use the prebuilt tesseract

.github/workflows/build-ruby-image.yml

@@ -12,6 +12,13 @@ jobs:
   build-ruby-image:
     runs-on: ubuntu-latest
     steps:
+      - name: Set env
+        run: >-
+          echo "TAG=${HEAD_TAG::8}" >> ${GITHUB_ENV};
+          echo ${HEAD_TAG::8}
+        env:
+          HEAD_TAG: ${{ inputs.tag || github.event.pull_request.head.sha || github.sha }}
+        shell: bash
       - name: Set up QEMU
         uses: docker/setup-qemu-action@v2
       - name: Set up Docker Buildx
@@ -48,8 +55,8 @@ jobs:
           cp ../Dockerfile .
           cp ../Gemfile* .
           cp -r ../derivative_rodeo .
-          docker build -t ghcr.io/scientist-softserv/space_stone/awsrubylayer:latest .
+          docker build -t ghcr.io/scientist-softserv/space_stone/awsrubylayer:${TAG} .
       - name: Push new docker image
         run: |
-          docker push ghcr.io/scientist-softserv/space_stone/awsrubylayer:latest
+          docker push ghcr.io/scientist-softserv/space_stone/awsrubylayer:${TAG}
         shell: bash

awslambda/Dockerfile

@@ -228,16 +228,13 @@ RUN bundle config set --local path build
 RUN bundle
 
 ###############################################################################
-# Zip all dependencies
+# The end
 ###############################################################################
 
-WORKDIR /opt
-# RUN zip -r /root/ProcessDocumentLayer.zip *
+RUN cp /opt/tesseract/share/tessdata/* /opt/share/tessdata/ && \
+  mkdir /opt/share/tessdata-best && cp -r /opt/share/tessdata/* /opt/share/tessdata-best && \
+  cd /opt/share/tessdata-best && curl -LO https://github.com/tesseract-ocr/tessdata_best/raw/main/eng.traineddata
 
-###############################################################################
-# Entrypoint: Copy zip file to host
-###############################################################################
-# TODO remove?
-# ENTRYPOINT ["/bin/cp", "/root/ProcessDocumentLayer.zip", "/output"]
+WORKDIR /opt
 
 CMD "/bin/bash"