Validate website field

ecosystem/validation/test_url.py

@@ -12,6 +12,8 @@
 
 """URL Validations"""
 
+from urllib.parse import urlparse
+
 import pytest
 
 # pylint: disable=missing-function-docstring, missing-class-docstring
@@ -20,7 +22,7 @@
 class TestURLs:
     @classmethod
     def get_all_urls(cls, member):
-        """recursevly search for URLs in the member data"""
+        """recursively search for URLs in the member data"""
         for key, value in member.to_dict().items():
             if isinstance(value, str) and value.startswith("http"):
                 yield getattr(member, key)
@@ -29,33 +31,114 @@ def get_all_urls(cls, member):
             else:
                 continue
 
+    @staticmethod
+    def normalize_url(url):
+        """Normalize URL for comparison."""
+
+        if url is None:
+            return None, None
+
+        parsed = urlparse(str(url))
+
+        hostname = (parsed.hostname or "").lower()
+
+        if hostname.startswith("www."):
+            hostname = hostname[4:]
+
+        path = parsed.path.rstrip("/")
+
+        return hostname, path
+
+    @staticmethod
+    def is_same_or_subpath(candidate, reference):
+        """
+        Check whether candidate duplicates or points inside reference.
+        """
+
+        candidate_host, candidate_path = TestURLs.normalize_url(candidate)
+
+        reference_host, reference_path = TestURLs.normalize_url(reference)
+
+        if candidate_host != reference_host:
+            return False
+
+        return candidate_path == reference_path or candidate_path.startswith(
+            reference_path + "/"
+        )
+
     def test_http(self, member):
         for url in TestURLs.get_all_urls(member):
             assert not str(url).startswith("http:"), f"{url} is not HTTPS"
 
     def test_025(self, member):
         """Documentation link has redundant suffix"""
+
         fields = ["url", "documentation", "reference_paper"]
+
         for field in fields:
             url = getattr(member, field)
+
             if url is None:
                 continue
-            if url.hostname.endswith("readthedocs.io"):
-                suffixes = ["en/latest/", "en/latest", "en"]
+
+            if url.hostname.lower().endswith("readthedocs.io"):
+                suffixes = [
+                    "en/latest/",
+                    "en/latest",
+                    "en",
+                ]
+
                 for suffix in suffixes:
                     assert not url.path.endswith(
                         suffix
                     ), f"{url} has redundant suffix: {suffix}"
 
     def test_026(self, member):
         """The /README.md is not documentation"""
+
         documentation_url = getattr(member, "documentation")
+
         if documentation_url is None:
             pytest.skip("No member.documentation")
-        if documentation_url.hostname.endswith("github.com"):
-            suffixes = ["main/README.md"]
+
+        if documentation_url.hostname.lower().endswith("github.com"):
+            suffixes = [
+                "main/README.md",
+                "blob/main/README.md",
+                "tree/main",
+            ]
+
             for suffix in suffixes:
                 assert not documentation_url.path.endswith(suffix), (
-                    "The field `member.documentation` can be empty. "
-                    "It does not have to be a link to the `README.md`."
+                    "The field `member.documentation` can be "
+                    "empty. It does not have to be a link to "
+                    "the repository's `README.md` or root."
                 )
+
+    def test_027(self, member):
+        """Website should not duplicate GitHub or PyPI URLs"""
+
+        website_url = getattr(member, "website")
+
+        if website_url is None:
+            pytest.skip("No member.website")
+
+        hostname = website_url.hostname or ""
+
+        # Reject GitHub repository URLs as website
+        repository_url = getattr(member, "url")
+
+        if repository_url is not None and repository_url.hostname.lower().endswith(
+            "github.com"
+        ):
+            assert not TestURLs.is_same_or_subpath(
+                website_url,
+                repository_url,
+            ), (
+                "The field `member.website` should not "
+                "duplicate the GitHub repository URL."
+            )
+
+        assert not (
+            hostname.endswith("pypi.org") or hostname.endswith("pypi.python.org")
+        ), ("The field `member.website` should not " "point to a PyPI project URL.")

resources/checks.toml

@@ -154,6 +154,15 @@ checker = "test_url.py::TestURLs::test_026"
 category = "SUBMISSION"
 importance = "RECOMMENDATION"
 
+[027]
+title = "Website should not be GitHub repository or PyPI URL"
+description = "The `member.website` field should point to the project's website instead of duplicating the GitHub repository or a PyPI project page."
+applies_to = "all"
+affects = ["member.website"]
+checker = "test_url.py::TestURLs::test_027"
+category = "SUBMISSION"
+importance = "RECOMMENDATION"
+
 [G05]
 title = "Archived GitHub repository should have `archived` support"
 applies_to = "GitHub projects"