Merge pull request #1039 from ocefpaf/remove_pkg_resources

Remove pkg resources

.pre-commit-config.yaml

@@ -1,6 +1,6 @@
 repos:
 - repo: https://github.com/pre-commit/pre-commit-hooks
-  rev: v4.5.0
+  rev: v4.6.0
   hooks:
     - id: trailing-whitespace
       exclude: compliance_checker/tests/data
@@ -19,7 +19,7 @@ repos:
         - test_requirements.txt
 
 - repo: https://github.com/psf/black
-  rev: 24.1.1
+  rev: 24.3.0
   hooks:
   - id: black
     language_version: python3
@@ -31,7 +31,7 @@ repos:
 
 
 - repo: https://github.com/astral-sh/ruff-pre-commit
-  rev: v0.2.0
+  rev: v0.3.5
   hooks:
     - id: ruff
 

compliance_checker/cf/cf_1_6.py

@@ -2764,12 +2764,7 @@ def check_cell_boundaries(self, ds):
             ):
                 valid = False
                 reasoning.append(
-                    "Dimension {} of boundary variable (for {}) must have at least {} elements to form a simplex/closed cell with previous dimensions {}.".format(
-                        boundary_variable.name,
-                        variable.name,
-                        len(variable.dimensions) + 1,
-                        boundary_variable.dimensions[:-1],
-                    ),
+                    f"Dimension {boundary_variable.name} of boundary variable (for {variable.name}) must have at least {len(variable.dimensions) + 1} elements to form a simplex/closed cell with previous dimensions {boundary_variable.dimensions[:-1]}.",
                 )
             result = Result(
                 BaseCheck.MEDIUM,

compliance_checker/cf/cf_1_7.py

@@ -283,12 +283,7 @@ def check_cell_boundaries(self, ds):
             ):
                 valid = False
                 reasoning.append(
-                    "Dimension {} of boundary variable (for {}) must have at least {} elements to form a simplex/closed cell with previous dimensions {}.".format(
-                        boundary_variable.name,
-                        variable.name,
-                        len(variable.dimensions) + 1,
-                        boundary_variable.dimensions[:-1],
-                    ),
+                    f"Dimension {boundary_variable.name} of boundary variable (for {variable.name}) must have at least {len(variable.dimensions) + 1} elements to form a simplex/closed cell with previous dimensions {boundary_variable.dimensions[:-1]}.",
                 )
 
             # 7.1 Required 3/5:

compliance_checker/cf/util.py

@@ -1,3 +1,4 @@
+import importlib.resources
 import itertools
 import os
 import sys
@@ -7,7 +8,6 @@
 from cf_units import Unit
 from lxml import etree
 from netCDF4 import Dataset
-from pkg_resources import resource_filename
 
 # copied from paegan
 # paegan may depend on these later
@@ -284,9 +284,9 @@ def download_cf_standard_name_table(version, location=None):
     if (
         location is None
     ):  # This case occurs when updating the packaged version from command line
-        location = resource_filename(
-            "compliance_checker",
-            "data/cf-standard-name-table.xml",
+        location = (
+            importlib.resources.files("compliance_checker")
+            / "data/cf-standard-name-table.xml"
         )
 
     if version == "latest":

compliance_checker/cfutil.py

@@ -3,13 +3,13 @@
 compliance_checker/cfutil.py
 """
 import csv
+import importlib.resources
 import re
 import warnings
 from collections import defaultdict
 from functools import lru_cache, partial
 
 from cf_units import Unit
-from pkg_resources import resource_filename
 
 _UNITLESS_DB = None
 _SEA_NAMES = None
@@ -128,7 +128,7 @@ def get_sea_names():
     if _SEA_NAMES is None:
         buf = {}
         with open(
-            resource_filename("compliance_checker", "data/seanames.csv"),
+            importlib.resources.files("compliance_checker") / "data/seanames.csv",
         ) as f:
             reader = csv.reader(f)
             for code, sea_name in reader:

compliance_checker/protocols/netcdf.py

@@ -17,6 +17,7 @@ def is_netcdf(url):
     :param str url: Location of file on the file system
     """
     # Try an obvious exclusion of remote resources
+    url = str(url)
     if url.startswith("http"):
         return False
 

compliance_checker/suite.py

@@ -13,7 +13,6 @@
 import warnings
 from collections import defaultdict
 from datetime import datetime, timezone
-from distutils.version import StrictVersion
 from operator import itemgetter
 from pathlib import Path
 from urllib.parse import urlparse
@@ -23,6 +22,7 @@
 from netCDF4 import Dataset
 from owslib.sos import SensorObservationService
 from owslib.swe.sensor.sml import SensorML
+from packaging.version import parse
 from pkg_resources import working_set
 
 from compliance_checker import __version__, tempnc
@@ -186,9 +186,8 @@ def _load_checkers(cls, checkers):
         for spec, versions in itertools.groupby(ver_checkers, itemgetter(0)):
             version_nums = [v[-1] for v in versions]
             try:
-                latest_version = str(max(StrictVersion(v) for v in version_nums))
-            # if the version can't be parsed as a StrictVersion, parse
-            # according to character collation
+                latest_version = str(max(parse(v) for v in version_nums))
+            # if the version can't be parsed, do it according to character collation
             except ValueError:
                 latest_version = max(version_nums)
             cls.checkers[spec] = cls.checkers[spec + ":latest"] = cls.checkers[
@@ -765,12 +764,9 @@ def generate_dataset(self, cdl_path):
 
         :param str cdl_path: Absolute path to cdl file that is used to generate netCDF file
         """
-        if (
-            ".cdl" in cdl_path
-        ):  # it's possible the filename doesn't have the .cdl extension
-            ds_str = cdl_path.replace(".cdl", ".nc")
-        else:
-            ds_str = cdl_path + ".nc"
+        if isinstance(cdl_path, str):
+            cdl_path = Path(cdl_path)
+        ds_str = cdl_path.with_suffix(".nc")
 
         # generate netCDF-4 file
         iostat = subprocess.run(

compliance_checker/tests/__init__.py

@@ -1,4 +1,5 @@
 import unittest
+from pathlib import Path
 
 from netCDF4 import Dataset
 
@@ -25,8 +26,8 @@ def load_dataset(self, nc_dataset):
         """
         Return a loaded NC Dataset for the given path
         """
-        if not isinstance(nc_dataset, str):
-            raise ValueError("nc_dataset should be a string")
+        if not isinstance(nc_dataset, (str, Path)):
+            raise ValueError("nc_dataset should be a valid path")
 
         nc_dataset = Dataset(nc_dataset, "r")
         self.addCleanup(nc_dataset.close)

compliance_checker/tests/conftest.py

@@ -1,11 +1,10 @@
+import importlib.resources
 import os
 import subprocess
 from itertools import chain
-from pathlib import Path
 
 import pytest
 from netCDF4 import Dataset
-from pkg_resources import resource_filename
 
 from compliance_checker.cf import util
 from compliance_checker.suite import CheckSuite
@@ -27,7 +26,9 @@ def static_files(cdl_stem):
     Returns the Path to a valid nc dataset\n
     replaces the old STATIC_FILES dict
     """
-    datadir = Path(resource_filename("compliance_checker", "tests/data")).resolve()
+    datadir = (
+        importlib.resources.files("compliance_checker").joinpath("tests/data").resolve()
+    )
     assert datadir.exists(), f"{datadir} not found"
 
     cdl_paths = glob_down(datadir, f"{cdl_stem}.cdl", 3)

compliance_checker/tests/resources.py

@@ -1,16 +1,14 @@
-import os
+import importlib.resources
 import subprocess
 
-from pkg_resources import resource_filename
-
 
 def get_filename(path):
     """
     Returns the path to a valid dataset
     """
-    filename = resource_filename("compliance_checker", path)
-    nc_path = filename.replace(".cdl", ".nc")
-    if not os.path.exists(nc_path):
+    filename = importlib.resources.files("compliance_checker") / path
+    nc_path = filename.with_suffix(".nc")
+    if not nc_path.exists():
         generate_dataset(filename, nc_path)
     return nc_path
 

compliance_checker/tests/test_cf_integration.py

@@ -45,9 +45,7 @@
             '§2.6.1 Conventions global attribute does not contain "CF-1.8"',
             f"standard_name visibility is not defined in Standard Name Table v{std_names._version}. Possible close match(es): ['visibility_in_air']",
             'Standard name modifier "data_quality" for variable visibility_qc is not a valid modifier according to CF Appendix C',
-            "standard_name wind_direction is not defined in Standard Name Table v{}. Possible close match(es): ['wind_to_direction', 'wind_from_direction', 'wind_gust_from_direction']".format(
-                std_names._version,
-            ),
+            f"standard_name wind_direction is not defined in Standard Name Table v{std_names._version}. Possible close match(es): ['wind_to_direction', 'wind_from_direction', 'wind_gust_from_direction']",
             'Standard name modifier "data_quality" for variable wind_direction_qc is not a valid modifier according to CF Appendix C',
             f"standard_name wind_gust is not defined in Standard Name Table v{std_names._version}. Possible close match(es): ['y_wind_gust', 'x_wind_gust', 'wind_speed_of_gust']",
             'Standard name modifier "data_quality" for variable wind_gust_qc is not a valid modifier according to CF Appendix C',

compliance_checker/tests/test_suite.py

@@ -1,35 +1,30 @@
+import importlib.resources
 import os
 import unittest
 from pathlib import Path
 
 import numpy as np
-from pkg_resources import resource_filename
 
 from compliance_checker.acdd import ACDDBaseCheck
 from compliance_checker.base import BaseCheck, GenericFile, Result
 from compliance_checker.suite import CheckSuite
 
 static_files = {
-    "2dim": resource_filename("compliance_checker", "tests/data/2dim-grid.nc"),
-    "bad_region": resource_filename("compliance_checker", "tests/data/bad_region.nc"),
-    "bad_data_type": resource_filename(
-        "compliance_checker",
-        "tests/data/bad_data_type.nc",
-    ),
-    "test_cdl": resource_filename("compliance_checker", "tests/data/test_cdl.cdl"),
-    "test_cdl_nc": resource_filename(
-        "compliance_checker",
-        "tests/data/test_cdl_nc_file.nc",
-    ),
-    "empty": resource_filename("compliance_checker", "tests/data/non-comp/empty.file"),
-    "ru07": resource_filename(
-        "compliance_checker",
-        "tests/data/ru07-20130824T170228_rt0.nc",
-    ),
-    "netCDF4": resource_filename(
-        "compliance_checker",
-        "tests/data/test_cdl_nc4_file.cdl",
-    ),
+    "2dim": importlib.resources.files("compliance_checker") / "tests/data/2dim-grid.nc",
+    "bad_region": importlib.resources.files("compliance_checker")
+    / "tests/data/bad_region.nc",
+    "bad_data_type": importlib.resources.files("compliance_checker")
+    / "tests/data/bad_data_type.nc",
+    "test_cdl": importlib.resources.files("compliance_checker")
+    / "tests/data/test_cdl.cdl",
+    "test_cdl_nc": importlib.resources.files("compliance_checker")
+    / "tests/data/test_cdl_nc_file.nc",
+    "empty": importlib.resources.files("compliance_checker")
+    / "tests/data/non-comp/empty.file",
+    "ru07": importlib.resources.files("compliance_checker")
+    / "tests/data/ru07-20130824T170228_rt0.nc",
+    "netCDF4": importlib.resources.files("compliance_checker")
+    / "tests/data/test_cdl_nc4_file.cdl",
 }
 
 
@@ -95,9 +90,9 @@ def test_generate_dataset_netCDF4(self):
         # create netCDF4 file
         ds_name = self.cs.generate_dataset(static_files["netCDF4"])
         # check if correct name is return
-        assert ds_name == static_files["netCDF4"].replace(".cdl", ".nc")
+        assert ds_name == static_files["netCDF4"].with_suffix(".nc")
         # check if netCDF4 file was created
-        assert os.path.isfile(static_files["netCDF4"].replace(".cdl", ".nc"))
+        assert os.path.isfile(static_files["netCDF4"].with_suffix(".nc"))
 
     def test_include_checks(self):
         ds = self.cs.load_dataset(static_files["bad_data_type"])
@@ -242,7 +237,7 @@ def test_cdl_file(self):
             )
         ds.close()
 
-        nc_file_path = static_files["test_cdl"].replace(".cdl", ".nc")
+        nc_file_path = static_files["test_cdl"].with_suffix(".nc")
         self.addCleanup(os.remove, nc_file_path)
 
         # Ok the scores should be equal!

pyproject.toml

@@ -7,7 +7,7 @@ requires = [
 ]
 
 [tool.ruff]
-select = [
+lint.select = [
     "A", # flake8-builtins
     "B", # flake8-bugbear
     "C4", # flake8-comprehensions
@@ -24,11 +24,11 @@ exclude = [
   "compliance_checker/cf/cf.py",
 ]
 
-ignore = [
+lint.ignore = [
   "E501",
 ]
 
-[tool.ruff.per-file-ignores]
+[tool.ruff.lint.per-file-ignores]
 "docs/source/conf.py" = [
   "E402",
   "A001",

requirements.txt

@@ -5,6 +5,7 @@ jinja2>=2.7.3
 lxml>=3.2.1
 netcdf4>=1.6.4
 owsLib>=0.8.3
+packaging
 pendulum>=1.2.4
 pygeoif>=0.6
 pyproj>=2.2.1