Support ExtractByMistral function spec declaration in Python SDK.

python/cocoindex/flow.py

@@ -60,8 +60,15 @@ def _create_data_slice(
 def _spec_kind(spec: Any) -> str:
     return spec.__class__.__name__
 
-def _spec_dump(spec: Any) -> dict[str, Any]:
-    return spec.__dict__
+def _spec_value_dump(spec: Any) -> Any:
+    """Recursively dump a spec object and its nested attributes to a dictionary."""
+    if hasattr(spec, '__dict__'):
+        return {k: _spec_value_dump(v) for k, v in spec.__dict__.items()}
+    elif isinstance(spec, (list, tuple)):
+        return [_spec_value_dump(item) for item in spec]
+    elif isinstance(spec, dict):
+        return {k: _spec_value_dump(v) for k, v in spec.items()}
+    return spec
 
 T = TypeVar('T')
 
@@ -161,7 +168,7 @@ def transform(self, fn_spec: op.FunctionSpec, /, name: str | None = None) -> Dat
             lambda target_scope, name:
                 flow_builder_state.engine_flow_builder.transform(
                     _spec_kind(fn_spec),
-                    _spec_dump(fn_spec),
+                    _spec_value_dump(fn_spec),
                     args,
                     target_scope,
                     flow_builder_state.field_name_builder.build_name(
@@ -252,7 +259,7 @@ def export(self, name: str, target_spec: op.StorageSpec, /, *,
             {"field_name": field_name, "metric": metric.value}
             for field_name, metric in vector_index]
         self._flow_builder_state.engine_flow_builder.export(
-            name, _spec_kind(target_spec), _spec_dump(target_spec),
+            name, _spec_kind(target_spec), _spec_value_dump(target_spec),
             index_options, self._engine_data_collector)
 
 
@@ -293,7 +300,7 @@ def add_source(self, spec: op.SourceSpec, /, name: str | None = None) -> DataSli
             self._state,
             lambda target_scope, name: self._state.engine_flow_builder.add_source(
                 _spec_kind(spec),
-                _spec_dump(spec),
+                _spec_value_dump(spec),
                 target_scope,
                 self._state.field_name_builder.build_name(
                     name, prefix=_to_snake_case(_spec_kind(spec))+'_'),

python/cocoindex/functions.py

@@ -1,4 +1,5 @@
 """All builtin functions."""
+from dataclasses import dataclass
 from typing import Annotated, Any
 
 import sentence_transformers
@@ -11,6 +12,20 @@ class SplitRecursively(op.FunctionSpec):
     chunk_overlap: int
     language: str | None = None
 
+@dataclass
+class MistralModelSpec:
+    """A specification for a Mistral model."""
+    model_id: str
+    isq_type: str
+
+class ExtractByMistral(op.FunctionSpec):
+    """Extract information from a text using a Mistral model."""
+
+    model: MistralModelSpec
+    # Expected to be generated by `cocoindex.typing.encode_enriched_type()`
+    output_type: dict[str, Any]
+    instructions: str | None = None
+
 class SentenceTransformerEmbed(op.FunctionSpec):
     """
     `SentenceTransformerEmbed` embeds a text into a vector space using the [SentenceTransformer](https://huggingface.co/sentence-transformers) library.

python/cocoindex/typing.py

@@ -2,7 +2,7 @@
 import collections
 import dataclasses
 import types
-from typing import Annotated, NamedTuple, Any, TypeVar, TYPE_CHECKING
+from typing import Annotated, NamedTuple, Any, TypeVar, TYPE_CHECKING, overload
 
 class Vector(NamedTuple):
     dim: int | None
@@ -182,6 +182,13 @@ def encode_enriched_type_info(enriched_type_info: AnalyzedTypeInfo) -> dict[str,
 
     return encoded
 
+@overload
+def encode_enriched_type(t: None) -> None:
+    ...
+
+@overload
+def encode_enriched_type(t: Any) -> dict[str, Any]:
+    ...
 
 def encode_enriched_type(t) -> dict[str, Any] | None:
     """