🙏 Trim dict keys

harambe/parser/parser.py

@@ -1,5 +1,5 @@
 from abc import ABC, abstractmethod
-from typing import Any, List, Optional, Type
+from typing import Any, List, Optional, Type, Union, Dict
 
 from pydantic import BaseModel, Extra, Field, NameEmail, ValidationError, create_model
 
@@ -50,7 +50,7 @@ def validate(self, data: dict[str, Any], base_url: URL) -> dict[str, Any]:
         self.model = self._schema_to_pydantic_model(self.schema)
 
         try:
-            return self.model(**data).model_dump()
+            return self.model(**trim_dict_keys(data)).model_dump()
         except ValidationError as validation_error:
             raise SchemaValidationError(
                 data=data, schema=self.schema, message=str(validation_error)
@@ -156,3 +156,13 @@ def _get_type(self, field: str) -> Type[Any]:
         if not field_type:
             raise ValueError(f"Unsupported field type: {field}")
         return field_type
+
+
+# TODO: Make this a root pre validator
+def trim_dict_keys(data: Union[Dict[str, Any], Any]) -> Union[Dict[str, Any], Any]:
+    if isinstance(data, dict):
+        return {key.strip(): trim_dict_keys(value) for key, value in data.items()}
+    elif isinstance(data, list):
+        return [trim_dict_keys(item) for item in data]
+    else:
+        return data

pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "harambe-sdk"
-version = "0.19.0"
+version = "0.19.10"
 description = "Data extraction SDK for Playwright 🐒🍌"
 authors = ["awtkns <[email protected]>"]
 readme = "README.md"

test/parser/test_parser.py

@@ -88,9 +88,9 @@
         (
             schemas.nested_lists_and_objects_schema,
             {
-                "departments": [
+                " departments": [
                     {
-                        "name": "Engineering",
+                        "name ": "Engineering",
                         "teams": [
                             {"team_name": "Backend", "members": ["Alice", "Bob"]}
                         ],
@@ -109,6 +109,17 @@
                 ]
             },
         ),
+        (
+            schemas.documents_schema,
+            {
+                "documents ": [
+                    {
+                        " title  ": "Document One",
+                        " document_url ": "/doc1",  # ✅ handles extra spaces
+                    },
+                ]
+            },
+        ),
         (
             schemas.enums_schema,
             {"season": "spring"},