Chore/improve ingest imports (#46)

Author: ctrahey

Diff for: CHANGELOG.md

@@ -1,3 +1,8 @@
+## 0.0.26
+
+* **Bump `unstructured-ingest` to 0.5.23**
+* **Change how we import FileData type for slimmer import graph at runtime**
+
 ## 0.0.25
 
 * **Remove message channels from input signature**

Diff for: requirements/cli.txt

@@ -2,7 +2,7 @@
 #    uv pip compile ./cli.in --output-file ./cli.txt --no-strip-extras --python-version 3.10
 annotated-types==0.7.0
     # via pydantic
-anyio==4.8.0
+anyio==4.9.0
     # via starlette
 asgiref==3.8.1
     # via opentelemetry-instrumentation-asgi
@@ -13,68 +13,68 @@ click==8.1.8
     #   uvicorn
 dataclasses-json==0.6.7
     # via unstructured-ingest
-deprecated==1.2.15
+deprecated==1.2.18
     # via
     #   opentelemetry-api
     #   opentelemetry-exporter-otlp-proto-grpc
     #   opentelemetry-semantic-conventions
 exceptiongroup==1.2.2
     # via anyio
-fastapi==0.115.6
+fastapi==0.115.12
     # via -r ./cli.in
-googleapis-common-protos==1.66.0
+googleapis-common-protos==1.69.2
     # via opentelemetry-exporter-otlp-proto-grpc
-grpcio==1.69.0
+grpcio==1.71.0
     # via opentelemetry-exporter-otlp-proto-grpc
 h11==0.14.0
     # via uvicorn
 idna==3.10
     # via anyio
-importlib-metadata==8.5.0
+importlib-metadata==8.6.1
     # via opentelemetry-api
-marshmallow==3.25.1
+marshmallow==3.26.1
     # via dataclasses-json
 mypy-extensions==1.0.0
     # via typing-inspect
-ndjson==0.3.1
-    # via unstructured-ingest
-numpy==2.2.2
-    # via pandas
-opentelemetry-api==1.29.0
+numpy==2.2.4
+    # via
+    #   pandas
+    #   unstructured-ingest
+opentelemetry-api==1.31.1
     # via
     #   opentelemetry-exporter-otlp-proto-grpc
     #   opentelemetry-instrumentation
     #   opentelemetry-instrumentation-asgi
     #   opentelemetry-instrumentation-fastapi
     #   opentelemetry-sdk
     #   opentelemetry-semantic-conventions
-opentelemetry-exporter-otlp-proto-common==1.29.0
+opentelemetry-exporter-otlp-proto-common==1.31.1
     # via opentelemetry-exporter-otlp-proto-grpc
-opentelemetry-exporter-otlp-proto-grpc==1.29.0
+opentelemetry-exporter-otlp-proto-grpc==1.31.1
     # via -r ./cli.in
-opentelemetry-instrumentation==0.50b0
+opentelemetry-instrumentation==0.52b1
     # via
     #   opentelemetry-instrumentation-asgi
     #   opentelemetry-instrumentation-fastapi
-opentelemetry-instrumentation-asgi==0.50b0
+opentelemetry-instrumentation-asgi==0.52b1
     # via opentelemetry-instrumentation-fastapi
-opentelemetry-instrumentation-fastapi==0.50b0
+opentelemetry-instrumentation-fastapi==0.52b1
     # via -r ./cli.in
-opentelemetry-proto==1.29.0
+opentelemetry-proto==1.31.1
     # via
     #   opentelemetry-exporter-otlp-proto-common
     #   opentelemetry-exporter-otlp-proto-grpc
-opentelemetry-sdk==1.29.0
+opentelemetry-sdk==1.31.1
     # via
     #   opentelemetry-exporter-otlp-proto-grpc
     #   unstructured-ingest
-opentelemetry-semantic-conventions==0.50b0
+opentelemetry-semantic-conventions==0.52b1
     # via
     #   opentelemetry-instrumentation
     #   opentelemetry-instrumentation-asgi
     #   opentelemetry-instrumentation-fastapi
     #   opentelemetry-sdk
-opentelemetry-util-http==0.50b0
+opentelemetry-util-http==0.52b1
     # via
     #   opentelemetry-instrumentation-asgi
     #   opentelemetry-instrumentation-fastapi
@@ -84,11 +84,11 @@ packaging==24.2
     #   opentelemetry-instrumentation
 pandas==2.2.3
     # via unstructured-ingest
-protobuf==5.29.3
+protobuf==5.29.4
     # via
     #   googleapis-common-protos
     #   opentelemetry-proto
-pydantic==2.10.5
+pydantic==2.10.6
     # via
     #   fastapi
     #   unstructured-ingest
@@ -98,13 +98,13 @@ python-dateutil==2.9.0.post0
     # via
     #   pandas
     #   unstructured-ingest
-pytz==2024.2
+pytz==2025.2
     # via pandas
 six==1.17.0
     # via python-dateutil
 sniffio==1.3.1
     # via anyio
-starlette==0.41.3
+starlette==0.46.1
     # via fastapi
 tqdm==4.67.1
     # via unstructured-ingest
@@ -120,9 +120,9 @@ typing-extensions==4.12.2
     #   uvicorn
 typing-inspect==0.9.0
     # via dataclasses-json
-tzdata==2025.1
+tzdata==2025.2
     # via pandas
-unstructured-ingest==0.4.0
+unstructured-ingest==0.5.23
     # via -r ./cli.in
 uvicorn==0.34.0
     # via -r ./cli.in

Diff for: requirements/constraints.txt

@@ -1 +1 @@
-unstructured-ingest>=0.4.0
+unstructured-ingest>=0.5.23

Diff for: requirements/lint.txt

@@ -2,19 +2,19 @@
 #    uv pip compile ./lint.in --output-file ./lint.txt --no-strip-extras --python-version 3.10
 autoflake==2.3.1
     # via -r ./lint.in
-black==24.10.0
+black==25.1.0
     # via -r ./lint.in
 click==8.1.8
     # via black
-flake8==7.1.1
+flake8==7.1.2
     # via
     #   -r ./lint.in
     #   flake8-print
 flake8-print==5.0.0
     # via -r ./lint.in
 mccabe==0.7.0
     # via flake8
-mypy==1.14.1
+mypy==1.15.0
     # via -r ./lint.in
 mypy-extensions==1.0.0
     # via
@@ -24,7 +24,7 @@ packaging==24.2
     # via black
 pathspec==0.12.1
     # via black
-platformdirs==4.3.6
+platformdirs==4.3.7
     # via black
 pycodestyle==2.12.1
     # via
@@ -34,7 +34,7 @@ pyflakes==3.2.0
     # via
     #   autoflake
     #   flake8
-ruff==0.9.2
+ruff==0.11.2
     # via -r ./lint.in
 tomli==2.2.1
     # via

Diff for: requirements/release.txt

@@ -2,7 +2,7 @@
 #    uv pip compile ./release.in --output-file ./release.txt --no-strip-extras --python-version 3.10
 backports-tarfile==1.2.0
     # via jaraco-context
-certifi==2024.12.14
+certifi==2025.1.31
     # via requests
 charset-normalizer==3.4.1
     # via requests
@@ -30,7 +30,7 @@ more-itertools==10.6.0
     # via
     #   jaraco-classes
     #   jaraco-functools
-nh3==0.2.20
+nh3==0.2.21
     # via readme-renderer
 packaging==24.2
     # via twine

Diff for: requirements/test.txt

@@ -1,8 +1,8 @@
 # This file was autogenerated by uv via the following command:
 #    uv pip compile ./test.in --output-file ./test.txt --no-strip-extras --python-version 3.10
-anyio==4.8.0
+anyio==4.9.0
     # via httpx
-certifi==2024.12.14
+certifi==2025.1.31
     # via
     #   httpcore
     #   httpx
@@ -20,13 +20,13 @@ idna==3.10
     # via
     #   anyio
     #   httpx
-iniconfig==2.0.0
+iniconfig==2.1.0
     # via pytest
 packaging==24.2
     # via pytest
 pluggy==1.5.0
     # via pytest
-pytest==8.3.4
+pytest==8.3.5
     # via -r ./test.in
 sniffio==1.3.1
     # via anyio

Diff for: requirements/validate.txt

@@ -1,6 +1,6 @@
 # This file was autogenerated by uv via the following command:
 #    uv pip compile ./validate.in --output-file ./validate.txt --no-strip-extras --python-version 3.10
-certifi==2024.12.14
+certifi==2025.1.31
     # via requests
 charset-normalizer==3.4.1
     # via requests

Diff for: test/api/test_api.py

@@ -4,7 +4,12 @@
 import pytest
 from fastapi.testclient import TestClient
 from pydantic import BaseModel
-from unstructured_ingest.v2.interfaces import BatchFileData, BatchItem, FileData, SourceIdentifiers
+from unstructured_ingest.v2.types.file_data import (
+    BatchFileData,
+    BatchItem,
+    FileData,
+    SourceIdentifiers,
+)
 
 from unstructured_platform_plugins.etl_uvicorn.api_generator import (
     EtlApiException,

Diff for: test/assets/dataclass_response.py

@@ -2,7 +2,7 @@
 from pathlib import Path
 from typing import Any, Optional, TypedDict, Union
 
-from unstructured_ingest.v2.interfaces import BatchFileData, FileData
+from unstructured_ingest.v2.types.file_data import BatchFileData, FileData
 
 
 class SampleFunctionResponse(TypedDict):

Diff for: test/assets/filedata_meta.py

@@ -2,7 +2,7 @@
 from typing import Optional, Union
 
 from pydantic import BaseModel
-from unstructured_ingest.v2.interfaces import BatchFileData, FileData, SourceIdentifiers
+from unstructured_ingest.v2.types.file_data import BatchFileData, FileData, SourceIdentifiers
 
 from unstructured_platform_plugins.schema import FileDataMeta, NewRecord
 

Diff for: test/test_schema.py

@@ -6,7 +6,7 @@
 import pytest
 from pydantic import BaseModel
 from typing_extensions import TypedDict
-from unstructured_ingest.v2.interfaces import FileData
+from unstructured_ingest.v2.types.file_data import FileData
 
 import unstructured_platform_plugins.schema.json_schema as js
 from unstructured_platform_plugins.schema.model import is_valid_input_dict, is_valid_response_dict

Diff for: test/test_utils.py

@@ -4,7 +4,7 @@
 
 import pytest
 from pydantic import BaseModel
-from unstructured_ingest.v2.interfaces import FileData, SourceIdentifiers
+from unstructured_ingest.v2.types.file_data import FileData, SourceIdentifiers
 from uvicorn.importer import import_from_string
 
 from unstructured_platform_plugins.etl_uvicorn import utils

Diff for: unstructured_platform_plugins/__version__.py

@@ -1 +1 @@
-__version__ = "0.0.25"  # pragma: no cover
+__version__ = "0.0.26"  # pragma: no cover

Diff for: unstructured_platform_plugins/etl_uvicorn/api_generator.py

@@ -11,7 +11,7 @@
 from opentelemetry.instrumentation.fastapi import FastAPIInstrumentor
 from pydantic import BaseModel, Field, create_model
 from starlette.responses import RedirectResponse
-from unstructured_ingest.v2.interfaces.file_data import file_data_from_dict
+from unstructured_ingest.v2.types.file_data import file_data_from_dict
 from uvicorn.config import LOG_LEVELS
 from uvicorn.importer import import_from_string
 

Diff for: unstructured_platform_plugins/schema/filedata_meta.py

@@ -1,7 +1,7 @@
 from typing import Any
 
 from pydantic import BaseModel, Field
-from unstructured_ingest.v2.interfaces import FileData
+from unstructured_ingest.v2.types.file_data import FileData
 
 
 class NewRecord(BaseModel):

Diff for: unstructured_platform_plugins/schema/json_schema.py

@@ -8,7 +8,7 @@
 
 from pydantic import BaseModel, create_model
 from pydantic.fields import FieldInfo, PydanticUndefined
-from unstructured_ingest.v2.interfaces import BatchFileData, FileData
+from unstructured_ingest.v2.types.file_data import BatchFileData, FileData
 
 from unstructured_platform_plugins.schema.utils import TypedParameter
 from unstructured_platform_plugins.type_hints import get_type_hints