✨ add data schema override to v2

Author: ianardee

mindee/input/inference_parameters.py

@@ -1,9 +1,41 @@
+import json
 from dataclasses import dataclass
-from typing import List, Optional
+from typing import List, Optional, Union
 
 from mindee.input.polling_options import PollingOptions
 
 
+class DataSchema:
+    """Modify the Data Schema."""
+
+    _replace: Optional[dict] = None
+
+    def __init__(self, replace: Optional[dict] = None):
+        self._replace = replace
+
+    @property
+    def replace(self):
+        """If set, completely replaces the data schema of the model."""
+        return self._replace
+
+    @replace.setter
+    def replace(self, value: Optional[Union[dict, str]]) -> None:
+        if value is None:
+            _replace = None
+        elif isinstance(value, str):
+            _replace = json.loads(value)
+        elif isinstance(value, dict):
+            _replace = value
+        else:
+            raise TypeError("Invalid type for data schema")
+        if _replace is not None and _replace == {}:
+            raise ValueError("Empty override provided")
+        self._replace = _replace
+
+    def __str__(self) -> str:
+        return json.dumps({"replace": self.replace})
+
+
 @dataclass
 class InferenceParameters:
     """Inference parameters to set when sending a file."""
@@ -30,4 +62,12 @@ class InferenceParameters:
     close_file: bool = True
     """Whether to close the file after parsing."""
     text_context: Optional[str] = None
-    """Additional text context used by the model during inference. Not recommended, for specific use only."""
+    """
+    Additional text context used by the model during inference.
+    Not recommended, for specific use only.
+    """
+    data_schema: Optional[DataSchema] = None
+    """
+    Dynamic changes to the data schema of the model for this inference.
+    Not recommended, for specific use only.
+    """

mindee/mindee_http/mindee_api_v2.py

@@ -98,8 +98,10 @@ def req_post_inference_enqueue(
             data["webhook_ids"] = params.webhook_ids
         if params.alias and len(params.alias):
             data["alias"] = params.alias
-        if params.text_context and (params.text_context):
+        if params.text_context and len(params.text_context):
             data["text_context"] = params.text_context
+        if params.data_schema is not None:
+            data["data_schema"] = str(params.data_schema)
 
         if isinstance(input_source, LocalInputSource):
             files = {"file": input_source.read_contents(params.close_file)}

mindee/parsing/v2/inference_active_options.py

@@ -1,6 +1,18 @@
 from mindee.parsing.common.string_dict import StringDict
 
 
+class DataSchemaActiveOptions:
+    """Data schema options activated during the inference."""
+
+    replace: bool
+
+    def __init__(self, raw_response: StringDict):
+        self.replace = raw_response["replace"]
+
+    def __str__(self) -> str:
+        return f"Data Schema\n-----------\n:Replace: {self.replace}"
+
+
 class InferenceActiveOptions:
     """Active options for the inference."""
 
@@ -29,13 +41,16 @@ class InferenceActiveOptions:
     Whether the text context feature was activated.
     When this feature is activated, the provided context is used to improve the accuracy of the inference.
     """
+    data_schema: DataSchemaActiveOptions
+    """Data schema options provided for the inference."""
 
     def __init__(self, raw_response: StringDict):
         self.raw_text = raw_response["raw_text"]
         self.polygon = raw_response["polygon"]
         self.confidence = raw_response["confidence"]
         self.rag = raw_response["rag"]
         self.text_context = raw_response["text_context"]
+        self.data_schema = DataSchemaActiveOptions(raw_response["data_schema"])
 
     def __str__(self) -> str:
         return (
@@ -44,4 +59,5 @@ def __str__(self) -> str:
             f"\n:Polygon: {self.polygon}"
             f"\n:Confidence: {self.confidence}"
             f"\n:RAG: {self.rag}"
+            f"\n:Text Context: {self.text_context}"
         )

tests/data

@@ -14,7 +14,7 @@ def file_path() -> Path:
 
 def _assert_local_response(local_response):
     fake_hmac_signing = "ogNjY44MhvKPGTtVsI8zG82JqWQa68woYQH"
-    signature = "b82a515c832fd2c4f4ce3a7e6f53c12e8d10e19223f6cf0e3a9809a7a3da26be"
+    signature = "1df388c992d87897fe61dfc56c444c58fc3c7369c31e2b5fd20d867695e93e85"
 
     assert local_response._file is not None
     assert not local_response.is_valid_hmac_signature(