Add pip wheel cache parser for origin.json files

src/packagedcode/__init__.py

@@ -171,6 +171,7 @@
     pypi.PypiEggHandler,
     # pypi.PypiSdistArchiveHandler,
     pypi.PypiWheelHandler,
+    pypi.PipCacheWheelHandler,
     pypi.PyprojectTomlHandler,
     pypi.PoetryPyprojectTomlHandler,
     pypi.PoetryLockHandler,

src/packagedcode/pypi.py

@@ -2661,3 +2661,77 @@ def get_requirement_from_section(section, sub_section):
     packages, _ = get_requirements_txt_dependencies(location=location)
     for req in packages:
         yield req.extracted_requirement
+
+
+class PipCacheWheelHandler(models.DatafileHandler):
+
+
+    datasource_id = 'pip_wheel_cache'
+    default_package_type = 'pypi'
+    default_primary_language = 'Python'
+    path_patterns = ('*/.cache/pip/wheels/*/*/*/*/origin.json',)
+    filetypes = ('json',)
+
+    @classmethod
+    def parse(cls, location, package_only=False):
+
+        with open(location, 'r') as f:
+            data = json.load(f)
+
+        url = data.get('url', '')
+        archive_info = data.get('archive_info', {})
+        sha256 = archive_info.get('hash', '').replace('sha256=', '')
+
+        if url:
+            filename = posixpath.basename(url)
+            name_version = filename.replace('.tar.gz', '').replace('.zip', '')
+
+            if '-' in name_version:
+                parts = name_version.rsplit('-', 1)
+                name = parts[0]
+                version = parts[1] if len(parts) > 1 else None
+            else:
+                name = name_version
+                version = None
+        else:
+            name = None
+            version = None
+
+        package_data = dict(
+            datasource_id=cls.datasource_id,
+            type=cls.default_package_type,
+            name=name,
+            version=version,
+            download_url=url,
+            sha256=sha256,
+        )
+        yield models.PackageData.from_data(package_data, package_only)
+
+    @classmethod
+    def assemble(cls, package_data, resource, codebase, package_adder):
+
+        name = package_data.get('name')
+        version = package_data.get('version')
+
+        if not name:
+            return
+
+        package = models.Package(
+            datasource_id=cls.datasource_id,
+            type=cls.default_package_type,
+            name=name,
+            version=version,
+            download_url=package_data.get('download_url'),
+            sha256=package_data.get('sha256'),
+        )
+
+        if version:
+            package.purl = PackageURL(
+                type='pypi',
+                name=name,
+                version=version,
+            ).to_string()
+
+        package_adder(package.package_uid, package, resource)
+
+        yield package

tests/packagedcode/data/pip_cache/wheels/construct/origin.json

@@ -0,0 +1 @@
+{"archive_info": {"hash": "sha256=7b2a3fd8e5f597a5aa1d614c3bd516fa065db01704c72a1efaaeec6ef23d8b45", "hashes": {"sha256": "7b2a3fd8e5f597a5aa1d614c3bd516fa065db01704c72a1efaaeec6ef23d8b45"}}, "url": "https://files.pythonhosted.org/packages/e0/b7/a4a032e94bcfdff481f2e6fecd472794d9da09f474a2185ed33b2c7cad64/construct-2.10.68.tar.gz"}

tests/packagedcode/data/pip_cache/wheels/construct/origin.json.expected

@@ -0,0 +1,46 @@
+[
+  {
+    "type": "pypi",
+    "namespace": null,
+    "name": "construct",
+    "version": "2.10.68",
+    "qualifiers": {},
+    "subpath": null,
+    "primary_language": null,
+    "description": null,
+    "release_date": null,
+    "parties": [],
+    "keywords": [],
+    "homepage_url": null,
+    "download_url": "https://files.pythonhosted.org/packages/e0/b7/a4a032e94bcfdff481f2e6fecd472794d9da09f474a2185ed33b2c7cad64/construct-2.10.68.tar.gz",
+    "size": null,
+    "sha1": null,
+    "md5": null,
+    "sha256": "7b2a3fd8e5f597a5aa1d614c3bd516fa065db01704c72a1efaaeec6ef23d8b45",
+    "sha512": null,
+    "bug_tracking_url": null,
+    "code_view_url": null,
+    "vcs_url": null,
+    "copyright": null,
+    "holder": null,
+    "declared_license_expression": null,
+    "declared_license_expression_spdx": null,
+    "license_detections": [],
+    "other_license_expression": null,
+    "other_license_expression_spdx": null,
+    "other_license_detections": [],
+    "extracted_license_statement": null,
+    "notice_text": null,
+    "source_packages": [],
+    "file_references": [],
+    "is_private": false,
+    "is_virtual": false,
+    "extra_data": {},
+    "dependencies": [],
+    "repository_homepage_url": null,
+    "repository_download_url": null,
+    "api_data_url": null,
+    "datasource_id": "pip_wheel_cache",
+    "purl": "pkg:pypi/construct@2.10.68"
+  }
+]

tests/packagedcode/test_pypi.py

@@ -753,3 +753,12 @@ def test_parse_setup_py(test_loc):
 def test_parse_more_setup_py(test_loc):
     check_setup_py_parsing(test_loc)
 
+
+class TestPipCacheWheelHandler(PackageTester):
+    test_data_dir = os.path.join(os.path.dirname(__file__), 'data')
+
+    def test_parse_pip_cache_wheel(self):
+        test_file = self.get_test_loc('pip_cache/wheels/construct/origin.json')
+        expected_loc = self.get_test_loc('pip_cache/wheels/construct/origin.json.expected')
+        package = pypi.PipCacheWheelHandler.parse(test_file)
+        self.check_packages_data(package, expected_loc, regen=REGEN_TEST_FIXTURES)