team
 License: 
                                          Apache License
                                    Version 2.0, January 2004
@@ -215,11 +215,11 @@ Classifier: Operating System :: MacOS :: MacOS X
 Classifier: Operating System :: Microsoft
 Classifier: Operating System :: POSIX
 Classifier: Programming Language :: Python :: 3
-Classifier: Programming Language :: Python :: 3.9
 Classifier: Programming Language :: Python :: 3.10
 Classifier: Programming Language :: Python :: 3.11
 Classifier: Programming Language :: Python :: 3.12
 Classifier: Programming Language :: Python :: 3.13
+Classifier: Programming Language :: Python :: 3.14
 Classifier: License :: OSI Approved :: Apache Software License
 Classifier: Topic :: Scientific/Engineering :: GIS
 Requires-Python: >=3.9

{geometamaker-0.2.0 → geometamaker-0.2.1}/tests/test_geometamaker.py

@@ -59,11 +59,11 @@ def create_vector(target_filepath, field_map=None, driver='GEOJSON'):
 
 def create_raster(
         numpy_dtype, target_path,
-        pixel_size=(1, 1), projection_epsg=4326,
+        pixel_size=(1, 1), raster_size=(2, 2), projection_epsg=4326,
         origin=(0, 0), n_bands=2, define_nodata=True):
     driver_name, creation_options = DEFAULT_GTIFF_CREATION_TUPLE_OPTIONS
     raster_driver = gdal.GetDriverByName(driver_name)
-    ny, nx = (2, 2)
+    nx, ny = raster_size
     gdal_type = gdal_array.NumericTypeCodeToGDALTypeCode(numpy_dtype)
     raster = raster_driver.Create(
         target_path, nx, ny, n_bands, gdal_type)
@@ -78,7 +78,7 @@ def create_raster(
     if projection_wkt is not None:
         raster.SetProjection(projection_wkt)
 
-    base_array = numpy.full((2, 2), 1, dtype=numpy_dtype)
+    base_array = numpy.full((nx, ny), 1, dtype=numpy_dtype)
 
     target_nodata = pygeoprocessing.choose_nodata(numpy_dtype)
 
@@ -96,7 +96,8 @@ class GeometamakerTests(unittest.TestCase):
 
     def setUp(self):
         """Override setUp function to create temp workspace directory."""
-        self.workspace_dir = tempfile.mkdtemp()
+        self.workspace_dir = tempfile.mkdtemp(
+            suffix='\U0001f60e')  # ensure unicode support
         self.patcher = patch('geometamaker.config.platformdirs.user_config_dir')
         self.mock_user_config_dir = self.patcher.start()
         self.mock_user_config_dir.return_value = self.workspace_dir
@@ -315,11 +316,34 @@ class GeometamakerTests(unittest.TestCase):
              'STATISTICS_VALID_PERCENT': '100'})
 
     def test_describe_raster_band_compute_statistics(self):
-        """Test band statistics will be included if they already exist."""
+        """Test band statistics will be computed if they do not already exist."""
+        import geometamaker
+
+        datasource_path = os.path.join(self.workspace_dir, 'raster.tif')
+        create_raster(numpy.int16, datasource_path, n_bands=1)
+
+        resource = geometamaker.describe(datasource_path, compute_stats=True)
+        self.assertEqual(
+            resource.data_model.bands[0].gdal_metadata,
+            {'STATISTICS_MINIMUM': '1',
+             'STATISTICS_MAXIMUM': '1',
+             'STATISTICS_MEAN': '1',
+             'STATISTICS_STDDEV': '0',
+             'STATISTICS_VALID_PERCENT': '100'})
+
+    def test_describe_raster_band_compute_statistics_valid_percent(self):
+        """Test band statistics will be computed if VALID_PERCENT is missing."""
         import geometamaker
 
         datasource_path = os.path.join(self.workspace_dir, 'raster.tif')
         create_raster(numpy.int16, datasource_path, n_bands=1)
+        raster = gdal.OpenEx(datasource_path)
+        band = raster.GetRasterBand(1)
+        # Deliberately set stats metadata that do not include VALID_PERCENT
+        band.SetMetadataItem('STATISTICS_MINIMUM', '1')
+        band.SetMetadataItem('STATISTICS_MAXIMUM', '1')
+        band.SetMetadataItem('STATISTICS_MEAN', '1')
+        band.SetMetadataItem('STATISTICS_STDDEV', '0')
 
         resource = geometamaker.describe(datasource_path, compute_stats=True)
         self.assertEqual(