llama-cloud 0.0.7__py3-none-any.whl → 0.0.9__py3-none-any.whl

llama_cloud/resources/pipelines/client.py

@@ -9,6 +9,7 @@ from ...core.client_wrapper import AsyncClientWrapper, SyncClientWrapper
 from ...core.jsonable_encoder import jsonable_encoder
 from ...core.remove_none_from_dict import remove_none_from_dict
 from ...errors.unprocessable_entity_error import UnprocessableEntityError
+from ...types.chat_params import ChatParams
 from ...types.cloud_document import CloudDocument
 from ...types.cloud_document_create import CloudDocumentCreate
 from ...types.configured_transformation_item import ConfiguredTransformationItem
@@ -30,6 +31,7 @@ from ...types.pipeline_file import PipelineFile
 from ...types.pipeline_file_create import PipelineFileCreate
 from ...types.pipeline_type import PipelineType
 from ...types.preset_retrieval_params import PresetRetrievalParams
+from ...types.retrieval_mode import RetrievalMode
 from ...types.retrieve_results import RetrieveResults
 from .types.pipeline_file_update_custom_metadata_value import PipelineFileUpdateCustomMetadataValue
 
@@ -52,19 +54,22 @@ class PipelinesClient:
     def search_pipelines(
         self,
         *,
-        project_name: str,
+        project_name: typing.Optional[str] = None,
         pipeline_name: typing.Optional[str] = None,
         pipeline_type: typing.Optional[PipelineType] = None,
+        project_id: typing.Optional[str] = None,
     ) -> typing.List[Pipeline]:
         """
         Search for pipelines by various parameters.
 
         Parameters:
-            - project_name: str.
+            - project_name: typing.Optional[str].
 
             - pipeline_name: typing.Optional[str].
 
             - pipeline_type: typing.Optional[PipelineType].
+
+            - project_id: typing.Optional[str].
         ---
         from llama_cloud import PipelineType
         from llama_cloud.client import LlamaCloud
@@ -73,7 +78,6 @@ class PipelinesClient:
             token="YOUR_TOKEN",
         )
         client.pipelines.search_pipelines(
-            project_name="string",
             pipeline_type=PipelineType.PLAYGROUND,
         )
         """
@@ -81,7 +85,12 @@ class PipelinesClient:
             "GET",
             urllib.parse.urljoin(f"{self._client_wrapper.get_base_url()}/", "api/v1/pipelines"),
             params=remove_none_from_dict(
-                {"project_name": project_name, "pipeline_name": pipeline_name, "pipeline_type": pipeline_type}
+                {
+                    "project_name": project_name,
+                    "pipeline_name": pipeline_name,
+                    "pipeline_type": pipeline_type,
+                    "project_id": project_id,
+                }
             ),
             headers=self._client_wrapper.get_headers(),
             timeout=60,
@@ -115,6 +124,7 @@ class PipelinesClient:
             PipelineCreate,
             PipelineType,
             PresetRetrievalParams,
+            RetrievalMode,
             SupportedEvalLlmModelNames,
         )
         from llama_cloud.client import LlamaCloud
@@ -133,6 +143,7 @@ class PipelinesClient:
                         filters=[],
                         condition=FilterCondition.AND,
                     ),
+                    retrieval_mode=RetrievalMode.CHUNKS,
                 ),
                 eval_parameters=EvalExecutionParams(
                     llm_model=SupportedEvalLlmModelNames.GPT_3_5_TURBO,
@@ -181,6 +192,7 @@ class PipelinesClient:
             PipelineCreate,
             PipelineType,
             PresetRetrievalParams,
+            RetrievalMode,
             SupportedEvalLlmModelNames,
         )
         from llama_cloud.client import LlamaCloud
@@ -199,6 +211,7 @@ class PipelinesClient:
                         filters=[],
                         condition=FilterCondition.AND,
                     ),
+                    retrieval_mode=RetrievalMode.CHUNKS,
                 ),
                 eval_parameters=EvalExecutionParams(
                     llm_model=SupportedEvalLlmModelNames.GPT_3_5_TURBO,
@@ -305,6 +318,7 @@ class PipelinesClient:
             LlamaParseParameters,
             MetadataFilters,
             PresetRetrievalParams,
+            RetrievalMode,
             SupportedEvalLlmModelNames,
         )
         from llama_cloud.client import LlamaCloud
@@ -323,6 +337,7 @@ class PipelinesClient:
                     filters=[],
                     condition=FilterCondition.AND,
                 ),
+                retrieval_mode=RetrievalMode.CHUNKS,
             ),
             eval_parameters=EvalExecutionParams(
                 llm_model=SupportedEvalLlmModelNames.GPT_3_5_TURBO,
@@ -644,12 +659,22 @@ class PipelinesClient:
             raise ApiError(status_code=_response.status_code, body=_response.text)
         raise ApiError(status_code=_response.status_code, body=_response_json)
 
-    def list_pipeline_files(self, pipeline_id: str) -> typing.List[PipelineFile]:
+    def list_pipeline_files(
+        self,
+        pipeline_id: str,
+        *,
+        data_source_id: typing.Optional[str] = None,
+        only_manually_uploaded: typing.Optional[bool] = None,
+    ) -> typing.List[PipelineFile]:
         """
         Get files for a pipeline.
 
         Parameters:
             - pipeline_id: str.
+
+            - data_source_id: typing.Optional[str].
+
+            - only_manually_uploaded: typing.Optional[bool].
         ---
         from llama_cloud.client import LlamaCloud
 
@@ -663,6 +688,9 @@ class PipelinesClient:
         _response = self._client_wrapper.httpx_client.request(
             "GET",
             urllib.parse.urljoin(f"{self._client_wrapper.get_base_url()}/", f"api/v1/pipelines/{pipeline_id}/files"),
+            params=remove_none_from_dict(
+                {"data_source_id": data_source_id, "only_manually_uploaded": only_manually_uploaded}
+            ),
             headers=self._client_wrapper.get_headers(),
             timeout=60,
         )
@@ -837,6 +865,65 @@ class PipelinesClient:
             raise ApiError(status_code=_response.status_code, body=_response.text)
         raise ApiError(status_code=_response.status_code, body=_response_json)
 
+    def import_pipeline_metadata(self, pipeline_id: str, *, upload_file: typing.IO) -> typing.Dict[str, str]:
+        """
+        Import metadata for a pipeline.
+
+        Parameters:
+            - pipeline_id: str.
+
+            - upload_file: typing.IO.
+        """
+        _response = self._client_wrapper.httpx_client.request(
+            "PUT",
+            urllib.parse.urljoin(f"{self._client_wrapper.get_base_url()}/", f"api/v1/pipelines/{pipeline_id}/metadata"),
+            data=jsonable_encoder({}),
+            files={"upload_file": upload_file},
+            headers=self._client_wrapper.get_headers(),
+            timeout=60,
+        )
+        if 200 <= _response.status_code < 300:
+            return pydantic.parse_obj_as(typing.Dict[str, str], _response.json())  # type: ignore
+        if _response.status_code == 422:
+            raise UnprocessableEntityError(pydantic.parse_obj_as(HttpValidationError, _response.json()))  # type: ignore
+        try:
+            _response_json = _response.json()
+        except JSONDecodeError:
+            raise ApiError(status_code=_response.status_code, body=_response.text)
+        raise ApiError(status_code=_response.status_code, body=_response_json)
+
+    def delete_pipeline_files_metadata(self, pipeline_id: str) -> None:
+        """
+        Delete metadata for all files in a pipeline.
+
+        Parameters:
+            - pipeline_id: str.
+        ---
+        from llama_cloud.client import LlamaCloud
+
+        client = LlamaCloud(
+            token="YOUR_TOKEN",
+        )
+        client.pipelines.delete_pipeline_files_metadata(
+            pipeline_id="string",
+        )
+        """
+        _response = self._client_wrapper.httpx_client.request(
+            "DELETE",
+            urllib.parse.urljoin(f"{self._client_wrapper.get_base_url()}/", f"api/v1/pipelines/{pipeline_id}/metadata"),
+            headers=self._client_wrapper.get_headers(),
+            timeout=60,
+        )
+        if 200 <= _response.status_code < 300:
+            return
+        if _response.status_code == 422:
+            raise UnprocessableEntityError(pydantic.parse_obj_as(HttpValidationError, _response.json()))  # type: ignore
+        try:
+            _response_json = _response.json()
+        except JSONDecodeError:
+            raise ApiError(status_code=_response.status_code, body=_response.text)
+        raise ApiError(status_code=_response.status_code, body=_response_json)
+
     def list_pipeline_data_sources(self, pipeline_id: str) -> typing.List[PipelineDataSource]:
         """
         Get data sources for a pipeline.
@@ -997,6 +1084,8 @@ class PipelinesClient:
         rerank_top_n: typing.Optional[int] = OMIT,
         alpha: typing.Optional[float] = OMIT,
         search_filters: typing.Optional[MetadataFilters] = OMIT,
+        files_top_k: typing.Optional[int] = OMIT,
+        retrieval_mode: typing.Optional[RetrievalMode] = OMIT,
         query: str,
     ) -> RetrieveResults:
         """
@@ -1017,9 +1106,13 @@ class PipelinesClient:
 
             - search_filters: typing.Optional[MetadataFilters]. Search filters for retrieval.
 
+            - files_top_k: typing.Optional[int]. Number of files to retrieve (only for retrieval mode files_via_metadata and files_via_content).
+
+            - retrieval_mode: typing.Optional[RetrievalMode]. The retrieval mode for the query.
+
             - query: str. The query to retrieve against.
         ---
-        from llama_cloud import FilterCondition, MetadataFilters
+        from llama_cloud import FilterCondition, MetadataFilters, RetrievalMode
         from llama_cloud.client import LlamaCloud
 
         client = LlamaCloud(
@@ -1031,6 +1124,7 @@ class PipelinesClient:
                 filters=[],
                 condition=FilterCondition.AND,
             ),
+            retrieval_mode=RetrievalMode.CHUNKS,
             query="string",
         )
         """
@@ -1047,6 +1141,10 @@ class PipelinesClient:
             _request["alpha"] = alpha
         if search_filters is not OMIT:
             _request["search_filters"] = search_filters
+        if files_top_k is not OMIT:
+            _request["files_top_k"] = files_top_k
+        if retrieval_mode is not OMIT:
+            _request["retrieval_mode"] = retrieval_mode
         _response = self._client_wrapper.httpx_client.request(
             "POST",
             urllib.parse.urljoin(f"{self._client_wrapper.get_base_url()}/", f"api/v1/pipelines/{pipeline_id}/retrieve"),
@@ -1133,6 +1231,58 @@ class PipelinesClient:
             raise ApiError(status_code=_response.status_code, body=_response.text)
         raise ApiError(status_code=_response.status_code, body=_response_json)
 
+    def chat(self, pipeline_id: str, *, retrieval_parameters: PresetRetrievalParams, data: ChatParams) -> typing.Any:
+        """
+        Parameters:
+            - pipeline_id: str.
+
+            - retrieval_parameters: PresetRetrievalParams.
+
+            - data: ChatParams.
+        ---
+        from llama_cloud import (
+            ChatParams,
+            FilterCondition,
+            MetadataFilters,
+            PresetRetrievalParams,
+            RetrievalMode,
+        )
+        from llama_cloud.client import LlamaCloud
+
+        client = LlamaCloud(
+            token="YOUR_TOKEN",
+        )
+        client.pipelines.chat(
+            pipeline_id="string",
+            retrieval_parameters=PresetRetrievalParams(
+                search_filters=MetadataFilters(
+                    filters=[],
+                    condition=FilterCondition.AND,
+                ),
+                retrieval_mode=RetrievalMode.CHUNKS,
+            ),
+            data=ChatParams(
+                messages=[],
+            ),
+        )
+        """
+        _response = self._client_wrapper.httpx_client.request(
+            "POST",
+            urllib.parse.urljoin(f"{self._client_wrapper.get_base_url()}/", f"api/v1/pipelines/{pipeline_id}/chat"),
+            json=jsonable_encoder({"retrieval_parameters": retrieval_parameters, "data": data}),
+            headers=self._client_wrapper.get_headers(),
+            timeout=60,
+        )
+        if 200 <= _response.status_code < 300:
+            return pydantic.parse_obj_as(typing.Any, _response.json())  # type: ignore
+        if _response.status_code == 422:
+            raise UnprocessableEntityError(pydantic.parse_obj_as(HttpValidationError, _response.json()))  # type: ignore
+        try:
+            _response_json = _response.json()
+        except JSONDecodeError:
+            raise ApiError(status_code=_response.status_code, body=_response.text)
+        raise ApiError(status_code=_response.status_code, body=_response_json)
+
     def list_pipeline_documents(
         self, pipeline_id: str, *, skip: typing.Optional[int] = None, limit: typing.Optional[int] = None
     ) -> typing.List[CloudDocument]:
@@ -1374,19 +1524,22 @@ class AsyncPipelinesClient:
     async def search_pipelines(
         self,
         *,
-        project_name: str,
+        project_name: typing.Optional[str] = None,
         pipeline_name: typing.Optional[str] = None,
         pipeline_type: typing.Optional[PipelineType] = None,
+        project_id: typing.Optional[str] = None,
     ) -> typing.List[Pipeline]:
         """
         Search for pipelines by various parameters.
 
         Parameters:
-            - project_name: str.
+            - project_name: typing.Optional[str].
 
             - pipeline_name: typing.Optional[str].
 
             - pipeline_type: typing.Optional[PipelineType].
+
+            - project_id: typing.Optional[str].
         ---
         from llama_cloud import PipelineType
         from llama_cloud.client import AsyncLlamaCloud
@@ -1395,7 +1548,6 @@ class AsyncPipelinesClient:
             token="YOUR_TOKEN",
         )
         await client.pipelines.search_pipelines(
-            project_name="string",
             pipeline_type=PipelineType.PLAYGROUND,
         )
         """
@@ -1403,7 +1555,12 @@ class AsyncPipelinesClient:
             "GET",
             urllib.parse.urljoin(f"{self._client_wrapper.get_base_url()}/", "api/v1/pipelines"),
             params=remove_none_from_dict(
-                {"project_name": project_name, "pipeline_name": pipeline_name, "pipeline_type": pipeline_type}
+                {
+                    "project_name": project_name,
+                    "pipeline_name": pipeline_name,
+                    "pipeline_type": pipeline_type,
+                    "project_id": project_id,
+                }
             ),
             headers=self._client_wrapper.get_headers(),
             timeout=60,
@@ -1437,6 +1594,7 @@ class AsyncPipelinesClient:
             PipelineCreate,
             PipelineType,
             PresetRetrievalParams,
+            RetrievalMode,
             SupportedEvalLlmModelNames,
         )
         from llama_cloud.client import AsyncLlamaCloud
@@ -1455,6 +1613,7 @@ class AsyncPipelinesClient:
                         filters=[],
                         condition=FilterCondition.AND,
                     ),
+                    retrieval_mode=RetrievalMode.CHUNKS,
                 ),
                 eval_parameters=EvalExecutionParams(
                     llm_model=SupportedEvalLlmModelNames.GPT_3_5_TURBO,
@@ -1503,6 +1662,7 @@ class AsyncPipelinesClient:
             PipelineCreate,
             PipelineType,
             PresetRetrievalParams,
+            RetrievalMode,
             SupportedEvalLlmModelNames,
         )
         from llama_cloud.client import AsyncLlamaCloud
@@ -1521,6 +1681,7 @@ class AsyncPipelinesClient:
                         filters=[],
                         condition=FilterCondition.AND,
                     ),
+                    retrieval_mode=RetrievalMode.CHUNKS,
                 ),
                 eval_parameters=EvalExecutionParams(
                     llm_model=SupportedEvalLlmModelNames.GPT_3_5_TURBO,
@@ -1627,6 +1788,7 @@ class AsyncPipelinesClient:
             LlamaParseParameters,
             MetadataFilters,
             PresetRetrievalParams,
+            RetrievalMode,
             SupportedEvalLlmModelNames,
         )
         from llama_cloud.client import AsyncLlamaCloud
@@ -1645,6 +1807,7 @@ class AsyncPipelinesClient:
                     filters=[],
                     condition=FilterCondition.AND,
                 ),
+                retrieval_mode=RetrievalMode.CHUNKS,
             ),
             eval_parameters=EvalExecutionParams(
                 llm_model=SupportedEvalLlmModelNames.GPT_3_5_TURBO,
@@ -1968,12 +2131,22 @@ class AsyncPipelinesClient:
             raise ApiError(status_code=_response.status_code, body=_response.text)
         raise ApiError(status_code=_response.status_code, body=_response_json)
 
-    async def list_pipeline_files(self, pipeline_id: str) -> typing.List[PipelineFile]:
+    async def list_pipeline_files(
+        self,
+        pipeline_id: str,
+        *,
+        data_source_id: typing.Optional[str] = None,
+        only_manually_uploaded: typing.Optional[bool] = None,
+    ) -> typing.List[PipelineFile]:
         """
         Get files for a pipeline.
 
         Parameters:
             - pipeline_id: str.
+
+            - data_source_id: typing.Optional[str].
+
+            - only_manually_uploaded: typing.Optional[bool].
         ---
         from llama_cloud.client import AsyncLlamaCloud
 
@@ -1987,6 +2160,9 @@ class AsyncPipelinesClient:
         _response = await self._client_wrapper.httpx_client.request(
             "GET",
             urllib.parse.urljoin(f"{self._client_wrapper.get_base_url()}/", f"api/v1/pipelines/{pipeline_id}/files"),
+            params=remove_none_from_dict(
+                {"data_source_id": data_source_id, "only_manually_uploaded": only_manually_uploaded}
+            ),
             headers=self._client_wrapper.get_headers(),
             timeout=60,
         )
@@ -2161,6 +2337,65 @@ class AsyncPipelinesClient:
             raise ApiError(status_code=_response.status_code, body=_response.text)
         raise ApiError(status_code=_response.status_code, body=_response_json)
 
+    async def import_pipeline_metadata(self, pipeline_id: str, *, upload_file: typing.IO) -> typing.Dict[str, str]:
+        """
+        Import metadata for a pipeline.
+
+        Parameters:
+            - pipeline_id: str.
+
+            - upload_file: typing.IO.
+        """
+        _response = await self._client_wrapper.httpx_client.request(
+            "PUT",
+            urllib.parse.urljoin(f"{self._client_wrapper.get_base_url()}/", f"api/v1/pipelines/{pipeline_id}/metadata"),
+            data=jsonable_encoder({}),
+            files={"upload_file": upload_file},
+            headers=self._client_wrapper.get_headers(),
+            timeout=60,
+        )
+        if 200 <= _response.status_code < 300:
+            return pydantic.parse_obj_as(typing.Dict[str, str], _response.json())  # type: ignore
+        if _response.status_code == 422:
+            raise UnprocessableEntityError(pydantic.parse_obj_as(HttpValidationError, _response.json()))  # type: ignore
+        try:
+            _response_json = _response.json()
+        except JSONDecodeError:
+            raise ApiError(status_code=_response.status_code, body=_response.text)
+        raise ApiError(status_code=_response.status_code, body=_response_json)
+
+    async def delete_pipeline_files_metadata(self, pipeline_id: str) -> None:
+        """
+        Delete metadata for all files in a pipeline.
+
+        Parameters:
+            - pipeline_id: str.
+        ---
+        from llama_cloud.client import AsyncLlamaCloud
+
+        client = AsyncLlamaCloud(
+            token="YOUR_TOKEN",
+        )
+        await client.pipelines.delete_pipeline_files_metadata(
+            pipeline_id="string",
+        )
+        """
+        _response = await self._client_wrapper.httpx_client.request(
+            "DELETE",
+            urllib.parse.urljoin(f"{self._client_wrapper.get_base_url()}/", f"api/v1/pipelines/{pipeline_id}/metadata"),
+            headers=self._client_wrapper.get_headers(),
+            timeout=60,
+        )
+        if 200 <= _response.status_code < 300:
+            return
+        if _response.status_code == 422:
+            raise UnprocessableEntityError(pydantic.parse_obj_as(HttpValidationError, _response.json()))  # type: ignore
+        try:
+            _response_json = _response.json()
+        except JSONDecodeError:
+            raise ApiError(status_code=_response.status_code, body=_response.text)
+        raise ApiError(status_code=_response.status_code, body=_response_json)
+
     async def list_pipeline_data_sources(self, pipeline_id: str) -> typing.List[PipelineDataSource]:
         """
         Get data sources for a pipeline.
@@ -2321,6 +2556,8 @@ class AsyncPipelinesClient:
         rerank_top_n: typing.Optional[int] = OMIT,
         alpha: typing.Optional[float] = OMIT,
         search_filters: typing.Optional[MetadataFilters] = OMIT,
+        files_top_k: typing.Optional[int] = OMIT,
+        retrieval_mode: typing.Optional[RetrievalMode] = OMIT,
         query: str,
     ) -> RetrieveResults:
         """
@@ -2341,9 +2578,13 @@ class AsyncPipelinesClient:
 
             - search_filters: typing.Optional[MetadataFilters]. Search filters for retrieval.
 
+            - files_top_k: typing.Optional[int]. Number of files to retrieve (only for retrieval mode files_via_metadata and files_via_content).
+
+            - retrieval_mode: typing.Optional[RetrievalMode]. The retrieval mode for the query.
+
             - query: str. The query to retrieve against.
         ---
-        from llama_cloud import FilterCondition, MetadataFilters
+        from llama_cloud import FilterCondition, MetadataFilters, RetrievalMode
         from llama_cloud.client import AsyncLlamaCloud
 
         client = AsyncLlamaCloud(
@@ -2355,6 +2596,7 @@ class AsyncPipelinesClient:
                 filters=[],
                 condition=FilterCondition.AND,
             ),
+            retrieval_mode=RetrievalMode.CHUNKS,
             query="string",
         )
         """
@@ -2371,6 +2613,10 @@ class AsyncPipelinesClient:
             _request["alpha"] = alpha
         if search_filters is not OMIT:
             _request["search_filters"] = search_filters
+        if files_top_k is not OMIT:
+            _request["files_top_k"] = files_top_k
+        if retrieval_mode is not OMIT:
+            _request["retrieval_mode"] = retrieval_mode
         _response = await self._client_wrapper.httpx_client.request(
             "POST",
             urllib.parse.urljoin(f"{self._client_wrapper.get_base_url()}/", f"api/v1/pipelines/{pipeline_id}/retrieve"),
@@ -2457,6 +2703,60 @@ class AsyncPipelinesClient:
             raise ApiError(status_code=_response.status_code, body=_response.text)
         raise ApiError(status_code=_response.status_code, body=_response_json)
 
+    async def chat(
+        self, pipeline_id: str, *, retrieval_parameters: PresetRetrievalParams, data: ChatParams
+    ) -> typing.Any:
+        """
+        Parameters:
+            - pipeline_id: str.
+
+            - retrieval_parameters: PresetRetrievalParams.
+
+            - data: ChatParams.
+        ---
+        from llama_cloud import (
+            ChatParams,
+            FilterCondition,
+            MetadataFilters,
+            PresetRetrievalParams,
+            RetrievalMode,
+        )
+        from llama_cloud.client import AsyncLlamaCloud
+
+        client = AsyncLlamaCloud(
+            token="YOUR_TOKEN",
+        )
+        await client.pipelines.chat(
+            pipeline_id="string",
+            retrieval_parameters=PresetRetrievalParams(
+                search_filters=MetadataFilters(
+                    filters=[],
+                    condition=FilterCondition.AND,
+                ),
+                retrieval_mode=RetrievalMode.CHUNKS,
+            ),
+            data=ChatParams(
+                messages=[],
+            ),
+        )
+        """
+        _response = await self._client_wrapper.httpx_client.request(
+            "POST",
+            urllib.parse.urljoin(f"{self._client_wrapper.get_base_url()}/", f"api/v1/pipelines/{pipeline_id}/chat"),
+            json=jsonable_encoder({"retrieval_parameters": retrieval_parameters, "data": data}),
+            headers=self._client_wrapper.get_headers(),
+            timeout=60,
+        )
+        if 200 <= _response.status_code < 300:
+            return pydantic.parse_obj_as(typing.Any, _response.json())  # type: ignore
+        if _response.status_code == 422:
+            raise UnprocessableEntityError(pydantic.parse_obj_as(HttpValidationError, _response.json()))  # type: ignore
+        try:
+            _response_json = _response.json()
+        except JSONDecodeError:
+            raise ApiError(status_code=_response.status_code, body=_response.text)
+        raise ApiError(status_code=_response.status_code, body=_response_json)
+
     async def list_pipeline_documents(
         self, pipeline_id: str, *, skip: typing.Optional[int] = None, limit: typing.Optional[int] = None
     ) -> typing.List[CloudDocument]: