pattrrnProperties matches all patterns against all keys

In draft 4, patternProperties creates an all-of set of schemas for each property. Each property gets compared against all patterns. And it must match all of the schemas for the patterns it matches. So, we cannot avoid compiling some of the patterns like I thought.
https://datatracker.ietf.org/doc/html/draft-fge-json-schema-validation-00#section-8.3

Author: cognifloyd

st2common/st2common/util/config_loader.py

@@ -137,20 +137,28 @@ def _get_object_properties_schema(object_schema, object_keys=None):
         pattern_properties = object_schema.get("patternProperties", {})
         # patternProperties can be a boolean or a dict
         if pattern_properties and isinstance(pattern_properties, dict):
-            for raw_pattern, pattern_schema in pattern_properties.items():
-                pattern = re.compile(raw_pattern)
-                for key in list(extra_keys):
+            # we need to match all extra_keys against all patterns
+            # and then compose the per-property schema from all
+            # the matched patterns' properties.
+            pattern_properties = {
+                re.compile(raw_pattern): pattern_schema
+                for raw_pattern, pattern_schema in pattern_properties.items()
+            }
+            for key in list(extra_keys):
+                key_schemas = []
+                for pattern, pattern_schema in pattern_properties.items():
                     if pattern.search(key):
-                        # update matched key
-                        flattened_properties_schema[key] = pattern_schema
-                        # don't check matched key against any more patterns
-                        # and don't overwrite with additionalProperties
-                        extra_keys.remove(key)
-
-                if not extra_keys:
-                    # nothing to check. Don't compile any more patterns
-                    # and don't look at additionalProperties.
-                    return flattened_properties_schema
+                        key_schemas.append(pattern_schema)
+                if key_schemas:
+                    composed_schema = {**schema for schema in key_schemas}
+                    # update matched key
+                    flattened_properties_schema[key] = composed_schema
+                    # don't overwrite matched key's schema with additionalProperties
+                    extra_keys.remove(key)
+
+            if not extra_keys:
+                # nothing else to check. Don't look at additionalProperties.
+                return flattened_properties_schema
 
         # fill in any remaining keys with additionalProperties
         additional_properties = object_schema.get("additionalProperties", {})