morphik 0.1.7__tar.gz → 0.1.9__tar.gz

{morphik-0.1.7 → morphik-0.1.9}/PKG-INFO

@@ -1,12 +1,12 @@
 Metadata-Version: 2.4
 Name: morphik
-Version: 0.1.7
+Version: 0.1.9
 Summary: Morphik Python Client
 Author-email: Morphik <founders@morphik.ai>
 Requires-Python: >=3.8
 Requires-Dist: httpx>=0.24.0
-Requires-Dist: pillow==10.4.0
-Requires-Dist: pydantic==2.10.3
+Requires-Dist: pillow>=10.4.0
+Requires-Dist: pydantic>=2.10.3
 Requires-Dist: pyjwt>=2.0.0
 Requires-Dist: requests>=2.32.3
 Description-Content-Type: text/markdown

{morphik-0.1.7 → morphik-0.1.9}/morphik/__init__.py

@@ -12,4 +12,4 @@ __all__ = [
     "Document",
 ]
 
-__version__ = "0.1.7"
+__version__ = "0.1.8"

{morphik-0.1.7 → morphik-0.1.9}/morphik/_internal.py

@@ -357,6 +357,7 @@ class _MorphikClientLogic:
         sources: List[Union[ChunkSource, Dict[str, Any]]],
         folder_name: Optional[Union[str, List[str]]],
         end_user_id: Optional[str],
+        use_colpali: bool = True,
     ) -> Dict[str, Any]:
         """Prepare request for batch_get_chunks endpoint"""
         source_dicts = []
@@ -366,15 +367,14 @@ class _MorphikClientLogic:
             else:
                 source_dicts.append(source.model_dump())
 
-        if folder_name or end_user_id:
-            request = {"sources": source_dicts}
-            if folder_name:
-                request["folder_name"] = folder_name
-            if end_user_id:
-                request["end_user_id"] = end_user_id
-            return request
-        # Return the dictionary structure { "sources": [...] } consistently.
-        return {"sources": source_dicts}
+        # Always include use_colpali flag so the server can decide how to
+        # enrich chunks.  Keep any additional scoping parameters.
+        request: Dict[str, Any] = {"sources": source_dicts, "use_colpali": use_colpali}
+        if folder_name:
+            request["folder_name"] = folder_name
+        if end_user_id:
+            request["end_user_id"] = end_user_id
+        return request
 
     def _prepare_create_graph_request(
         self,

{morphik-0.1.7 → morphik-0.1.9}/morphik/async_.py

@@ -460,6 +460,7 @@ class AsyncFolder:
         self,
         sources: List[Union[ChunkSource, Dict[str, Any]]],
         additional_folders: Optional[List[str]] = None,
+        use_colpali: bool = True,
     ) -> List[FinalChunkResult]:
         """
         Retrieve specific chunks by their document ID and chunk number in a single batch operation within this folder.
@@ -467,12 +468,15 @@ class AsyncFolder:
         Args:
             sources: List of ChunkSource objects or dictionaries with document_id and chunk_number
             additional_folders: Optional list of additional folder names to further scope operations
+            use_colpali: Whether to use ColPali-style embedding model
 
         Returns:
             List[FinalChunkResult]: List of chunk results
         """
         merged = self._merge_folders(additional_folders)
-        request = self._client._logic._prepare_batch_get_chunks_request(sources, merged, None)
+        request = self._client._logic._prepare_batch_get_chunks_request(
+            sources, merged, None, use_colpali
+        )
         response = await self._client._request("POST", "batch/chunks", data=request)
         return self._client._logic._parse_chunk_result_list_response(response)
 
@@ -998,6 +1002,7 @@ class AsyncUserScope:
         self,
         sources: List[Union[ChunkSource, Dict[str, Any]]],
         folder_name: Optional[Union[str, List[str]]] = None,
+        use_colpali: bool = True,
     ) -> List[FinalChunkResult]:
         """
         Retrieve specific chunks by their document ID and chunk number in a single batch operation for this end user.
@@ -1005,11 +1010,14 @@ class AsyncUserScope:
         Args:
             sources: List of ChunkSource objects or dictionaries with document_id and chunk_number
             folder_name: Optional folder name (or list of names) to scope the request
+            use_colpali: Whether to use ColPali-style embedding model
 
         Returns:
             List[FinalChunkResult]: List of chunk results
         """
-        request = self._client._logic._prepare_batch_get_chunks_request(sources, self._folder_name, self._end_user_id)
+        request = self._client._logic._prepare_batch_get_chunks_request(
+            sources, self._folder_name, self._end_user_id, use_colpali
+        )
         response = await self._client._request("POST", "batch/chunks", data=request)
         return self._client._logic._parse_chunk_result_list_response(response)
 
@@ -2184,6 +2192,7 @@ class AsyncMorphik:
         self,
         sources: List[Union[ChunkSource, Dict[str, Any]]],
         folder_name: Optional[Union[str, List[str]]] = None,
+        use_colpali: bool = True,
     ) -> List[FinalChunkResult]:
         """
         Retrieve specific chunks by their document ID and chunk number in a single batch operation.
@@ -2191,6 +2200,7 @@ class AsyncMorphik:
         Args:
             sources: List of ChunkSource objects or dictionaries with document_id and chunk_number
             folder_name: Optional folder name (or list of names) to scope the request
+            use_colpali: Whether to use ColPali-style embedding model
 
         Returns:
             List[FinalChunkResult]: List of chunk results
@@ -2215,7 +2225,9 @@ class AsyncMorphik:
                 print(f"Chunk from {chunk.document_id}, number {chunk.chunk_number}: {chunk.content[:50]}...")
             ```
         """
-        request = self._logic._prepare_batch_get_chunks_request(sources, folder_name, None)
+        request = self._logic._prepare_batch_get_chunks_request(
+            sources, folder_name, None, use_colpali
+        )
         response = await self._request("POST", "batch/chunks", data=request)
         return self._logic._parse_chunk_result_list_response(response)
 

{morphik-0.1.7 → morphik-0.1.9}/morphik/sync.py

@@ -474,27 +474,23 @@ class Folder:
         self,
         sources: List[Union[ChunkSource, Dict[str, Any]]],
         additional_folders: Optional[List[str]] = None,
+        use_colpali: bool = True,
     ) -> List[FinalChunkResult]:
         """
-        Retrieve specific chunks by their document ID and chunk number in a single batch operation within this folder.
+        Retrieve specific chunks by their document ID and chunk number in this folder.
 
         Args:
             sources: List of ChunkSource objects or dictionaries with document_id and chunk_number
             additional_folders: Optional list of extra folders to include in the scope
+            use_colpali: Whether to request multimodal chunks when available
 
         Returns:
             List[FinalChunkResult]: List of chunk results
         """
-        # Convert to list of dictionaries if needed
-        source_dicts = []
-        for source in sources:
-            if isinstance(source, dict):
-                source_dicts.append(source)
-            else:
-                source_dicts.append(source.model_dump())
-
         merged = self._merge_folders(additional_folders)
-        request = {"sources": source_dicts, "folder_name": merged}
+        request = self._client._logic._prepare_batch_get_chunks_request(
+            sources, merged, None, use_colpali
+        )
 
         response = self._client._request("POST", "batch/chunks", data=request)
         return self._client._logic._parse_chunk_result_list_response(response)
@@ -1078,30 +1074,23 @@ class UserScope:
         self,
         sources: List[Union[ChunkSource, Dict[str, Any]]],
         additional_folders: Optional[List[str]] = None,
+        use_colpali: bool = True,
     ) -> List[FinalChunkResult]:
         """
-        Retrieve specific chunks by their document ID and chunk number in a single batch operation for this end user.
+        Retrieve specific chunks by their document ID and chunk number in this folder.
 
         Args:
             sources: List of ChunkSource objects or dictionaries with document_id and chunk_number
             additional_folders: Optional list of extra folders to include in the scope
+            use_colpali: Whether to request multimodal chunks when available
 
         Returns:
             List[FinalChunkResult]: List of chunk results
         """
-        # Convert to list of dictionaries if needed
-        source_dicts = []
-        for source in sources:
-            if isinstance(source, dict):
-                source_dicts.append(source)
-            else:
-                source_dicts.append(source.model_dump())
-
         merged = self._merge_folders(additional_folders)
-        request = {"sources": source_dicts, "end_user_id": self._end_user_id}
-
-        if merged:
-            request["folder_name"] = merged
+        request = self._client._logic._prepare_batch_get_chunks_request(
+            sources, merged, None, use_colpali
+        )
 
         response = self._client._request("POST", "batch/chunks", data=request)
         return self._client._logic._parse_chunk_result_list_response(response)
@@ -2344,7 +2333,10 @@ class Morphik:
         return docs
 
     def batch_get_chunks(
-        self, sources: List[Union[ChunkSource, Dict[str, Any]]], folder_name: Optional[Union[str, List[str]]] = None
+        self,
+        sources: List[Union[ChunkSource, Dict[str, Any]]],
+        folder_name: Optional[Union[str, List[str]]] = None,
+        use_colpali: bool = True,
     ) -> List[FinalChunkResult]:
         """
         Retrieve specific chunks by their document ID and chunk number.
@@ -2376,7 +2368,9 @@ class Morphik:
                 print(f"Chunk from {chunk.document_id}, number {chunk.chunk_number}: {chunk.content[:50]}...")
             ```
         """
-        request = self._logic._prepare_batch_get_chunks_request(sources, folder_name, None)
+        request = self._logic._prepare_batch_get_chunks_request(
+            sources, folder_name, None, use_colpali
+        )
         response = self._request("POST", "batch/chunks", data=request)
         return self._logic._parse_chunk_result_list_response(response)
 

{morphik-0.1.7 → morphik-0.1.9}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "hatchling.build"
 
 [project]
 name = "morphik"
-version = "0.1.7"
+version = "0.1.9"
 authors = [
     { name = "Morphik", email = "founders@morphik.ai" },
 ]
@@ -14,9 +14,9 @@ requires-python = ">=3.8"
 dependencies = [
     "httpx>=0.24.0",
     "pyjwt>=2.0.0",
-    "pydantic==2.10.3",
+    "pydantic>=2.10.3",
     "requests>=2.32.3",
-    "pillow==10.4.0"
+    "pillow>=10.4.0"
 ]
 
 [tool.hatch.build.targets.wheel]