Merge branch 'main' into CORE-3587/better-element-ids

CHANGELOG.md

@@ -1,4 +1,4 @@
-## 0.13.3-dev8
+## 0.13.3
 
 ### Enhancements
 * **Unique and deterministic hash IDs for elements** Element IDs produced by any partitioning function are now deterministic and unique at the document level by default. Before, hashes were based only on text; however, they now also take into account the element's sequence number on a page, the page's number in the document, and the document's file name.
@@ -16,6 +16,7 @@
 
 * **Add support for extracting text from tag tails in HTML**. This fix adds ability to generate separate elements using tag tails.
 * **Add support for extracting text from `<b>` tags in HTML** Now `partition_html()` can extract text from `<b>` tags inside container tags (like `<div>`, `<pre>`).
+* **Fix pip-compile make target** Missing base.in dependency missing from requirments make file added
 
 ## 0.13.2
 

docs/requirements.txt

@@ -30,7 +30,7 @@ docutils==0.18.1
     #   sphinx-tabs
 furo==2023.7.26
     # via -r ./build.in
-idna==3.6
+idna==3.7
     # via
     #   -c ./base.txt
     #   requests

requirements/Makefile

@@ -8,7 +8,7 @@ INGEST_REQUIREMENTSTXT := $(patsubst %.in,%.txt,$(INGEST_REQUIREMENTS))
 
 
 .PHONY: all
-all: compile-base compile-ingest
+all: compile-all-base compile-ingest
 
 .PHONY: compile-test
 compile-test:
@@ -19,12 +19,16 @@ compile-dev:
 	pip-compile --upgrade dev.in
 
 .PHONY: compile-base
-compile-base: compile-test compile-dev
-	@$(foreach file,$(BASE_REQUIREMENTS),echo "compiling: $(file)" && pip-compile --upgrade $(file);)
+compile-base:
+	pip-compile --upgrade base.in
+
+.PHONY: compile-all-base
+compile-all-base: compile-base compile-test compile-dev
+	@$(foreach file,$(BASE_REQUIREMENTS),echo -e "\n\ncompiling: $(file)" && pip-compile --no-strip-extras --upgrade $(file) || exit;)
 
 .PHONY: compile-ingest
 compile-ingest:
-	@$(foreach file,$(INGEST_REQUIREMENTS),echo "compiling: $(file)" && pip-compile --upgrade $(file);)
+	@$(foreach file,$(INGEST_REQUIREMENTS),echo -e "\n\ncompiling: $(file)" && pip-compile --no-strip-extras --upgrade $(file) || exit;)
 
 .PHONY: clean
 clean: clean-base clean-ingest

requirements/base.txt

@@ -29,11 +29,11 @@ emoji==2.11.0
     # via -r ./base.in
 filetype==1.2.0
     # via -r ./base.in
-idna==3.6
+idna==3.7
     # via
     #   requests
     #   unstructured-client
-joblib==1.3.2
+joblib==1.4.0
     # via nltk
 jsonpath-python==1.0.6
     # via unstructured-client
@@ -67,9 +67,9 @@ python-iso639==2024.2.7
     # via -r ./base.in
 python-magic==0.4.27
     # via -r ./base.in
-rapidfuzz==3.7.0
+rapidfuzz==3.8.1
     # via -r ./base.in
-regex==2023.12.25
+regex==2024.4.16
     # via nltk
 requests==2.31.0
     # via

requirements/build.txt

@@ -30,7 +30,7 @@ docutils==0.18.1
     #   sphinx-tabs
 furo==2023.7.26
     # via -r ./build.in
-idna==3.6
+idna==3.7
     # via
     #   -c ./base.txt
     #   requests

requirements/deps/constraints.txt

@@ -22,8 +22,6 @@ Office365-REST-Python-Client<2.4.3
 # unstructured-inference to be upgraded when unstructured library is upgraded
 # https://github.com/Unstructured-IO/unstructured/issues/1458
 # unstructured-inference
-# unable to build wheel for arm on 0.3.3+
-safetensors<=0.3.2
 # use the known compatible version of weaviate and unstructured.pytesseract
 unstructured.pytesseract>=0.3.12
 weaviate-client>3.25.0
@@ -38,7 +36,6 @@ torch>2
 # pinned in unstructured paddleocr
 opencv-python==4.8.0.76
 opencv-contrib-python==4.8.0.76
-onnxruntime==1.15.1
 platformdirs==3.10.0
 
 # TODO: Constraint due to langchain, remove when that gets updated:

requirements/dev.txt

@@ -73,15 +73,15 @@ defusedxml==0.7.1
     # via nbconvert
 distlib==0.3.8
     # via virtualenv
-exceptiongroup==1.2.0
+exceptiongroup==1.2.1
     # via
     #   -c ./test.txt
     #   anyio
 executing==2.0.1
     # via stack-data
 fastjsonschema==2.19.1
     # via nbformat
-filelock==3.13.3
+filelock==3.13.4
     # via virtualenv
 fqdn==1.5.1
     # via jsonschema
@@ -91,9 +91,9 @@ httpcore==1.0.5
     # via httpx
 httpx==0.27.0
     # via jupyterlab
-identify==2.5.35
+identify==2.5.36
     # via pre-commit
-idna==3.6
+idna==3.7
     # via
     #   -c ./base.txt
     #   -c ./test.txt
@@ -134,7 +134,7 @@ jinja2==3.1.3
     #   jupyterlab
     #   jupyterlab-server
     #   nbconvert
-json5==0.9.24
+json5==0.9.25
     # via jupyterlab-server
 jsonpointer==2.4
     # via jsonschema
@@ -170,9 +170,9 @@ jupyter-core==5.7.2
     #   qtconsole
 jupyter-events==0.10.0
     # via jupyter-server
-jupyter-lsp==2.2.4
+jupyter-lsp==2.2.5
     # via jupyterlab
-jupyter-server==2.13.0
+jupyter-server==2.14.0
     # via
     #   jupyter-lsp
     #   jupyterlab
@@ -181,11 +181,11 @@ jupyter-server==2.13.0
     #   notebook-shim
 jupyter-server-terminals==0.5.3
     # via jupyter-server
-jupyterlab==4.1.5
+jupyterlab==4.1.6
     # via notebook
 jupyterlab-pygments==0.3.0
     # via nbconvert
-jupyterlab-server==2.25.4
+jupyterlab-server==2.26.0
     # via
     #   jupyterlab
     #   notebook
@@ -195,7 +195,7 @@ markupsafe==2.1.5
     # via
     #   jinja2
     #   nbconvert
-matplotlib-inline==0.1.6
+matplotlib-inline==0.1.7
     # via
     #   ipykernel
     #   ipython
@@ -216,7 +216,7 @@ nest-asyncio==1.6.0
     # via ipykernel
 nodeenv==1.8.0
     # via pre-commit
-notebook==7.1.2
+notebook==7.1.3
     # via jupyter
 notebook-shim==0.2.4
     # via
@@ -294,7 +294,7 @@ pyyaml==6.0.1
     #   -c ./test.txt
     #   jupyter-events
     #   pre-commit
-pyzmq==25.1.2
+pyzmq==26.0.2
     # via
     #   ipykernel
     #   jupyter-client
@@ -327,7 +327,7 @@ rpds-py==0.18.0
     # via
     #   jsonschema
     #   referencing
-send2trash==1.8.2
+send2trash==1.8.3
     # via jupyter-server
 six==1.16.0
     # via
@@ -368,7 +368,7 @@ tornado==6.4
     #   jupyterlab
     #   notebook
     #   terminado
-traitlets==5.14.2
+traitlets==5.14.3
     # via
     #   comm
     #   ipykernel
@@ -401,7 +401,7 @@ urllib3==1.26.18
     #   -c ./base.txt
     #   -c ./test.txt
     #   requests
-virtualenv==20.25.1
+virtualenv==20.25.3
     # via pre-commit
 wcwidth==0.2.13
     # via prompt-toolkit

requirements/extra-csv.txt

@@ -8,7 +8,7 @@ numpy==1.26.4
     # via
     #   -c ./base.txt
     #   pandas
-pandas==2.2.1
+pandas==2.2.2
     # via -r ./extra-csv.in
 python-dateutil==2.9.0.post0
     # via

requirements/extra-paddleocr.txt

@@ -39,7 +39,7 @@ cython==3.0.10
     # via unstructured-paddleocr
 et-xmlfile==1.1.0
     # via openpyxl
-flask==3.0.2
+flask==3.0.3
     # via
     #   flask-babel
     #   visualdl
@@ -49,7 +49,7 @@ fonttools==4.51.0
     # via matplotlib
 future==1.0.0
     # via bce-python-sdk
-idna==3.6
+idna==3.7
     # via
     #   -c ./base.txt
     #   requests
@@ -63,7 +63,7 @@ importlib-metadata==7.1.0
     # via flask
 importlib-resources==6.4.0
     # via matplotlib
-itsdangerous==2.1.2
+itsdangerous==2.2.0
     # via flask
 jinja2==3.1.3
     # via
@@ -129,7 +129,7 @@ packaging==23.2
     #   matplotlib
     #   scikit-image
     #   visualdl
-pandas==2.2.1
+pandas==2.2.2
     # via visualdl
 pdf2image==1.17.0
     # via unstructured-paddleocr
@@ -168,7 +168,7 @@ pytz==2024.1
     # via
     #   flask-babel
     #   pandas
-rapidfuzz==3.7.0
+rapidfuzz==3.8.1
     # via
     #   -c ./base.txt
     #   unstructured-paddleocr
@@ -188,7 +188,7 @@ scipy==1.10.1
     #   -c ././deps/constraints.txt
     #   imgaug
     #   scikit-image
-shapely==2.0.3
+shapely==2.0.4
     # via
     #   imgaug
     #   unstructured-paddleocr
@@ -200,7 +200,7 @@ six==1.16.0
     #   imgaug
     #   python-dateutil
     #   visualdl
-tifffile==2024.2.12
+tifffile==2024.4.18
     # via scikit-image
 tqdm==4.66.2
     # via

requirements/extra-pdf-image.in

@@ -9,7 +9,7 @@ pillow_heif
 pypdf
 # Do not move to constraints.in, otherwise unstructured-inference will not be upgraded
 # when unstructured library is.
-unstructured-inference==0.7.25
+unstructured-inference==0.7.27
 # unstructured fork of pytesseract that provides an interface to allow for multiple output formats
 # from one tesseract call
 unstructured.pytesseract>=0.3.12