build(deps): bump unstructured.paddleocr to 2.8.1.0 (#3561)

### Summary - Bump `unstructured.paddleocr` to 2.8.1.0 - Remove `opencv-python` and `opencv-contrib-python` constraint pins - Fix `0.15.7` changelog
Unstructured-IO · Aug 23, 2024 · ac10ba4 · ac10ba4
1 parent 32bb77a
commit ac10ba4
Show file tree

Hide file tree

Showing 45 changed files with 97 additions and 115 deletions.
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -1,15 +1,26 @@
-## 0.15.8-dev3
+## 0.15.8-dev4
 
 ### Enhancements
 
+* **Bump unstructured.paddleocr to 2.8.1.0.**
+
 ### Features
 
 ### Fixes
 
-* **Fix NLTK data download path to prevent nested directories**. Resolved an issue where a nested "nltk_data" directory was created within the parent "nltk_data" directory when it already existed. This fix prevents errors in checking for existing downloads and loading models from NLTK data.
 * **Minify text_as_html from DOCX.** Previously `.metadata.text_as_html` for DOCX tables was "bloated" with whitespace and noise elements introduced by `tabulate` that produced over-chunking and lower "semantic density" of elements. Reduce HTML to minimum character count without preserving all text.
 * **Fall back to filename extension-based file-type detection for unidentified OLE files.** Resolves a problem where a DOC file that could not be detected as such by `filetype` was incorrectly identified as a MSG file.
 
+## 0.15.7
+
+### Enhancements
+
+### Features
+
+### Fixes
+
+* **Fix NLTK data download path to prevent nested directories**. Resolved an issue where a nested "nltk_data" directory was created within the parent "nltk_data" directory when it already existed. This fix prevents errors in checking for existing downloads and loading models from NLTK data.
+
 ## 0.15.6
 
 ### Enhancements

diff --git a/requirements/base.txt b/requirements/base.txt
@@ -43,7 +43,7 @@ httpcore==1.0.5
     # via httpx
 httpx==0.27.0
     # via unstructured-client
-idna==3.7
+idna==3.8
     # via
     #   anyio
     #   httpx

diff --git a/requirements/deps/constraints.txt b/requirements/deps/constraints.txt
@@ -27,9 +27,6 @@ tokenizers>=0.19,<0.20
 pycocotools>=2.0.7
 # NOTE(crag): python3.8-python3.11 compat (if it ends up being required)
 torch>2
-# pinned in unstructured paddleocr
-opencv-python==4.8.0.76
-opencv-contrib-python==4.8.0.76
 platformdirs==3.10.0
 
 # TODO: Constaint due to boto, with python before 3.10 not requiring openssl 1.1.1, remove when that gets

diff --git a/requirements/dev.txt b/requirements/dev.txt
@@ -68,7 +68,7 @@ filelock==3.15.4
     # via virtualenv
 identify==2.6.0
     # via pre-commit
-idna==3.7
+idna==3.8
     # via
     #   -c ./base.txt
     #   -c ./test.txt
@@ -94,7 +94,7 @@ ipython-genutils==0.2.0
     # via
     #   nbclassic
     #   notebook
-ipywidgets==8.1.3
+ipywidgets==8.1.5
     # via jupyter
 jedi==0.19.1
     # via ipython
@@ -140,7 +140,7 @@ jupyter-server-terminals==0.5.3
     # via jupyter-server
 jupyterlab-pygments==0.3.0
     # via nbconvert
-jupyterlab-widgets==3.0.11
+jupyterlab-widgets==3.0.13
     # via ipywidgets
 markupsafe==2.1.5
     # via
@@ -256,7 +256,7 @@ pyyaml==6.0.2
     #   -c ./test.txt
     #   jupyter-events
     #   pre-commit
-pyzmq==26.1.1
+pyzmq==26.2.0
     # via
     #   ipykernel
     #   jupyter-client
@@ -352,7 +352,7 @@ wheel==0.44.0
     # via
     #   -c ././deps/constraints.txt
     #   pip-tools
-widgetsnbextension==4.0.11
+widgetsnbextension==4.0.13
     # via ipywidgets
 zipp==3.20.0
     # via importlib-metadata

diff --git a/requirements/extra-paddleocr.in b/requirements/extra-paddleocr.in
@@ -2,4 +2,4 @@
 -c base.txt
 
 paddlepaddle==3.0.0b1
-unstructured.paddleocr==2.8.0.1
+unstructured.paddleocr==2.8.1.0
diff --git a/requirements/extra-paddleocr.txt b/requirements/extra-paddleocr.txt
@@ -10,10 +10,6 @@ anyio==4.4.0
     #   httpx
 astor==0.8.1
     # via paddlepaddle
-attrdict==2.0.1
-    # via unstructured-paddleocr
-cachetools==5.5.0
-    # via premailer
 certifi==2024.7.4
     # via
     #   -c ././deps/constraints.txt
@@ -27,18 +23,12 @@ charset-normalizer==3.3.2
     #   requests
 contourpy==1.2.1
     # via matplotlib
-cssselect==1.2.0
-    # via premailer
-cssutils==2.11.1
-    # via premailer
 cycler==0.12.1
     # via matplotlib
 cython==3.0.11
     # via unstructured-paddleocr
 decorator==5.1.1
     # via paddlepaddle
-et-xmlfile==1.1.0
-    # via openpyxl
 exceptiongroup==1.2.2
     # via
     #   -c ./base.txt
@@ -57,7 +47,7 @@ httpx==0.27.0
     # via
     #   -c ./base.txt
     #   paddlepaddle
-idna==3.7
+idna==3.8
     # via
     #   -c ./base.txt
     #   anyio
@@ -69,23 +59,14 @@ imageio==2.35.1
     #   scikit-image
 imgaug==0.4.0
     # via unstructured-paddleocr
-importlib-resources==6.4.3
+importlib-resources==6.4.4
     # via matplotlib
 kiwisolver==1.4.5
     # via matplotlib
-lanms-neo==1.0.2
-    # via unstructured-paddleocr
 lazy-loader==0.4
     # via scikit-image
-lxml==5.3.0
-    # via
-    #   -c ./base.txt
-    #   premailer
-    #   unstructured-paddleocr
 matplotlib==3.9.2
     # via imgaug
-more-itertools==10.4.0
-    # via cssutils
 networkx==3.2.1
     # via
     #   paddlepaddle
@@ -106,17 +87,12 @@ numpy==1.26.4
     #   shapely
     #   tifffile
     #   unstructured-paddleocr
-opencv-contrib-python==4.8.0.76
-    # via
-    #   -c ././deps/constraints.txt
-    #   unstructured-paddleocr
-opencv-python==4.8.0.76
+opencv-contrib-python==4.10.0.84
+    # via unstructured-paddleocr
+opencv-python==4.10.0.84
     # via
-    #   -c ././deps/constraints.txt
     #   imgaug
     #   unstructured-paddleocr
-openpyxl==3.1.5
-    # via unstructured-paddleocr
 opt-einsum==3.3.0
     # via paddlepaddle
 packaging==24.1
@@ -138,8 +114,6 @@ pillow==10.4.0
     #   pdf2image
     #   scikit-image
     #   unstructured-paddleocr
-premailer==3.10.0
-    # via unstructured-paddleocr
 protobuf==4.23.4
     # via
     #   -c ././deps/constraints.txt
@@ -161,7 +135,6 @@ rapidfuzz==3.9.6
 requests==2.32.3
     # via
     #   -c ./base.txt
-    #   premailer
     #   unstructured-paddleocr
 scikit-image==0.24.0
     # via
@@ -178,7 +151,6 @@ shapely==2.0.6
 six==1.16.0
     # via
     #   -c ./base.txt
-    #   attrdict
     #   imgaug
     #   python-dateutil
 sniffio==1.3.1
@@ -197,7 +169,7 @@ typing-extensions==4.12.2
     #   -c ./base.txt
     #   anyio
     #   paddlepaddle
-unstructured-paddleocr==2.8.0.1
+unstructured-paddleocr==2.8.1.0
     # via -r ./extra-paddleocr.in
 urllib3==1.26.19
     # via

diff --git a/requirements/extra-pdf-image.txt b/requirements/extra-pdf-image.txt
@@ -57,7 +57,7 @@ googleapis-common-protos==1.63.2
     # via
     #   google-api-core
     #   grpcio-status
-grpcio==1.65.5
+grpcio==1.66.0
     # via
     #   -c ././deps/constraints.txt
     #   google-api-core
@@ -72,11 +72,11 @@ huggingface-hub==0.24.6
     #   unstructured-inference
 humanfriendly==10.0
     # via coloredlogs
-idna==3.7
+idna==3.8
     # via
     #   -c ./base.txt
     #   requests
-importlib-resources==6.4.3
+importlib-resources==6.4.4
     # via matplotlib
 iopath==0.1.10
     # via layoutparser
@@ -122,9 +122,8 @@ onnx==1.16.2
     #   unstructured-inference
 onnxruntime==1.19.0
     # via unstructured-inference
-opencv-python==4.8.0.76
+opencv-python==4.10.0.84
     # via
-    #   -c ././deps/constraints.txt
     #   layoutparser
     #   unstructured-inference
 packaging==24.1
@@ -269,7 +268,7 @@ tqdm==4.66.5
     #   huggingface-hub
     #   iopath
     #   transformers
-transformers==4.44.1
+transformers==4.44.2
     # via unstructured-inference
 typing-extensions==4.12.2
     # via

diff --git a/requirements/huggingface.txt b/requirements/huggingface.txt
@@ -30,7 +30,7 @@ huggingface-hub==0.24.6
     # via
     #   tokenizers
     #   transformers
-idna==3.7
+idna==3.8
     # via
     #   -c ./base.txt
     #   requests
@@ -99,7 +99,7 @@ tqdm==4.66.5
     #   huggingface-hub
     #   sacremoses
     #   transformers
-transformers==4.44.1
+transformers==4.44.2
     # via -r ./huggingface.in
 typing-extensions==4.12.2
     # via

diff --git a/requirements/ingest/airtable.txt b/requirements/ingest/airtable.txt
@@ -15,7 +15,7 @@ charset-normalizer==3.3.2
     # via
     #   -c ./ingest/../base.txt
     #   requests
-idna==3.7
+idna==3.8
     # via
     #   -c ./ingest/../base.txt
     #   requests

diff --git a/requirements/ingest/astradb.txt b/requirements/ingest/astradb.txt
@@ -57,7 +57,7 @@ httpx[http2]==0.27.0
     #   astrapy
 hyperframe==6.0.1
     # via h2
-idna==3.7
+idna==3.8
     # via
     #   -c ./ingest/../base.txt
     #   anyio

diff --git a/requirements/ingest/azure-cognitive-search.txt b/requirements/ingest/azure-cognitive-search.txt
@@ -19,7 +19,7 @@ charset-normalizer==3.3.2
     # via
     #   -c ./ingest/../base.txt
     #   requests
-idna==3.7
+idna==3.8
     # via
     #   -c ./ingest/../base.txt
     #   requests

diff --git a/requirements/ingest/azure.txt b/requirements/ingest/azure.txt
@@ -54,7 +54,7 @@ fsspec==2024.6.1
     # via
     #   -r ./ingest/azure.in
     #   adlfs
-idna==3.7
+idna==3.8
     # via
     #   -c ./ingest/../base.txt
     #   requests

diff --git a/requirements/ingest/box.txt b/requirements/ingest/box.txt
@@ -8,7 +8,7 @@ attrs==24.2.0
     # via boxsdk
 boxfs==0.3.0
     # via -r ./ingest/box.in
-boxsdk[jwt]==3.12.0
+boxsdk[jwt]==3.13.0
     # via boxfs
 certifi==2024.7.4
     # via
@@ -27,7 +27,7 @@ fsspec==2024.6.1
     # via
     #   -r ./ingest/box.in
     #   boxfs
-idna==3.7
+idna==3.8
     # via
     #   -c ./ingest/../base.txt
     #   requests

diff --git a/requirements/ingest/chroma.txt b/requirements/ingest/chroma.txt
@@ -62,7 +62,7 @@ google-auth==2.34.0
     # via kubernetes
 googleapis-common-protos==1.63.2
     # via opentelemetry-exporter-otlp-proto-grpc
-grpcio==1.65.5
+grpcio==1.66.0
     # via
     #   -c ./ingest/../deps/constraints.txt
     #   chromadb
@@ -78,15 +78,15 @@ huggingface-hub==0.24.6
     # via tokenizers
 humanfriendly==10.0
     # via coloredlogs
-idna==3.7
+idna==3.8
     # via
     #   -c ./ingest/../base.txt
     #   anyio
     #   httpx
     #   requests
 importlib-metadata==8.4.0
     # via -r ./ingest/chroma.in
-importlib-resources==6.4.3
+importlib-resources==6.4.4
     # via chromadb
 kubernetes==30.1.0
     # via chromadb
@@ -129,7 +129,7 @@ packaging==24.1
     #   build
     #   huggingface-hub
     #   onnxruntime
-posthog==3.5.0
+posthog==3.5.2
     # via chromadb
 protobuf==4.23.4
     # via

diff --git a/requirements/ingest/clarifai.txt b/requirements/ingest/clarifai.txt
@@ -21,11 +21,11 @@ contextlib2==21.6.0
     # via schema
 googleapis-common-protos==1.63.2
     # via clarifai-grpc
-grpcio==1.65.5
+grpcio==1.66.0
     # via
     #   -c ./ingest/../deps/constraints.txt
     #   clarifai-grpc
-idna==3.7
+idna==3.8
     # via
     #   -c ./ingest/../base.txt
     #   requests

diff --git a/requirements/ingest/confluence.txt b/requirements/ingest/confluence.txt
@@ -21,7 +21,7 @@ charset-normalizer==3.3.2
     #   requests
 deprecated==1.2.14
     # via atlassian-python-api
-idna==3.7
+idna==3.8
     # via
     #   -c ./ingest/../base.txt
     #   requests