unstructured-ingest 0.0.22__py3-none-any.whl → 0.0.23__py3-none-any.whl

unstructured_ingest/__version__.py

@@ -1 +1 @@
-__version__ = "0.0.22"  # pragma: no cover
+__version__ = "0.0.23"  # pragma: no cover

unstructured_ingest/cli/interfaces.py

@@ -415,11 +415,11 @@ class CliEmbeddingConfig(EmbeddingConfig, CliMixin):
     @staticmethod
     def get_cli_options() -> t.List[click.Option]:
         embed_providers = [
-            "langchain-openai",
-            "langchain-huggingface",
-            "langchain-aws-bedrock",
-            "langchain-vertexai",
-            "langchain-voyageai",
+            "openai",
+            "huggingface",
+            "aws-bedrock",
+            "vertexai",
+            "voyageai",
             "octoai",
         ]
         options = [

unstructured_ingest/embed/__init__.py

@@ -1,17 +0,0 @@
-from unstructured_ingest.embed.bedrock import BedrockEmbeddingEncoder
-from unstructured_ingest.embed.huggingface import HuggingFaceEmbeddingEncoder
-from unstructured_ingest.embed.mixedbreadai import MixedbreadAIEmbeddingEncoder
-from unstructured_ingest.embed.octoai import OctoAIEmbeddingEncoder
-from unstructured_ingest.embed.openai import OpenAIEmbeddingEncoder
-from unstructured_ingest.embed.vertexai import VertexAIEmbeddingEncoder
-from unstructured_ingest.embed.voyageai import VoyageAIEmbeddingEncoder
-
-EMBEDDING_PROVIDER_TO_CLASS_MAP = {
-    "langchain-openai": OpenAIEmbeddingEncoder,
-    "langchain-huggingface": HuggingFaceEmbeddingEncoder,
-    "langchain-aws-bedrock": BedrockEmbeddingEncoder,
-    "langchain-vertexai": VertexAIEmbeddingEncoder,
-    "langchain-voyageai": VoyageAIEmbeddingEncoder,
-    "mixedbread-ai": MixedbreadAIEmbeddingEncoder,
-    "octoai": OctoAIEmbeddingEncoder,
-}

unstructured_ingest/embed/vertexai.py

@@ -45,7 +45,7 @@ class VertexAIEmbeddingConfig(EmbeddingConfig):
         extras="embed-vertexai",
     )
     def get_client(self) -> "TextEmbeddingModel":
-        """Creates a Langchain VertexAI python client to embed elements."""
+        """Creates a VertexAI python client to embed elements."""
         from vertexai.language_models import TextEmbeddingModel
 
         self.register_application_credentials()

unstructured_ingest/embed/voyageai.py

@@ -20,11 +20,11 @@ class VoyageAIEmbeddingConfig(EmbeddingConfig):
     timeout_in_seconds: Optional[int] = None
 
     @requires_dependencies(
-        ["langchain", "langchain_voyageai"],
+        ["voyageai"],
         extras="embed-voyageai",
     )
     def get_client(self) -> "VoyageAIClient":
-        """Creates a Langchain VoyageAI python client to embed elements."""
+        """Creates a VoyageAI python client to embed elements."""
         from voyageai import Client as VoyageAIClient
 
         client = VoyageAIClient(

unstructured_ingest/interfaces.py

@@ -204,14 +204,14 @@ class EmbeddingConfig(BaseConfig):
         if self.model_name:
             kwargs["model_name"] = self.model_name
         # TODO make this more dynamic to map to encoder configs
-        if self.provider == "langchain-openai":
+        if self.provider == "openai":
             from unstructured_ingest.embed.openai import (
                 OpenAIEmbeddingConfig,
                 OpenAIEmbeddingEncoder,
             )
 
             return OpenAIEmbeddingEncoder(config=OpenAIEmbeddingConfig(**kwargs))
-        elif self.provider == "langchain-huggingface":
+        elif self.provider == "huggingface":
             from unstructured_ingest.embed.huggingface import (
                 HuggingFaceEmbeddingConfig,
                 HuggingFaceEmbeddingEncoder,
@@ -225,7 +225,7 @@ class EmbeddingConfig(BaseConfig):
             )
 
             return OctoAIEmbeddingEncoder(config=OctoAiEmbeddingConfig(**kwargs))
-        elif self.provider == "langchain-aws-bedrock":
+        elif self.provider == "aws-bedrock":
             from unstructured_ingest.embed.bedrock import (
                 BedrockEmbeddingConfig,
                 BedrockEmbeddingEncoder,
@@ -238,14 +238,14 @@ class EmbeddingConfig(BaseConfig):
                     region_name=self.aws_region,
                 )
             )
-        elif self.provider == "langchain-vertexai":
+        elif self.provider == "vertexai":
             from unstructured_ingest.embed.vertexai import (
                 VertexAIEmbeddingConfig,
                 VertexAIEmbeddingEncoder,
             )
 
             return VertexAIEmbeddingEncoder(config=VertexAIEmbeddingConfig(**kwargs))
-        elif self.provider == "langchain-voyageai":
+        elif self.provider == "voyageai":
             from unstructured_ingest.embed.voyageai import (
                 VoyageAIEmbeddingConfig,
                 VoyageAIEmbeddingEncoder,

unstructured_ingest/v2/processes/embedder.py

@@ -15,11 +15,11 @@ if TYPE_CHECKING:
 class EmbedderConfig(BaseModel):
     embedding_provider: Optional[
         Literal[
-            "langchain-openai",
-            "langchain-huggingface",
-            "langchain-aws-bedrock",
-            "langchain-vertexai",
-            "langchain-voyageai",
+            "openai",
+            "huggingface",
+            "aws-bedrock",
+            "vertexai",
+            "voyageai",
             "octoai",
             "mixedbread-ai",
         ]
@@ -114,22 +114,22 @@ class EmbedderConfig(BaseModel):
         if self.embedding_model_name:
             kwargs["model_name"] = self.embedding_model_name
         # TODO make this more dynamic to map to encoder configs
-        if self.embedding_provider == "langchain-openai":
+        if self.embedding_provider == "openai":
             return self.get_openai_embedder(embedding_kwargs=kwargs)
 
-        if self.embedding_provider == "langchain-huggingface":
+        if self.embedding_provider == "huggingface":
             return self.get_huggingface_embedder(embedding_kwargs=kwargs)
 
         if self.embedding_provider == "octoai":
             return self.get_octoai_embedder(embedding_kwargs=kwargs)
 
-        if self.embedding_provider == "langchain-aws-bedrock":
+        if self.embedding_provider == "aws-bedrock":
             return self.get_bedrock_embedder()
 
-        if self.embedding_provider == "langchain-vertexai":
+        if self.embedding_provider == "vertexai":
             return self.get_vertexai_embedder(embedding_kwargs=kwargs)
 
-        if self.embedding_provider == "langchain-voyageai":
+        if self.embedding_provider == "voyageai":
             return self.get_voyageai_embedder(embedding_kwargs=kwargs)
         if self.embedding_provider == "mixedbread-ai":
             return self.get_mixedbread_embedder(embedding_kwargs=kwargs)

{unstructured_ingest-0.0.22.dist-info → unstructured_ingest-0.0.23.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: unstructured-ingest
-Version: 0.0.22
+Version: 0.0.23
 Summary: A library that prepares raw documents for downstream ML tasks.
 Home-page: https://github.com/Unstructured-IO/unstructured-ingest
 Author: Unstructured Technologies
@@ -22,20 +22,20 @@ Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
 Requires-Python: >=3.9.0,<3.13
 Description-Content-Type: text/markdown
 License-File: LICENSE.md
-Requires-Dist: pandas
-Requires-Dist: dataclasses-json
-Requires-Dist: opentelemetry-sdk
+Requires-Dist: pydantic>=2.7
 Requires-Dist: python-dateutil
 Requires-Dist: click
+Requires-Dist: opentelemetry-sdk
+Requires-Dist: pandas
+Requires-Dist: dataclasses-json
 Requires-Dist: tqdm
-Requires-Dist: pydantic>=2.7
 Provides-Extra: airtable
 Requires-Dist: pyairtable; extra == "airtable"
 Provides-Extra: astradb
 Requires-Dist: astrapy; extra == "astradb"
 Provides-Extra: azure
-Requires-Dist: adlfs; extra == "azure"
 Requires-Dist: fsspec; extra == "azure"
+Requires-Dist: adlfs; extra == "azure"
 Provides-Extra: azure-cognitive-search
 Requires-Dist: azure-search-documents; extra == "azure-cognitive-search"
 Provides-Extra: bedrock
@@ -44,8 +44,8 @@ Provides-Extra: biomed
 Requires-Dist: bs4; extra == "biomed"
 Requires-Dist: requests; extra == "biomed"
 Provides-Extra: box
-Requires-Dist: boxfs; extra == "box"
 Requires-Dist: fsspec; extra == "box"
+Requires-Dist: boxfs; extra == "box"
 Provides-Extra: chroma
 Requires-Dist: chromadb; extra == "chroma"
 Provides-Extra: clarifai
@@ -60,8 +60,8 @@ Requires-Dist: unstructured[tsv]; extra == "csv"
 Provides-Extra: databricks-volumes
 Requires-Dist: databricks-sdk; extra == "databricks-volumes"
 Provides-Extra: delta-table
-Requires-Dist: deltalake; extra == "delta-table"
 Requires-Dist: fsspec; extra == "delta-table"
+Requires-Dist: deltalake; extra == "delta-table"
 Provides-Extra: discord
 Requires-Dist: discord-py; extra == "discord"
 Provides-Extra: doc
@@ -69,8 +69,8 @@ Requires-Dist: unstructured[docx]; extra == "doc"
 Provides-Extra: docx
 Requires-Dist: unstructured[docx]; extra == "docx"
 Provides-Extra: dropbox
-Requires-Dist: dropboxdrivefs; extra == "dropbox"
 Requires-Dist: fsspec; extra == "dropbox"
+Requires-Dist: dropboxdrivefs; extra == "dropbox"
 Provides-Extra: elasticsearch
 Requires-Dist: elasticsearch[async]; extra == "elasticsearch"
 Provides-Extra: embed-huggingface
@@ -88,18 +88,18 @@ Provides-Extra: epub
 Requires-Dist: unstructured[epub]; extra == "epub"
 Provides-Extra: gcs
 Requires-Dist: bs4; extra == "gcs"
-Requires-Dist: gcsfs; extra == "gcs"
 Requires-Dist: fsspec; extra == "gcs"
+Requires-Dist: gcsfs; extra == "gcs"
 Provides-Extra: github
-Requires-Dist: pygithub>1.58.0; extra == "github"
 Requires-Dist: requests; extra == "github"
+Requires-Dist: pygithub>1.58.0; extra == "github"
 Provides-Extra: gitlab
 Requires-Dist: python-gitlab; extra == "gitlab"
 Provides-Extra: google-drive
 Requires-Dist: google-api-python-client; extra == "google-drive"
 Provides-Extra: hubspot
-Requires-Dist: hubspot-api-client; extra == "hubspot"
 Requires-Dist: urllib3; extra == "hubspot"
+Requires-Dist: hubspot-api-client; extra == "hubspot"
 Provides-Extra: jira
 Requires-Dist: atlassian-python-api; extra == "jira"
 Provides-Extra: kafka
@@ -115,16 +115,16 @@ Requires-Dist: pymongo; extra == "mongodb"
 Provides-Extra: msg
 Requires-Dist: unstructured[msg]; extra == "msg"
 Provides-Extra: notion
+Requires-Dist: backoff; extra == "notion"
 Requires-Dist: httpx; extra == "notion"
-Requires-Dist: notion-client; extra == "notion"
 Requires-Dist: htmlBuilder; extra == "notion"
-Requires-Dist: backoff; extra == "notion"
+Requires-Dist: notion-client; extra == "notion"
 Provides-Extra: odt
 Requires-Dist: unstructured[odt]; extra == "odt"
 Provides-Extra: onedrive
+Requires-Dist: bs4; extra == "onedrive"
 Requires-Dist: msal; extra == "onedrive"
 Requires-Dist: Office365-REST-Python-Client; extra == "onedrive"
-Requires-Dist: bs4; extra == "onedrive"
 Provides-Extra: openai
 Requires-Dist: openai; extra == "openai"
 Requires-Dist: tiktoken; extra == "openai"
@@ -156,13 +156,13 @@ Requires-Dist: unstructured[rst]; extra == "rst"
 Provides-Extra: rtf
 Requires-Dist: unstructured[rtf]; extra == "rtf"
 Provides-Extra: s3
-Requires-Dist: s3fs; extra == "s3"
 Requires-Dist: fsspec; extra == "s3"
+Requires-Dist: s3fs; extra == "s3"
 Provides-Extra: salesforce
 Requires-Dist: simple-salesforce; extra == "salesforce"
 Provides-Extra: sftp
-Requires-Dist: fsspec; extra == "sftp"
 Requires-Dist: paramiko; extra == "sftp"
+Requires-Dist: fsspec; extra == "sftp"
 Provides-Extra: sharepoint
 Requires-Dist: msal; extra == "sharepoint"
 Requires-Dist: Office365-REST-Python-Client; extra == "sharepoint"

{unstructured_ingest-0.0.22.dist-info → unstructured_ingest-0.0.23.dist-info}/RECORD

@@ -1,7 +1,7 @@
 unstructured_ingest/__init__.py,sha256=U4S_2y3zgLZVfMenHRaJFBW8yqh2mUBuI291LGQVOJ8,35
-unstructured_ingest/__version__.py,sha256=s86WG6OgfRK15ii3W6pb0AczHH8T1QB33x5r6twQpCY,43
+unstructured_ingest/__version__.py,sha256=HgbcmBIk6mQp0Bz81M53L-kPIBJnMYIFOGkRL73EChs,43
 unstructured_ingest/error.py,sha256=qDncnJgbf5ils956RcO2CGlAKYDT5OaEM9Clv1JVTNc,1448
-unstructured_ingest/interfaces.py,sha256=LuY-85clq5iyOtDd9vDABjK2MZCm6wRWK53cdb4SROY,31411
+unstructured_ingest/interfaces.py,sha256=0r0gQoHJQ4DVSQEVbUPBA3N6WyvGMkR1u6U2SwUvoAQ,31361
 unstructured_ingest/logger.py,sha256=S5nSqGcABoQyeicgRnBQFjDScCaTvFVivOCvbo-laL0,4479
 unstructured_ingest/main.py,sha256=82G_7eG4PNhc_xIqj4Y_sFbDV9VI-nwSfsfJQMzovMk,169
 unstructured_ingest/processor.py,sha256=XKKrvbxsb--5cDzz4hB3-GfWZYyIjJ2ah8FpzQKF_DM,2760
@@ -9,7 +9,7 @@ unstructured_ingest/cli/__init__.py,sha256=9kNcBOHuXON5lB1MJU9QewEhwPmId56vXqB29
 unstructured_ingest/cli/cli.py,sha256=LutBTBYMqboKw8cputHVszpenyfnySzcUC15ifwuYyg,1049
 unstructured_ingest/cli/cmd_factory.py,sha256=UdHm1KacTombpF6DxyTSwTCuApsKHUYw_kVu5Nhcy3Y,364
 unstructured_ingest/cli/common.py,sha256=I0El08FHz5kxw7iz0VWOWPrvcJD1rBgXJSwVIpVmmwU,204
-unstructured_ingest/cli/interfaces.py,sha256=6kMmTVm5ia6wUIdOMRu_uErkcIeWndr_6fhPc1AnoYM,24134
+unstructured_ingest/cli/interfaces.py,sha256=nWZVXAoLEP08eDPj10c2nwHNbd-HXOHFa4YvEdUJ8y8,24084
 unstructured_ingest/cli/utils.py,sha256=KNhkFNKOeEihc8HlvMz_MTbYVQNFklrBKbC8xg9h1xE,7982
 unstructured_ingest/cli/base/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 unstructured_ingest/cli/base/cmd.py,sha256=BbfjA2v203Jh-7DL6bzxQ7fOeNixd5BsBMuzXz6h5IQ,583
@@ -166,15 +166,15 @@ unstructured_ingest/connector/notion/types/database_properties/title.py,sha256=O
 unstructured_ingest/connector/notion/types/database_properties/unique_id.py,sha256=H9lKi8rCDPtKmuu7j9CnJoTUr6YmzIF4oXbv_OxuN9k,1162
 unstructured_ingest/connector/notion/types/database_properties/url.py,sha256=iXQ2tVUm9UlKVtDA0NQiFIRJ5PHYW9wOaWt2vFfSVCg,862
 unstructured_ingest/connector/notion/types/database_properties/verification.py,sha256=J_DLjY-v2T6xDGMQ7FkI0YMKMA6SG6Y3yYW7qUD1hKA,2334
-unstructured_ingest/embed/__init__.py,sha256=whnTiGsSbNqaLObr058CKG5iGxk5OkN_41IBEtHQYW8,900
+unstructured_ingest/embed/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 unstructured_ingest/embed/bedrock.py,sha256=5-pKWwOEGHKOHa06wYuKOhvT8Xu72ke6nrpCnRtkAaU,3872
 unstructured_ingest/embed/huggingface.py,sha256=ku_JQr72KBG8n5b6KRkXIbeBGzdgLw_KKIEm1dFK3oM,2729
 unstructured_ingest/embed/interfaces.py,sha256=L5WimR69bmEvliIBlZ8wOCH_YDA9DWteCu6QEsKCV5I,1113
 unstructured_ingest/embed/mixedbreadai.py,sha256=NSrAt1_bjphTHLUnlzzWSBU25UBCZlpYaLdWSRSGyqs,5504
 unstructured_ingest/embed/octoai.py,sha256=0zxAUAMzodGkqMwqMkEvSfgWLNHtEnhdvUofvJDQD1A,2368
 unstructured_ingest/embed/openai.py,sha256=4Ee4A2rQ8OlSh_yiJSFmok_qqRDi1A3KyayB5YiPLFw,2058
-unstructured_ingest/embed/vertexai.py,sha256=pclpjjacvGTex74tD_3yZKhR-X5BR0fAHbgZUE62uh0,3608
-unstructured_ingest/embed/voyageai.py,sha256=8l-EGVdOcR9jLoc0DV4aPOvJsvRnInI8w12C9jxiUWA,2664
+unstructured_ingest/embed/vertexai.py,sha256=cgyRyTm_dO_qyedwbIhOQIFvKjCqZBoDh606ykzTYHI,3598
+unstructured_ingest/embed/voyageai.py,sha256=6BWNJUZOqkHSMaO2XPVZVYAVRrAtpMWQZEKp0qgp20Q,2631
 unstructured_ingest/enhanced_dataclass/__init__.py,sha256=gDZOUsv5eo-8jm4Yu7DdDwi101aGbfG7JctTdOYnTOM,151
 unstructured_ingest/enhanced_dataclass/core.py,sha256=d6aUkDynuKX87cHx9_N5UDUWrvISR4jYRFRTvd_avlI,3038
 unstructured_ingest/enhanced_dataclass/dataclasses.py,sha256=aZMsoCzAGRb8Rmh3BTSBFtNr6FmFTY93KYGLk3gYJKQ,1949
@@ -304,7 +304,7 @@ unstructured_ingest/v2/pipeline/steps/upload.py,sha256=zlgXgwReX9TBOdfTpS9hETah4
 unstructured_ingest/v2/processes/__init__.py,sha256=FaHWSCGyc7GWVnAsNEUUj7L8hT8gCVY3_hUE2VzWtUg,462
 unstructured_ingest/v2/processes/chunker.py,sha256=76PrpCSd8k3DpfdZcl8I10u7vciKzhSV9ZByrrp302g,5476
 unstructured_ingest/v2/processes/connector_registry.py,sha256=vkEe6jpgdYtZCxMj59s5atWGgmPuxAEXRUoTt-MJ7wc,2198
-unstructured_ingest/v2/processes/embedder.py,sha256=ZBCIm0oHxWmtUEQYyAjXACqTYPt3LnvXLtoFhu6mu8A,6077
+unstructured_ingest/v2/processes/embedder.py,sha256=nFYiOmIJwWLodBt_cC-E5h7zmYB9t3hLu2BWtBStm3g,5977
 unstructured_ingest/v2/processes/filter.py,sha256=kjUmMw2SDq2bme0JCAOxs6cJriIG6Ty09KOznS-xz08,2145
 unstructured_ingest/v2/processes/partitioner.py,sha256=bpqmZDsKKi6qtxNWdIWBfQmr1ccQUhU0axecpGAUf_4,7739
 unstructured_ingest/v2/processes/uncompress.py,sha256=Z_XfsITGdyaRwhtNUc7bMj5Y2jLuBge8KoK4nxhqKag,2425
@@ -339,9 +339,9 @@ unstructured_ingest/v2/processes/connectors/fsspec/gcs.py,sha256=-_pYHbsBG9FyRyN
 unstructured_ingest/v2/processes/connectors/fsspec/s3.py,sha256=je1BDqFWlyMfPa4oAMMNFQLLQtCY9quuqx3xjTwF8OQ,6251
 unstructured_ingest/v2/processes/connectors/fsspec/sftp.py,sha256=dwpyqDq0qceCBWX3zM1hiUlgXB4hzX6ObOr-sh-5CJs,6926
 unstructured_ingest/v2/processes/connectors/fsspec/utils.py,sha256=jec_Qfe2hbfahBuY-u8FnvHuv933AI5HwPFjOL3kEEY,456
-unstructured_ingest-0.0.22.dist-info/LICENSE.md,sha256=SxkKP_62uIAKb9mb1eH7FH4Kn2aYT09fgjKpJt5PyTk,11360
-unstructured_ingest-0.0.22.dist-info/METADATA,sha256=JteQ0j_D7rNBM7XNdb3dSgJMx26-OV8mLUVqubvhSMY,7108
-unstructured_ingest-0.0.22.dist-info/WHEEL,sha256=GV9aMThwP_4oNCtvEC2ec3qUYutgWeAzklro_0m4WJQ,91
-unstructured_ingest-0.0.22.dist-info/entry_points.txt,sha256=gUAAFnjFPnBgThJSEbw0N5ZjxtaKlT1s9e05_arQrNw,70
-unstructured_ingest-0.0.22.dist-info/top_level.txt,sha256=QaTxTcjfM5Hr9sZJ6weOJvSe5ESQc0F8AWkhHInTCf8,20
-unstructured_ingest-0.0.22.dist-info/RECORD,,
+unstructured_ingest-0.0.23.dist-info/LICENSE.md,sha256=SxkKP_62uIAKb9mb1eH7FH4Kn2aYT09fgjKpJt5PyTk,11360
+unstructured_ingest-0.0.23.dist-info/METADATA,sha256=iWfV6hzGvmClCO7_huz8s-h9FST1mJsc-mUHZQaGQU4,7108
+unstructured_ingest-0.0.23.dist-info/WHEEL,sha256=GV9aMThwP_4oNCtvEC2ec3qUYutgWeAzklro_0m4WJQ,91
+unstructured_ingest-0.0.23.dist-info/entry_points.txt,sha256=gUAAFnjFPnBgThJSEbw0N5ZjxtaKlT1s9e05_arQrNw,70
+unstructured_ingest-0.0.23.dist-info/top_level.txt,sha256=QaTxTcjfM5Hr9sZJ6weOJvSe5ESQc0F8AWkhHInTCf8,20
+unstructured_ingest-0.0.23.dist-info/RECORD,,