Disable uri-reference format check. Consistently use same validator a…

…cross codebase

altair/utils/display.py

@@ -4,10 +4,9 @@
 from typing import Callable, Dict
 import uuid
 
-from jsonschema import validate
-
 from .plugin_registry import PluginRegistry
 from .mimebundle import spec_to_mimebundle
+from .schemapi import validate_jsonschema
 
 
 # ==============================================================================
@@ -126,7 +125,7 @@ def _validate(self):
         # type: () -> None
         """Validate the spec against the schema."""
         schema_dict = json.loads(pkgutil.get_data(*self.schema_path).decode("utf-8"))
-        validate(self.spec, schema_dict)
+        validate_jsonschema(self.spec, schema_dict)
 
     def _repr_mimebundle_(self, include=None, exclude=None):
         """Return a MIME bundle for display in Jupyter frontends."""

altair/utils/schemapi.py

@@ -2,6 +2,7 @@
 # tools/generate_schema_wrapper.py. Do not modify directly.
 import collections
 import contextlib
+import functools
 import inspect
 import json
 from typing import Any
@@ -10,6 +11,7 @@
 import numpy as np
 import pandas as pd
 
+JSONSCHEMA_VALIDATOR = jsonschema.Draft7Validator
 
 # If DEBUG_MODE is True, then schema objects are converted to dict and
 # validated at creation time. This slows things down, particularly for
@@ -40,6 +42,35 @@ def debug_mode(arg):
         DEBUG_MODE = original
 
 
+@functools.wraps(jsonschema.validate)
+def validate_jsonschema(*args, **kwargs):
+    # We always want to use the same jsonschema validator across the whole codebase
+    validator_cls = JSONSCHEMA_VALIDATOR
+    kwargs["cls"] = validator_cls
+
+    removed_format_checkers = []
+    try:
+        # The "uri-reference" checker fails for some of the Vega-Lite
+        # schemas as URIs in "$ref" are not encoded,
+        # e.g. '#/definitions/ValueDefWithCondition<MarkPropFieldOrDatumDef,
+        # (Gradient|string|null)>' would be a valid $ref in a Vega-Lite schema but
+        # it is not a valid URI reference due to the characters such as '<'.
+        # This is fine and we can disable this format check below-
+        for format_name in ["uri-reference"]:
+            try:
+                checker = validator_cls.FORMAT_CHECKER.checkers.pop(format_name)
+                removed_format_checkers.append((format_name, checker))
+            except KeyError:
+                continue
+        output = jsonschema.validate(*args, **kwargs)
+    finally:
+        # Restore the original set of checkers as the jsonschema package
+        # might also be used by other packages
+        for format_name, checker in removed_format_checkers:
+            validator_cls.FORMAT_CHECKER.checkers[format_name] = checker
+    return output
+
+
 def _subclasses(cls):
     """Breadth-first sequence of all classes which inherit from cls."""
     seen = set()
@@ -155,7 +186,7 @@ class SchemaBase(object):
     _schema = None
     _rootschema = None
     _class_is_valid_at_instantiation = True
-    _validator = jsonschema.Draft7Validator
+    _validator = JSONSCHEMA_VALIDATOR
 
     def __init__(self, *args, **kwds):
         # Two valid options for initialization, which should be handled by
@@ -445,9 +476,7 @@ def validate(cls, instance, schema=None):
         if schema is None:
             schema = cls._schema
         resolver = jsonschema.RefResolver.from_schema(cls._rootschema or cls._schema)
-        return jsonschema.validate(
-            instance, schema, cls=cls._validator, resolver=resolver
-        )
+        return validate_jsonschema(instance, schema, resolver=resolver)
 
     @classmethod
     def resolve_references(cls, schema=None):
@@ -466,7 +495,7 @@ def validate_property(cls, name, value, schema=None):
         value = _todict(value, validate=False, context={})
         props = cls.resolve_references(schema or cls._schema).get("properties", {})
         resolver = jsonschema.RefResolver.from_schema(cls._rootschema or cls._schema)
-        return jsonschema.validate(value, props.get(name, {}), resolver=resolver)
+        return validate_jsonschema(value, props.get(name, {}), resolver=resolver)
 
     def __dir__(self):
         return list(self._kwds.keys())
@@ -560,7 +589,7 @@ def from_dict(
             for possible_schema in schemas:
                 resolver = jsonschema.RefResolver.from_schema(rootschema)
                 try:
-                    jsonschema.validate(dct, possible_schema, resolver=resolver)
+                    validate_jsonschema(dct, possible_schema, resolver=resolver)
                 except jsonschema.ValidationError:
                     continue
                 else:

tools/schemapi/schemapi.py

@@ -1,5 +1,6 @@
 import collections
 import contextlib
+import functools
 import inspect
 import json
 from typing import Any
@@ -8,6 +9,7 @@
 import numpy as np
 import pandas as pd
 
+JSONSCHEMA_VALIDATOR = jsonschema.Draft7Validator
 
 # If DEBUG_MODE is True, then schema objects are converted to dict and
 # validated at creation time. This slows things down, particularly for
@@ -38,6 +40,35 @@ def debug_mode(arg):
         DEBUG_MODE = original
 
 
+@functools.wraps(jsonschema.validate)
+def validate_jsonschema(*args, **kwargs):
+    # We always want to use the same jsonschema validator across the whole codebase
+    validator_cls = JSONSCHEMA_VALIDATOR
+    kwargs["cls"] = validator_cls
+
+    removed_format_checkers = []
+    try:
+        # The "uri-reference" checker fails for some of the Vega-Lite
+        # schemas as URIs in "$ref" are not encoded,
+        # e.g. '#/definitions/ValueDefWithCondition<MarkPropFieldOrDatumDef,
+        # (Gradient|string|null)>' would be a valid $ref in a Vega-Lite schema but
+        # it is not a valid URI reference due to the characters such as '<'.
+        # This is fine and we can disable this format check below-
+        for format_name in ["uri-reference"]:
+            try:
+                checker = validator_cls.FORMAT_CHECKER.checkers.pop(format_name)
+                removed_format_checkers.append((format_name, checker))
+            except KeyError:
+                continue
+        output = jsonschema.validate(*args, **kwargs)
+    finally:
+        # Restore the original set of checkers as the jsonschema package
+        # might also be used by other packages
+        for format_name, checker in removed_format_checkers:
+            validator_cls.FORMAT_CHECKER.checkers[format_name] = checker
+    return output
+
+
 def _subclasses(cls):
     """Breadth-first sequence of all classes which inherit from cls."""
     seen = set()
@@ -153,7 +184,7 @@ class SchemaBase(object):
     _schema = None
     _rootschema = None
     _class_is_valid_at_instantiation = True
-    _validator = jsonschema.Draft7Validator
+    _validator = JSONSCHEMA_VALIDATOR
 
     def __init__(self, *args, **kwds):
         # Two valid options for initialization, which should be handled by
@@ -443,9 +474,7 @@ def validate(cls, instance, schema=None):
         if schema is None:
             schema = cls._schema
         resolver = jsonschema.RefResolver.from_schema(cls._rootschema or cls._schema)
-        return jsonschema.validate(
-            instance, schema, cls=cls._validator, resolver=resolver
-        )
+        return validate_jsonschema(instance, schema, resolver=resolver)
 
     @classmethod
     def resolve_references(cls, schema=None):
@@ -464,7 +493,7 @@ def validate_property(cls, name, value, schema=None):
         value = _todict(value, validate=False, context={})
         props = cls.resolve_references(schema or cls._schema).get("properties", {})
         resolver = jsonschema.RefResolver.from_schema(cls._rootschema or cls._schema)
-        return jsonschema.validate(value, props.get(name, {}), resolver=resolver)
+        return validate_jsonschema(value, props.get(name, {}), resolver=resolver)
 
     def __dir__(self):
         return list(self._kwds.keys())
@@ -558,7 +587,7 @@ def from_dict(
             for possible_schema in schemas:
                 resolver = jsonschema.RefResolver.from_schema(rootschema)
                 try:
-                    jsonschema.validate(dct, possible_schema, resolver=resolver)
+                    validate_jsonschema(dct, possible_schema, resolver=resolver)
                 except jsonschema.ValidationError:
                     continue
                 else: