ci: Adds workflow to publish docker builds (#377)

Author: ryannikolaidis

.github/workflows/docker-publish.yml

@@ -0,0 +1,46 @@
+name: Build And Push Docker Image
+
+on:
+  push:
+    branches:
+      - main
+
+env:
+  DOCKER_REGISTRY: "quay.io"
+  DOCKER_NAMESPACE: "unstructured-io"
+  IMAGE_PLATFORMS: linux/amd64
+  PACKAGE: "unstructured"
+  PIP_VERSION: "22.2.1"
+  BUILD_TYPE_TAG_SUFFIX: ""
+
+jobs:
+  build:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Checkout code
+        uses: actions/checkout@v3
+
+      - name: Set up QEMU
+        uses: docker/setup-qemu-action@v1
+
+      - name: Set up Docker Buildx
+        uses: docker/setup-buildx-action@v1
+
+      - name: Login to Quay.io
+        uses: docker/login-action@v1
+        with:
+          registry: quay.io
+          username: ${{ secrets.QUAY_IO_ROBOT_USERNAME }}
+          password: ${{ secrets.QUAY_IO_ROBOT_TOKEN }}
+
+      # TODO(rniko): add a step to test the built image before pushing. Will add after https://unstructured-ai.atlassian.net/browse/CORE-745 is done
+      - name: Build and push Docker image
+        run: |
+          VERSION=$(grep -Po '(?<=__version__ = ")[^"]*' unstructured/__version__.py)
+          GIT_SHA=$(git rev-parse --short HEAD)
+          IMAGE_NAME=${{ env.PACKAGE }}
+          docker buildx build --platform=${{ env.IMAGE_PLATFORMS }} --provenance=false --push \
+          --build-arg PIP_VERSION=${{ env.PIP_VERSION }} \
+          -t ${{ env.DOCKER_REGISTRY }}/${{ env.DOCKER_NAMESPACE }}/${IMAGE_NAME}:${GIT_SHA} \
+          -t ${{ env.DOCKER_REGISTRY }}/${{ env.DOCKER_NAMESPACE }}/${IMAGE_NAME}:${VERSION} \
+          -t ${{ env.DOCKER_REGISTRY }}/${{ env.DOCKER_NAMESPACE }}/${IMAGE_NAME}:latest .

Dockerfile

@@ -8,6 +8,12 @@ ARG UNSTRUCTURED
 RUN yum -y update && \
   yum -y install poppler-utils xz-devel which
 
+# Enable the EPEL repository
+RUN yum install -y epel-release && yum clean all
+
+# Install pandoc
+RUN yum install -y pandoc && yum clean all
+
 # Note(austin) Get a recent tesseract from this repo
 # See https://tesseract-ocr.github.io/tessdoc/Installation.html
 # PDF and images:
@@ -44,6 +50,7 @@ ENV PATH="/home/usr/.local/bin:${PATH}"
 COPY example-docs example-docs
 
 COPY requirements/base.txt requirements-base.txt
+COPY requirements/test.txt requirements-test.txt
 COPY requirements/huggingface.txt requirements-huggingface.txt
 COPY requirements/dev.txt requirements-dev.txt
 # PDFs and images
@@ -52,6 +59,7 @@ COPY requirements/local-inference.txt requirements-local-inference.txt
 
 RUN python3.8 -m pip install pip==${PIP_VERSION} \
   && pip install --no-cache -r requirements-base.txt \
+  && pip install --no-cache -r requirements-test.txt \
   && pip install --no-cache -r requirements-huggingface.txt \
   && pip install --no-cache -r requirements-dev.txt \
   # PDFs and images