fix: correctly install mesa-gl for arm (#3647)

### Summary Fixes the `arm64` image builds, which will be available again starting in version `0.15.13`. A fix was implemented upstream in Unstructured-IO/base-images#47 and a workaround that installed `x86` packages in the `unstructured` repo was removed. ### Testing See [this job](https://github.com/Unstructured-IO/unstructured/actions/runs/10948943594/job/30401108059?pr=3647) for a successful `arm64` build on the feature branch.
Unstructured-IO · Sep 20, 2024 · 7d66a23 · 7d66a23
1 parent 0ed69a1
commit 7d66a23
Show file tree

Hide file tree

Showing 36 changed files with 198 additions and 156 deletions.
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -1,4 +1,4 @@
-## 0.15.13-dev3
+## 0.15.13
 
 ### Enhancements
 
@@ -9,6 +9,7 @@
 ### Fixes
 
 * **Fixes high memory overhead for intersection area computation** Using `numpy.float32` for coordinates and remove intermediate variables to reduce memory usage when computing intersection areas
+* **Fixes the `arm64` image build** `arm64` builds are now fixed and will be available against starting with the `0.15.13` release.
 
 ## 0.15.12
 

diff --git a/Dockerfile b/Dockerfile
@@ -9,17 +9,6 @@ COPY unstructured unstructured
 COPY test_unstructured test_unstructured
 COPY example-docs example-docs
 
-# NOTE(robinson) - temporary workaround to install mesa-gl 24.1.0 because
-# libgallum is missing in mesa-gl 24.2.0 from the wolfi package manager
-RUN wget "https://utic-public-cf.s3.amazonaws.com/mesa-gl-24.1.0-r0.718c913d.apk" && \
-  wget "https://utic-public-cf.s3.amazonaws.com/mesa-glapi-24.1.0-r0.4390a503.apk" && \
-  apk del mesa-gl && \
-  apk add --allow-untrusted mesa-gl-24.1.0-r0.718c913d.apk && \
-  apk add --allow-untrusted mesa-glapi-24.1.0-r0.4390a503.apk && \
-  rm mesa-gl-24.1.0-r0.718c913d.apk && \
-  rm mesa-glapi-24.1.0-r0.4390a503.apk
-
-
 RUN chown -R notebook-user:notebook-user /app && \
   apk add font-ubuntu git && \
   fc-cache -fv && \

diff --git a/requirements/base.txt b/requirements/base.txt
@@ -4,7 +4,7 @@
 #
 #    pip-compile ./base.in
 #
-anyio==4.4.0
+anyio==4.5.0
     # via httpx
 backoff==2.2.1
     # via -r ./base.in
@@ -16,6 +16,8 @@ certifi==2024.8.30
     #   httpx
     #   requests
     #   unstructured-client
+cffi==1.17.1
+    # via cryptography
 chardet==5.2.0
     # via -r ./base.in
 charset-normalizer==3.3.2
@@ -26,13 +28,15 @@ click==8.1.7
     # via
     #   nltk
     #   python-oxmsg
+cryptography==43.0.1
+    # via unstructured-client
 dataclasses-json==0.6.7
     # via
     #   -r ./base.in
     #   unstructured-client
 deepdiff==8.0.1
     # via unstructured-client
-emoji==2.12.1
+emoji==2.13.0
     # via -r ./base.in
 exceptiongroup==1.2.2
     # via anyio
@@ -82,7 +86,9 @@ packaging==24.1
     #   unstructured-client
 psutil==6.0.0
     # via -r ./base.in
-pypdf==4.3.1
+pycparser==2.22
+    # via cffi
+pypdf==5.0.0
     # via unstructured-client
 python-dateutil==2.9.0.post0
     # via unstructured-client
@@ -133,7 +139,7 @@ typing-inspect==0.9.0
     # via
     #   dataclasses-json
     #   unstructured-client
-unstructured-client==0.25.8
+unstructured-client==0.25.9
     # via -r ./base.in
 urllib3==1.26.20
     # via

diff --git a/requirements/dev.txt b/requirements/dev.txt
@@ -15,7 +15,7 @@ click==8.1.7
     #   pip-tools
 distlib==0.3.8
     # via virtualenv
-filelock==3.16.0
+filelock==3.16.1
     # via virtualenv
 identify==2.6.1
     # via pre-commit
@@ -32,7 +32,7 @@ packaging==24.1
     #   build
 pip-tools==7.4.1
     # via -r ./dev.in
-platformdirs==4.3.3
+platformdirs==4.3.6
     # via
     #   -c ./test.txt
     #   virtualenv
@@ -51,7 +51,7 @@ tomli==2.0.1
     #   -c ./test.txt
     #   build
     #   pip-tools
-virtualenv==20.26.4
+virtualenv==20.26.5
     # via pre-commit
 wheel==0.44.0
     # via pip-tools

diff --git a/requirements/extra-paddleocr.txt b/requirements/extra-paddleocr.txt
@@ -4,7 +4,7 @@
 #
 #    pip-compile ./extra-paddleocr.in
 #
-anyio==4.4.0
+anyio==4.5.0
     # via
     #   -c ./base.txt
     #   httpx
@@ -113,7 +113,7 @@ pillow==10.4.0
     #   pdf2image
     #   scikit-image
     #   unstructured-paddleocr
-protobuf==4.25.4
+protobuf==4.25.5
     # via
     #   -c ././deps/constraints.txt
     #   paddlepaddle

diff --git a/requirements/extra-pdf-image.txt b/requirements/extra-pdf-image.txt
@@ -13,7 +13,9 @@ certifi==2024.8.30
     #   -c ./base.txt
     #   requests
 cffi==1.17.1
-    # via cryptography
+    # via
+    #   -c ./base.txt
+    #   cryptography
 charset-normalizer==3.3.2
     # via
     #   -c ./base.txt
@@ -24,14 +26,16 @@ coloredlogs==15.0.1
 contourpy==1.3.0
     # via matplotlib
 cryptography==43.0.1
-    # via pdfminer-six
+    # via
+    #   -c ./base.txt
+    #   pdfminer-six
 cycler==0.12.1
     # via matplotlib
 deprecated==1.2.14
     # via pikepdf
 effdet==0.4.1
     # via -r ./extra-pdf-image.in
-filelock==3.16.0
+filelock==3.16.1
     # via
     #   huggingface-hub
     #   torch
@@ -44,9 +48,9 @@ fsspec==2024.9.0
     # via
     #   huggingface-hub
     #   torch
-google-api-core[grpc]==2.19.2
+google-api-core[grpc]==2.20.0
     # via google-cloud-vision
-google-auth==2.34.0
+google-auth==2.35.0
     # via
     #   google-api-core
     #   google-cloud-vision
@@ -63,7 +67,7 @@ grpcio==1.66.1
     #   grpcio-status
 grpcio-status==1.62.3
     # via google-api-core
-huggingface-hub==0.24.7
+huggingface-hub==0.25.0
     # via
     #   timm
     #   tokenizers
@@ -166,7 +170,7 @@ proto-plus==1.24.0
     # via
     #   google-api-core
     #   google-cloud-vision
-protobuf==4.25.4
+protobuf==4.25.5
     # via
     #   -c ././deps/constraints.txt
     #   google-api-core
@@ -185,10 +189,12 @@ pyasn1-modules==0.4.1
 pycocotools==2.0.8
     # via effdet
 pycparser==2.22
-    # via cffi
+    # via
+    #   -c ./base.txt
+    #   cffi
 pyparsing==3.1.4
     # via matplotlib
-pypdf==4.3.1
+pypdf==5.0.0
     # via
     #   -c ./base.txt
     #   -r ./extra-pdf-image.in
@@ -236,7 +242,7 @@ six==1.16.0
     # via
     #   -c ./base.txt
     #   python-dateutil
-sympy==1.13.2
+sympy==1.13.3
     # via
     #   onnxruntime
     #   torch

diff --git a/requirements/huggingface.txt b/requirements/huggingface.txt
@@ -16,7 +16,7 @@ click==8.1.7
     # via
     #   -c ./base.txt
     #   sacremoses
-filelock==3.16.0
+filelock==3.16.1
     # via
     #   huggingface-hub
     #   torch
@@ -25,7 +25,7 @@ fsspec==2024.9.0
     # via
     #   huggingface-hub
     #   torch
-huggingface-hub==0.24.7
+huggingface-hub==0.25.0
     # via
     #   tokenizers
     #   transformers
@@ -82,7 +82,7 @@ six==1.16.0
     # via
     #   -c ./base.txt
     #   langdetect
-sympy==1.13.2
+sympy==1.13.3
     # via torch
 tokenizers==0.19.1
     # via

diff --git a/requirements/ingest/airtable.txt b/requirements/ingest/airtable.txt
@@ -22,9 +22,9 @@ inflection==0.5.1
     # via pyairtable
 pyairtable==2.3.3
     # via -r ./ingest/airtable.in
-pydantic==2.9.1
+pydantic==2.9.2
     # via pyairtable
-pydantic-core==2.23.3
+pydantic-core==2.23.4
     # via pydantic
 requests==2.32.3
     # via

diff --git a/requirements/ingest/astradb.txt b/requirements/ingest/astradb.txt
@@ -4,15 +4,15 @@
 #
 #    pip-compile ./ingest/astradb.in
 #
-anyio==4.4.0
+anyio==4.5.0
     # via
     #   -c ./ingest/../base.txt
     #   httpx
 astrapy==1.4.2
     # via -r ./ingest/astradb.in
 cassandra-driver==3.29.2
     # via cassio
-cassio==0.1.8
+cassio==0.1.9
     # via astrapy
 certifi==2024.8.30
     # via
@@ -70,7 +70,7 @@ packaging==24.1
     # via
     #   -c ./ingest/../base.txt
     #   deprecation
-pymongo==4.8.0
+pymongo==4.9.1
     # via astrapy
 requests==2.32.3
     # via

diff --git a/requirements/ingest/azure.txt b/requirements/ingest/azure.txt
@@ -23,16 +23,17 @@ azure-core==1.31.0
     #   azure-storage-blob
 azure-datalake-store==0.0.53
     # via adlfs
-azure-identity==1.17.1
+azure-identity==1.18.0
     # via adlfs
-azure-storage-blob==12.22.0
+azure-storage-blob==12.23.0
     # via adlfs
 certifi==2024.8.30
     # via
     #   -c ./ingest/../base.txt
     #   requests
 cffi==1.17.1
     # via
+    #   -c ./ingest/../base.txt
     #   azure-datalake-store
     #   cryptography
 charset-normalizer==3.3.2
@@ -41,6 +42,7 @@ charset-normalizer==3.3.2
     #   requests
 cryptography==43.0.1
     # via
+    #   -c ./ingest/../base.txt
     #   azure-identity
     #   azure-storage-blob
     #   msal
@@ -74,11 +76,11 @@ multidict==6.1.0
 portalocker==2.10.1
     # via msal-extensions
 pycparser==2.22
-    # via cffi
-pyjwt[crypto]==2.9.0
     # via
-    #   msal
-    #   pyjwt
+    #   -c ./ingest/../base.txt
+    #   cffi
+pyjwt[crypto]==2.9.0
+    # via msal
 requests==2.32.3
     # via
     #   -c ./ingest/../base.txt

diff --git a/requirements/ingest/box.txt b/requirements/ingest/box.txt
@@ -15,13 +15,17 @@ certifi==2024.8.30
     #   -c ./ingest/../base.txt
     #   requests
 cffi==1.17.1
-    # via cryptography
+    # via
+    #   -c ./ingest/../base.txt
+    #   cryptography
 charset-normalizer==3.3.2
     # via
     #   -c ./ingest/../base.txt
     #   requests
 cryptography==43.0.1
-    # via boxsdk
+    # via
+    #   -c ./ingest/../base.txt
+    #   boxsdk
 fsspec==2024.9.0
     # via
     #   -r ./ingest/box.in
@@ -31,7 +35,9 @@ idna==3.10
     #   -c ./ingest/../base.txt
     #   requests
 pycparser==2.22
-    # via cffi
+    # via
+    #   -c ./ingest/../base.txt
+    #   cffi
 pyjwt==2.9.0
     # via boxsdk
 python-dateutil==2.9.0.post0