morphik 0.1.6__py3-none-any.whl → 0.1.8__py3-none-any.whl

morphik/__init__.py

@@ -12,4 +12,4 @@ __all__ = [
     "Document",
 ]
 
-__version__ = "0.1.6"
+__version__ = "0.1.8"

morphik/_internal.py

@@ -11,9 +11,9 @@ from PIL import Image
 from PIL.Image import Image as PILImage
 from pydantic import BaseModel, Field
 
+from .models import ChunkSource  # Prompt override models
 from .models import (
     ChunkResult,
-    ChunkSource,  # Prompt override models
     CompletionResponse,
     Document,
     DocumentResult,
@@ -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/async_.py

@@ -8,9 +8,9 @@ import httpx
 from pydantic import BaseModel
 
 from ._internal import FinalChunkResult, RuleOrDict, _MorphikClientLogic
+from .models import CompletionResponse  # Prompt override models
 from .models import (
     ChunkSource,
-    CompletionResponse,  # Prompt override models
     Document,
     DocumentResult,
     FolderInfo,
@@ -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/models.py

@@ -303,15 +303,15 @@ class Graph(BaseModel):
 
     @property
     def is_processing(self) -> bool:
-        return self.status == "processing"
+        return self.system_metadata.get("status") == "processing"
 
     @property
     def is_completed(self) -> bool:
-        return self.status == "completed"
+        return self.system_metadata.get("status") == "completed"
 
     @property
     def is_failed(self) -> bool:
-        return self.status == "failed"
+        return self.system_metadata.get("status") == "failed"
 
     @property
     def error(self) -> str | None:

morphik/sync.py

@@ -8,9 +8,9 @@ import httpx
 from pydantic import BaseModel
 
 from ._internal import FinalChunkResult, RuleOrDict, _MorphikClientLogic
+from .models import CompletionResponse  # Prompt override models
 from .models import (
     ChunkSource,
-    CompletionResponse,  # Prompt override models
     Document,
     DocumentResult,
     FolderInfo,
@@ -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/tests/test_async.py

@@ -4,9 +4,8 @@ import uuid
 from pathlib import Path
 
 import pytest
-from pydantic import BaseModel, Field
-
 from morphik.async_ import AsyncMorphik
+from pydantic import BaseModel, Field
 
 # Set to your local Morphik server - use localhost by default
 # Default client connects to localhost:8000 automatically

morphik/tests/test_sync.py

@@ -4,9 +4,8 @@ import uuid
 from pathlib import Path
 
 import pytest
-from pydantic import BaseModel, Field
-
 from morphik.sync import Morphik
+from pydantic import BaseModel, Field
 
 # Set to your local Morphik server - use localhost by default
 # Default client connects to localhost:8000 automatically

{morphik-0.1.6.dist-info → morphik-0.1.8.dist-info}/METADATA

@@ -1,14 +1,16 @@
 Metadata-Version: 2.4
 Name: morphik
-Version: 0.1.6
+Version: 0.1.8
 Summary: Morphik Python Client
 Author-email: Morphik <founders@morphik.ai>
 Requires-Python: >=3.8
+Requires-Dist: build>=1.2.2.post1
 Requires-Dist: httpx>=0.24.0
 Requires-Dist: pillow==10.4.0
 Requires-Dist: pydantic==2.10.3
 Requires-Dist: pyjwt>=2.0.0
 Requires-Dist: requests>=2.32.3
+Requires-Dist: twine>=6.1.0
 Description-Content-Type: text/markdown
 
 # Morphik

morphik-0.1.8.dist-info/RECORD

@@ -0,0 +1,18 @@
+morphik/__init__.py,sha256=aefTQR9UbZAzbPCup-ooUscRpR8ExPknjd9b_GhMRQA,242
+morphik/_internal.py,sha256=T0rwSrgbLEMq5_YYfIyGmCIF6I8-RNtgISYFXhEoJMo,18808
+morphik/async_.py,sha256=HKUwMO2LGassUz5SqNeGA0YUvaki442ogVmfHrmKt-8,96490
+morphik/exceptions.py,sha256=v4XGmfq5B0KrZEF6M1ID8A50-45-SRAQZTrXGXM6n0Q,260
+morphik/models.py,sha256=JFMeGLbEke-Bls08JXCBv3y_z4eI3PBW_mbX-989i5I,20942
+morphik/rules.py,sha256=fw0RovS0Pwtff8Dvo3nkM3Wl6WtR3ykSaxsU_sxdXKI,2565
+morphik/sync.py,sha256=o8KOJPD4QYx6xw7aNQADK-FOI37h0otddFoJtWfV1Ig,100887
+morphik/tests/README.md,sha256=jtJDDK8cS5E4SbygFQDy7t6Y-kQwNYtZajRwVJDR62U,1069
+morphik/tests/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+morphik/tests/example_usage.py,sha256=ls8n7355q-8gY43pZLKd4SzI-01MdFeXbT8bZ4U8MCg,11561
+morphik/tests/test_async.py,sha256=HJY9g5eTMNxyxzQth29n93Df51vW7jOb7a9OSevIxTY,13901
+morphik/tests/test_sync.py,sha256=Reqa25Q259mCr-tzWzc1RDcs5KZuDfBkJRKOyyxhDtE,13480
+morphik/tests/test_docs/sample1.txt,sha256=Fx6TElSiKdxyFeBp1iHthzHctFVZm38DrqcbdZMoidY,507
+morphik/tests/test_docs/sample2.txt,sha256=PE97gPv59J27A7CSNvi_0tRBIN3Mj6pyTFElCLfs3TE,686
+morphik/tests/test_docs/sample3.txt,sha256=OzrnJ_XsDUntEV0jk-ansa3_KIa6GnpvS5EVmlh6BHo,732
+morphik-0.1.8.dist-info/METADATA,sha256=Gc8XVr-VmlZLe4a8YmhvGDrUlNpSu_9-8yh1DisiT94,3439
+morphik-0.1.8.dist-info/WHEEL,sha256=qtCwoSJWgHk21S1Kb4ihdzI2rlJ1ZKaIurTj_ngOhyQ,87
+morphik-0.1.8.dist-info/RECORD,,

morphik-0.1.6.dist-info/RECORD

@@ -1,18 +0,0 @@
-morphik/__init__.py,sha256=71PM_1UCwD19H11SscczHDND_lETA2jrOX_mOAA7r38,242
-morphik/_internal.py,sha256=uKP2mP1G6thNYzxnAK_H8AlqPuEBQFH_8yK9XqVGuM4,18756
-morphik/async_.py,sha256=agF0NAE9P5_tcXpYodTZFMVt0jhYgoCuZ3ikuFsIUVI,96058
-morphik/exceptions.py,sha256=v4XGmfq5B0KrZEF6M1ID8A50-45-SRAQZTrXGXM6n0Q,260
-morphik/models.py,sha256=fj6o9K1zMualdivFL7Hef8mS4WRSLJu04bCCNvkZabw,20873
-morphik/rules.py,sha256=fw0RovS0Pwtff8Dvo3nkM3Wl6WtR3ykSaxsU_sxdXKI,2565
-morphik/sync.py,sha256=LeUsuf3BlGhRlB4jvdOtddF1T4Hkzo5r-qBM9UV1a6Y,101116
-morphik/tests/README.md,sha256=jtJDDK8cS5E4SbygFQDy7t6Y-kQwNYtZajRwVJDR62U,1069
-morphik/tests/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-morphik/tests/example_usage.py,sha256=ls8n7355q-8gY43pZLKd4SzI-01MdFeXbT8bZ4U8MCg,11561
-morphik/tests/test_async.py,sha256=M-gET1TD_glTMYZl0DZjRFCb3Qe8n8hRAmQARwbGU-A,13902
-morphik/tests/test_sync.py,sha256=uEKByKr0woCRKuZrI6Ovz-JVrK7KfZHLSJT7zqj1d7U,13481
-morphik/tests/test_docs/sample1.txt,sha256=Fx6TElSiKdxyFeBp1iHthzHctFVZm38DrqcbdZMoidY,507
-morphik/tests/test_docs/sample2.txt,sha256=PE97gPv59J27A7CSNvi_0tRBIN3Mj6pyTFElCLfs3TE,686
-morphik/tests/test_docs/sample3.txt,sha256=OzrnJ_XsDUntEV0jk-ansa3_KIa6GnpvS5EVmlh6BHo,732
-morphik-0.1.6.dist-info/METADATA,sha256=QjimRoP10I1tZbd5AGIZ-Wixe1osyDho5jLILhVT1mc,3377
-morphik-0.1.6.dist-info/WHEEL,sha256=qtCwoSJWgHk21S1Kb4ihdzI2rlJ1ZKaIurTj_ngOhyQ,87
-morphik-0.1.6.dist-info/RECORD,,