Fix looking up anchors from non-canonical URIs.

E.g. if we have http://example.com, a retrieval URI,
we need to be able to lookup anchors within it even if the
resource has some internal ID.

Author: Julian

docs/changes.rst

@@ -2,6 +2,12 @@
 Changelog
 =========
 
+v0.27.0
+-------
+
+* Support looking up anchors from non-canonical URIs.
+  In other words, if you add a resource at the URI ``http://example.com``, then looking up the anchor ``http://example.com#foo`` now works even if the resource has some internal ``$id`` saying its canonical URI is ``http://somethingelse.example.com``.
+
 v0.26.4
 -------
 

referencing/_core.py

@@ -349,14 +349,20 @@ def remove(self, uri: URI):
 
     def anchor(self, uri: URI, name: str):
         """
-        Retrieve the given anchor, which must already have been found.
+        Retrieve a given anchor from a resource which must already be crawled.
         """
-        value = self._anchors.get((uri, name))
+        resource = self.get(uri)
+        if resource is None:
+            canonical_uri = uri
+        else:
+            canonical_uri = resource.id() or uri
+
+        value = self._anchors.get((canonical_uri, name))
         if value is not None:
             return Retrieved(value=value, registry=self)
 
         registry = self.crawl()
-        value = registry._anchors.get((uri, name))
+        value = registry._anchors.get((canonical_uri, name))
         if value is not None:
             return Retrieved(value=value, registry=registry)
         if "/" in name: