JSONSchemata 0.0.0.dev0__tar.gz → 0.0.0.dev1__tar.gz

{jsonschemata-0.0.0.dev0/src/JSONSchemata.egg-info → jsonschemata-0.0.0.dev1}/PKG-INFO

@@ -1,7 +1,7 @@
 Metadata-Version: 2.1
 Name: JSONSchemata
-Version: 0.0.0.dev0
+Version: 0.0.0.dev1
 Requires-Python: >=3.10
 License-File: LICENSE
 Requires-Dist: jsonschema<5,>=4.21.0
-Requires-Dist: referencing==0.35.1
+Requires-Dist: referencing>=0.35.1

{jsonschemata-0.0.0.dev0 → jsonschemata-0.0.0.dev1}/pyproject.toml

@@ -17,10 +17,10 @@ namespaces = true
 # ----------------------------------------- Project Metadata -------------------------------------
 #
 [project]
-version = "0.0.0.dev0"
+version = "0.0.0.dev1"
 name = "JSONSchemata"
 dependencies = [
     "jsonschema >= 4.21.0, < 5",
-    "referencing == 0.35.1",
+    "referencing >= 0.35.1",
 ]
 requires-python = ">=3.10"

{jsonschemata-0.0.0.dev0 → jsonschemata-0.0.0.dev1/src/JSONSchemata.egg-info}/PKG-INFO

@@ -1,7 +1,7 @@
 Metadata-Version: 2.1
 Name: JSONSchemata
-Version: 0.0.0.dev0
+Version: 0.0.0.dev1
 Requires-Python: >=3.10
 License-File: LICENSE
 Requires-Dist: jsonschema<5,>=4.21.0
-Requires-Dist: referencing==0.35.1
+Requires-Dist: referencing>=0.35.1

{jsonschemata-0.0.0.dev0 → jsonschemata-0.0.0.dev1}/src/JSONSchemata.egg-info/SOURCES.txt

@@ -6,4 +6,6 @@ src/JSONSchemata.egg-info/dependency_links.txt
 src/JSONSchemata.egg-info/not-zip-safe
 src/JSONSchemata.egg-info/requires.txt
 src/JSONSchemata.egg-info/top_level.txt
-src/jsonschemata/__init__.py
+src/jsonschemata/__init__.py
+src/jsonschemata/edit.py
+src/jsonschemata/registry.py

{jsonschemata-0.0.0.dev0 → jsonschemata-0.0.0.dev1}/src/JSONSchemata.egg-info/requires.txt

@@ -1,2 +1,2 @@
 jsonschema<5,>=4.21.0
-referencing==0.35.1
+referencing>=0.35.1

jsonschemata-0.0.0.dev1/src/jsonschemata/__init__.py

@@ -0,0 +1 @@
+from jsonschemata import edit, registry

jsonschemata-0.0.0.dev1/src/jsonschemata/edit.py

@@ -0,0 +1,27 @@
+def required_last(schema: dict) -> dict:
+    """Modify JSON schema to recursively put all 'required' fields at the end of the schema.
+
+    This is done because otherwise the 'required' fields
+    are checked by jsonschema before filling the defaults,
+    which can cause the validation to fail.
+
+    Returns
+    -------
+    dict
+        Modified schema.
+        Note that the input schema is modified in-place,
+        so the return value is a reference to the (now modified) input schema.
+    """
+    if "required" in schema:
+        schema["required"] = schema.pop("required")
+    for key in ["anyOf", "allOf", "oneOf", "prefixItems"]:
+        if key in schema:
+            for subschema in schema[key]:
+                required_last(subschema)
+    for key in ["if", "then", "else", "not", "items", "additionalProperties"]:
+        if key in schema and isinstance(schema[key], dict):
+            required_last(schema[key])
+    if "properties" in schema and isinstance(schema["properties"], dict):
+        for subschema in schema["properties"].values():
+            required_last(subschema)
+    return schema

jsonschemata-0.0.0.dev1/src/jsonschemata/registry.py

@@ -0,0 +1,100 @@
+from typing import Callable as _Callable
+import referencing as _referencing
+from referencing import jsonschema as _ref_jsonschema, retrieval as _ref_retrieval
+import pkgdata as _pkgdata
+import pyserials as _ps
+import pylinks as _pl
+
+from jsonschemata import edit as _edit
+
+
+def make(
+    dynamic: bool = False,
+    crawl: bool = True,
+    add_resources: list[dict | _referencing.Resource | tuple[str, dict | _referencing.Resource]] | None = None,
+    add_resources_default_spec: _referencing.Specification = _ref_jsonschema.DRAFT202012,
+    retrieval_func: _Callable[[str], str | _referencing.Resource] = None,
+) -> _referencing.Registry:
+    """Create a registry of all JSON schemas.
+
+    Parameters
+    ----------
+    dynamic : bool, default: False
+        If True, any reference that is not found in the registry will be
+        [dynamically retrieved](https://referencing.readthedocs.io/en/stable/intro/#dynamically-retrieving-resources)
+        and [cached](https://referencing.readthedocs.io/en/stable/intro/#caching).
+        The retrieval works as follows:
+        If the reference URI starts with "http" or "https", the URI is fetched using an HTTP GET request.
+        Otherwise, the URI is assumed to be a local filepath,
+        which can be either absolute or relative to the current working directory.
+        You can also provide a custom retrieval function using the `retrieval_func` parameter.
+    crawl : bool, default: True
+        Pre [crawl](https://referencing.readthedocs.io/en/stable/api/#referencing.Registry.crawl)
+        all resources so that the registry is
+        [fully ready](https://referencing.readthedocs.io/en/stable/schema-packages/).
+    add_resources: list[dict | referencing.Resource | tuple[str, dict | referencing.Resource]] | None, default: None
+        A list of additional resources to add to the registry. Each resource can be a dictionary or
+        a [`referencing.Resource`](https://referencing.readthedocs.io/en/stable/api/#referencing.Resource)
+        object. If a resource does not have an "$id",
+        the ID must be provided along with the resource as a tuple of (ID, resource).
+    add_resources_default_spec: referencing.Specification, default: referencing.jsonschema.DRAFT202012
+        The default specification to use when creating an additional resource from a dictionary.
+    retrieval_func: Callable[[str], str | referencing.Resource], optional
+        A custom retrieval function to use when `dynamic` is True.
+        The function should take a URI as input and return the reference schema.
+        If you want the retrieval function to also cache the retrieved references,
+        the function must be decorated with the
+        [`@referencing.retrieval.to_cached_resource`](https://referencing.readthedocs.io/en/stable/api/#referencing.retrieval.to_cached_resource)
+        decorator, in which case the function must return the reference schema as a JSON string
+        (cf. [Referencing Docs](https://referencing.readthedocs.io/en/stable/intro/#caching)).
+        If the decorator is not used, the function must return the schema as a `referencing.Resource` instead.
+        Note that this retrieval function will only be used when `dynamic` is set to `True`.
+    Returns
+    -------
+    referencing.Registry
+        A [`referencing.Registry`](https://referencing.readthedocs.io/en/stable/api/#referencing.Registry)
+        object containing all resources.
+    """
+
+    @_ref_retrieval.to_cached_resource()
+    def retrieve_url(uri: str) -> str:
+        if uri.startswith(("http://", "https://")):
+            return _pl.http.request(url=uri, response_type="str")
+        return _ps.write.to_json_string(_ps.read.from_file(path=uri, toml_as_dict=True), sort_keys=False)
+
+    schema_dir_path = _pkgdata.get_package_path_from_caller(top_level=True) / "_data"
+    resources = []
+    for schema_filepath in schema_dir_path.glob("**/*.yaml"):
+        schema_dict = _ps.read.yaml_from_file(path=schema_filepath)
+        _edit.required_last(schema_dict)
+        schema = _referencing.Resource.from_contents(
+            schema_dict, default_specification=_ref_jsonschema.DRAFT202012
+        )
+        resources.append(schema)
+
+    id_resources: list[tuple[str, _referencing.Resource]] = []
+    for add_resource in add_resources or []:
+        resource_id = None
+        if isinstance(add_resource, dict):
+            add_resource = _referencing.Resource.from_contents(
+                add_resource, default_specification=add_resources_default_spec
+            )
+        elif isinstance(add_resource, tuple):
+            resource_id, resource_dict = add_resource
+            add_resource = _referencing.Resource.from_contents(
+                resource_dict, default_specification=add_resources_default_spec,
+            )
+        if resource_id:
+            id_resources.append((resource_id, add_resource))
+        else:
+            resources.append(add_resource)
+
+    registry = _referencing.Registry(
+        retrieve=retrieval_func or retrieve_url
+    ) if dynamic else _referencing.Registry()
+    registry = resources @ registry
+    if id_resources:
+        registry = registry.with_resources(id_resources)
+    if crawl:
+        registry.crawl()
+    return registry