chore(deps): bump unstructured to 0.13.2 (Unstructured-IO#402)

Bump `unstructured` dependency to 0.13.2
Test adjusted to reflect table extraction change introduced in unstructured 0.13.0

---------

Co-authored-by: Filip Knefel <[email protected]>
Co-authored-by: Paweł Kmiecik <[email protected]>

Author: 3 people

CHANGELOG.md

@@ -1,7 +1,8 @@
-## 0.0.66-dev1
+## 0.0.66-dev2
 
 * Add support for `unique_element_ids` parameter.
 * Add max lifetime, via MAX_LIFETIME_SECONDS env-var, to API containers
+* Bump unstructured to 0.13.2
 
 ## 0.0.65
 

requirements/base.txt

@@ -19,7 +19,6 @@ beautifulsoup4==4.12.3
 certifi==2024.2.2
     # via
     #   requests
-    #   unstructured
     #   unstructured-client
 cffi==1.16.0
     # via cryptography
@@ -29,7 +28,6 @@ charset-normalizer==3.3.2
     # via
     #   pdfminer-six
     #   requests
-    #   unstructured
     #   unstructured-client
 click==8.1.3
     # via
@@ -38,48 +36,46 @@ click==8.1.3
     #   uvicorn
 coloredlogs==15.0.1
     # via onnxruntime
-contourpy==1.2.0
+contourpy==1.2.1
     # via matplotlib
 cryptography==42.0.5
     # via pdfminer-six
 cycler==0.12.1
     # via matplotlib
 dataclasses-json==0.6.4
-    # via
-    #   unstructured
-    #   unstructured-client
-deepdiff==6.7.1
+    # via unstructured
+dataclasses-json-speakeasy==0.5.11
     # via unstructured-client
 deprecated==1.2.14
     # via pikepdf
 effdet==0.4.1
     # via layoutparser
-emoji==2.10.1
+emoji==2.11.0
     # via unstructured
 et-xmlfile==1.1.0
     # via openpyxl
 exceptiongroup==1.2.0
     # via anyio
-fastapi==0.110.0
+fastapi==0.110.1
     # via -r requirements/base.in
-filelock==3.13.1
+filelock==3.13.4
     # via
     #   huggingface-hub
     #   torch
     #   transformers
 filetype==1.2.0
     # via unstructured
-flatbuffers==24.3.7
+flatbuffers==24.3.25
     # via onnxruntime
-fonttools==4.49.0
+fonttools==4.51.0
     # via matplotlib
-fsspec==2024.2.0
+fsspec==2024.3.1
     # via
     #   huggingface-hub
     #   torch
 h11==0.14.0
     # via uvicorn
-huggingface-hub==0.21.4
+huggingface-hub==0.22.2
     # via
     #   timm
     #   tokenizers
@@ -96,7 +92,7 @@ iopath==0.1.10
     # via layoutparser
 jinja2==3.1.3
     # via torch
-joblib==1.3.2
+joblib==1.4.0
     # via nltk
 jsonpath-python==1.0.6
     # via unstructured-client
@@ -106,21 +102,22 @@ langdetect==1.0.9
     # via unstructured
 layoutparser[layoutmodels,tesseract]==0.3.4
     # via unstructured-inference
-lxml==5.1.0
+lxml==5.2.1
     # via
     #   pikepdf
     #   python-docx
     #   python-pptx
     #   unstructured
-markdown==3.5.2
+markdown==3.6
     # via unstructured
 markupsafe==2.1.5
     # via jinja2
 marshmallow==3.21.1
     # via
     #   dataclasses-json
+    #   dataclasses-json-speakeasy
     #   unstructured-client
-matplotlib==3.8.3
+matplotlib==3.8.4
     # via pycocotools
 mpmath==1.3.0
     # via sympy
@@ -130,7 +127,7 @@ mypy-extensions==1.0.0
     # via
     #   typing-inspect
     #   unstructured-client
-networkx==3.2.1
+networkx==3.3
     # via
     #   torch
     #   unstructured
@@ -154,7 +151,7 @@ olefile==0.47
     # via msg-parser
 omegaconf==2.3.0
     # via effdet
-onnx==1.15.0
+onnx==1.16.0
     # via
     #   unstructured
     #   unstructured-inference
@@ -166,8 +163,6 @@ opencv-python==4.9.0.80
     #   unstructured-inference
 openpyxl==3.1.2
     # via unstructured
-ordered-set==4.1.0
-    # via deepdiff
 packaging==24.0
     # via
     #   huggingface-hub
@@ -193,9 +188,9 @@ pdfminer-six==20231228
     #   unstructured
 pdfplumber==0.11.0
     # via layoutparser
-pikepdf==8.13.0
+pikepdf==8.15.0
     # via unstructured
-pillow==10.2.0
+pillow==10.3.0
     # via
     #   layoutparser
     #   matplotlib
@@ -207,19 +202,19 @@ pillow==10.2.0
     #   python-pptx
     #   torchvision
     #   unstructured-pytesseract
-pillow-heif==0.15.0
+pillow-heif==0.16.0
     # via unstructured
 portalocker==2.8.2
     # via iopath
-protobuf==4.25.3
+protobuf==5.26.1
     # via
     #   onnx
     #   onnxruntime
 psutil==5.9.8
     # via -r requirements/base.in
 pycocotools==2.0.7
     # via effdet
-pycparser==2.21
+pycparser==2.22
     # via cffi
 pycryptodome==3.20.0
     # via -r requirements/base.in
@@ -231,12 +226,11 @@ pypandoc==1.13
     # via unstructured
 pyparsing==3.1.2
     # via matplotlib
-pypdf==4.1.0
+pypdf==4.2.0
     # via
     #   -r requirements/base.in
     #   unstructured
-    #   unstructured-client
-pypdfium2==4.28.0
+pypdfium2==4.29.0
     # via pdfplumber
 pytesseract==0.3.10
     # via layoutparser
@@ -264,7 +258,7 @@ pyyaml==6.0.1
     #   omegaconf
     #   timm
     #   transformers
-rapidfuzz==3.6.2
+rapidfuzz==3.8.1
     # via
     #   unstructured
     #   unstructured-inference
@@ -286,7 +280,7 @@ safetensors==0.3.2
     #   -c requirements/constraints.in
     #   timm
     #   transformers
-scipy==1.12.0
+scipy==1.13.0
     # via layoutparser
 six==1.16.0
     # via
@@ -297,7 +291,7 @@ sniffio==1.3.1
     # via anyio
 soupsieve==2.5
     # via beautifulsoup4
-starlette==0.36.3
+starlette==0.37.2
     # via fastapi
 sympy==1.12
     # via
@@ -309,13 +303,13 @@ timm==0.9.16
     # via effdet
 tokenizers==0.15.2
     # via transformers
-torch==2.2.1
+torch==2.2.2
     # via
     #   effdet
     #   layoutparser
     #   timm
     #   torchvision
-torchvision==0.17.1
+torchvision==0.17.2
     # via
     #   effdet
     #   layoutparser
@@ -328,14 +322,15 @@ tqdm==4.66.2
     #   transformers
 transformers==4.37.1
     # via unstructured-inference
-typing-extensions==4.10.0
+typing-extensions==4.11.0
     # via
     #   anyio
     #   fastapi
     #   huggingface-hub
     #   iopath
     #   pydantic
     #   pydantic-core
+    #   pypdf
     #   python-docx
     #   torch
     #   typing-inspect
@@ -345,24 +340,23 @@ typing-extensions==4.10.0
 typing-inspect==0.9.0
     # via
     #   dataclasses-json
+    #   dataclasses-json-speakeasy
     #   unstructured-client
 tzdata==2024.1
     # via pandas
-unstructured[local-inference]==0.12.5
-    # via
-    #   -r requirements/base.in
-    #   unstructured
-unstructured-client==0.21.1
+unstructured[local-inference]==0.13.2
+    # via -r requirements/base.in
+unstructured-client==0.18.0
     # via unstructured
-unstructured-inference==0.7.23
+unstructured-inference==0.7.25
     # via unstructured
 unstructured-pytesseract==0.3.12
     # via unstructured
 urllib3==2.2.1
     # via
     #   requests
     #   unstructured-client
-uvicorn==0.28.0
+uvicorn==0.29.0
     # via -r requirements/base.in
 wrapt==1.16.0
     # via