Merge pull request #52 from gazpachoking/recursive_madness

Allow self referential references

Author: gazpachoking

.github/workflows/test.yml

@@ -15,6 +15,7 @@ jobs:
           - "3.8"
           - "3.9"
           - "3.10"
+          - "3.11"
 
     steps:
       - uses: actions/checkout@v3
@@ -35,4 +36,4 @@ jobs:
           flake8 . --count --exit-zero --max-complexity=10 --max-line-length=127 --statistics
       - name: Test with pytest
         run: |
-          pytest tests.py
+          pytest tests.py

README.md

@@ -7,7 +7,7 @@
 
 `jsonref` is a library for automatic dereferencing of [JSON
 Reference](https://datatracker.ietf.org/doc/html/draft-pbryan-zyp-json-ref-03)
-objects for Python (supporting Python 3.3+).
+objects for Python (supporting Python 3.7+).
 
 This library lets you use a data structure with JSON reference objects,
 as if the references had been replaced with the referent data.

docs/index.rst

@@ -8,7 +8,7 @@ jsonref
 
 ``jsonref`` is a library for automatic dereferencing of
 `JSON Reference <https://tools.ietf.org/id/draft-pbryan-zyp-json-ref-03.html>`_
-objects for Python (supporting Python 3.3+).
+objects for Python (supporting Python 3.7+).
 
 .. testcode::
 

jsonref.py

@@ -17,7 +17,7 @@
 
 from proxytypes import LazyProxy
 
-__version__ = "1.0.1"
+__version__ = "1.1.0"
 
 
 class JsonRefError(Exception):
@@ -124,22 +124,20 @@ def callback(self):
         uri, fragment = urlparse.urldefrag(self.full_uri)
 
         # If we already looked this up, return a reference to the same object
-        if uri in self.store:
-            result = self.resolve_pointer(self.store[uri], fragment)
-        else:
+        if uri not in self.store:
             # Remote ref
             try:
                 base_doc = self.loader(uri)
             except Exception as e:
                 raise self._error(
                     "%s: %s" % (e.__class__.__name__, str(e)), cause=e
                 ) from e
-
-            kwargs = self._ref_kwargs
-            kwargs["base_uri"] = uri
-            kwargs["recursing"] = False
-            base_doc = _replace_refs(base_doc, **kwargs)
-            result = self.resolve_pointer(base_doc, fragment)
+            base_doc = _replace_refs(
+                base_doc, **{**self._ref_kwargs, "base_uri": uri, "recursing": False}
+            )
+        else:
+            base_doc = self.store[uri]
+        result = self.resolve_pointer(base_doc, fragment)
         if result is self:
             raise self._error("Reference refers directly to itself.")
         if hasattr(result, "__subject__"):
@@ -174,6 +172,9 @@ def resolve_pointer(self, document, pointer):
                     part = int(part)
                 except ValueError:
                     pass
+            # If a reference points inside itself, it must mean inside reference object, not the referent data
+            if document is self:
+                document = self.__reference__
             try:
                 document = document[part]
             except (TypeError, LookupError) as e:
@@ -362,25 +363,7 @@ def _replace_refs(
             base_uri = urlparse.urljoin(base_uri, id_)
             store_uri = base_uri
 
-    try:
-        if not isinstance(obj["$ref"], str):
-            raise TypeError
-    except (TypeError, LookupError):
-        pass
-    else:
-        return JsonRef(
-            obj,
-            base_uri=base_uri,
-            loader=loader,
-            jsonschema=jsonschema,
-            load_on_repr=load_on_repr,
-            merge_props=merge_props,
-            _path=path,
-            _store=store,
-        )
-
-    # If our obj was not a json reference object, iterate through it,
-    # replacing children with JsonRefs
+    # First recursively iterate through our object, replacing children with JsonRefs
     if isinstance(obj, Mapping):
         obj = {
             k: _replace_refs(
@@ -411,8 +394,24 @@ def _replace_refs(
             )
             for i, v in enumerate(obj)
         ]
+
+    # If this object itself was a reference, replace it with a JsonRef
+    if isinstance(obj, Mapping) and isinstance(obj.get("$ref"), str):
+        obj = JsonRef(
+            obj,
+            base_uri=base_uri,
+            loader=loader,
+            jsonschema=jsonschema,
+            load_on_repr=load_on_repr,
+            merge_props=merge_props,
+            _path=path,
+            _store=store,
+        )
+
+    # Store the document with all references replaced in our cache
     if store_uri is not None:
         store[store_uri] = obj
+
     return obj
 
 
@@ -432,7 +431,7 @@ def load(
     proxied to their referent data.
 
     :param fp: File-like object containing JSON document
-    :param kwargs: This function takes any of the keyword arguments from
+    :param **kwargs: This function takes any of the keyword arguments from
         :func:`replace_refs`. Any other keyword arguments will be passed to
         :func:`json.load`
 
@@ -469,7 +468,7 @@ def loads(
     proxied to their referent data.
 
     :param s: String containing JSON document
-    :param kwargs: This function takes any of the keyword arguments from
+    :param **kwargs: This function takes any of the keyword arguments from
         :func:`replace_refs`. Any other keyword arguments will be passed to
         :func:`json.loads`
 
@@ -505,7 +504,7 @@ def load_uri(
     data.
 
     :param uri: URI to fetch the JSON from
-    :param kwargs: This function takes any of the keyword arguments from
+    :param **kwargs: This function takes any of the keyword arguments from
         :func:`replace_refs`
 
     """

pyproject.toml

@@ -7,7 +7,7 @@ authors = [
 license = {text = "MIT"}
 readme = "README.md"
 dynamic = ["version"]
-requires-python = ">=3.3"
+requires-python = ">=3.7"
 dependencies = []
 
 [project.urls]
@@ -28,4 +28,3 @@ build-backend = "pdm.pep517.api"
 
 [tool.isort]
 profile = "black"
-

tests.py

@@ -113,6 +113,18 @@ def test_extra_ref_attributes(self, parametrized_replace_refs):
             }
         }
 
+    def test_refs_inside_extra_props(self, parametrized_replace_refs):
+        """This seems really dubious per the spec... but OpenAPI 3.1 spec does it."""
+        docs = {
+            "a.json": {
+                "file": "a",
+                "b": {"$ref": "b.json#/ba", "extra": {"$ref": "b.json#/bb"}},
+            },
+            "b.json": {"ba": {"a": 1}, "bb": {"b": 2}},
+        }
+        result = parametrized_replace_refs(docs["a.json"], loader=docs.get, merge_props=True)
+        assert result == {"file": "a", "b": {"a": 1, "extra": {"b": 2}}}
+
     def test_recursive_extra(self, parametrized_replace_refs):
         json = {"a": {"$ref": "#", "extra": "foo"}}
         result = parametrized_replace_refs(json, merge_props=True)
@@ -234,6 +246,16 @@ def test_recursive_data_structures_remote_fragment(self):
         result = replace_refs(json1, base_uri="/json1", loader=loader)
         assert result["a"].__subject__ is result
 
+    def test_self_referent_reference(self, parametrized_replace_refs):
+        json = {"$ref": "#/sub", "sub": [1, 2]}
+        result = parametrized_replace_refs(json)
+        assert result == json["sub"]
+
+    def test_self_referent_reference_w_merge(self, parametrized_replace_refs):
+        json = {"$ref": "#/sub", "extra": "aoeu", "sub": {"main": "aoeu"}}
+        result = parametrized_replace_refs(json, merge_props=True)
+        assert result == {"main": "aoeu", "extra": "aoeu", "sub": {"main": "aoeu"}}
+
     def test_custom_loader(self):
         data = {"$ref": "foo"}
         loader = mock.Mock(return_value=42)