compile tesseract in container

.gitlab-ci.yml

@@ -9,36 +9,62 @@ variables:
 stages:
   - build
   - push
+  - clean
 
 before_script:
   - docker login -u gitlab-ci-token -p $CI_JOB_TOKEN $CI_REGISTRY
 
-Build:
+build_image:
   script:
-    - docker build --pull -t $CI_REGISTRY_IMAGE:tmp .
+    - docker build -t $INTERMEDIATE_IMAGE_TAG .
-    - docker push $CI_REGISTRY_IMAGE:tmp
+    - docker push $INTERMEDIATE_IMAGE_TAG
   stage: build
   tags:
-  - docker
+    - docker
+  variables:
+    INTERMEDIATE_IMAGE_TAG: $CI_REGISTRY_IMAGE:$CI_COMMIT_REF_SLUG
 
-Push latest:
+push_master:
   only:
     - master
   script:
-    - docker pull $CI_REGISTRY_IMAGE:tmp
+    - docker pull $INTERMEDIATE_IMAGE_TAG
-    - docker tag $CI_REGISTRY_IMAGE:tmp $CI_REGISTRY_IMAGE:latest
+    - docker tag $INTERMEDIATE_IMAGE_TAG $IMAGE_TAG
-    - docker push $CI_REGISTRY_IMAGE:latest
+    - docker push $IMAGE_TAG
   stage: push
   tags:
-  - docker
+    - docker
+  variables:
+    IMAGE_TAG: $CI_REGISTRY_IMAGE:latest
+    INTERMEDIATE_IMAGE_TAG: $CI_REGISTRY_IMAGE:$CI_COMMIT_REF_SLUG
 
-Push tag:
+push_other:
+  except:
+    - master
   only:
+    - branches
     - tags
   script:
-    - docker pull $CI_REGISTRY_IMAGE:tmp
+    - docker pull $INTERMEDIATE_IMAGE_TAG
-    - docker tag $CI_REGISTRY_IMAGE:tmp $CI_REGISTRY_IMAGE:$CI_COMMIT_REF_NAME
+    - docker tag $INTERMEDIATE_IMAGE_TAG $IMAGE_TAG
-    - docker push $CI_REGISTRY_IMAGE:$CI_COMMIT_REF_NAME
+    - docker push $IMAGE_TAG
   stage: push
   tags:
-  - docker
+    - docker
+  variables:
+    IMAGE_TAG: $CI_REGISTRY_IMAGE:CI_COMMIT_REF_NAME
+    INTERMEDIATE_IMAGE_TAG: $CI_REGISTRY_IMAGE:$CI_COMMIT_REF_SLUG
+
+delete_image:
+  before_script:
+    - apk add --no-cache curl
+    - curl --fail --show-error --location "https://github.com/genuinetools/reg/releases/download/v$REG_VERSION/reg-linux-amd64" --output /usr/local/bin/reg
+    - echo "$REG_SHA256  /usr/local/bin/reg" | sha256sum -c -
+    - chmod a+x /usr/local/bin/reg
+  script:
+    - /usr/local/bin/reg rm -d --auth-url $CI_REGISTRY -u $CI_REGISTRY_USER -p $CI_REGISTRY_PASSWORD $INTERMEDIATE_IMAGE_TAG
+  stage: clean
+  variables:
+    INTERMEDIATE_IMAGE_TAG: $CI_REGISTRY_IMAGE:$CI_COMMIT_REF_SLUG
+    REG_SHA256: ade837fc5224acd8c34732bf54a94f579b47851cc6a7fd5899a98386b782e228
+    REG_VERSION: 0.16.1

Dockerfile

@@ -1,73 +1,88 @@
 FROM debian:10-slim
 
 
-LABEL maintainer="inf_sfb1288@lists.uni-bielefeld.de"
+LABEL authors="Patrick Jentsch <p.jentsch@uni-bielefeld.de>, Stephan Porada <sporada@uni-bielefeld.de>"
 
 
 ENV LANG=C.UTF-8
 
 
-RUN apt-get update \
+RUN apt-get update
+
+
+## Install pyFlow ##
+ENV PYFLOW_RELEASE=1.1.20
+ADD "https://github.com/Illumina/pyflow/releases/download/v${PYFLOW_RELEASE}/pyflow-${PYFLOW_RELEASE}.tar.gz" .
+RUN tar -xzf "pyflow-${PYFLOW_RELEASE}.tar.gz" \
+ && cd "pyflow-${PYFLOW_RELEASE}" \
  && apt-get install -y --no-install-recommends \
-      apt-transport-https \
-      build-essential \
-      ca-certificates \
-      gnupg2 \
-      ghostscript \
       python2.7 \
-      python3.7 \
+ && python2.7 setup.py build install \
-      wget \
+ && cd .. \
-      zip
+ && rm -r "pyflow-${PYFLOW_RELEASE}" "pyflow-${PYFLOW_RELEASE}.tar.gz"
 
 
-ENV OCROPY_VERSION 1.3.3
+## Install ocropy ##
-ADD "https://github.com/tmbdev/ocropy/archive/v${OCROPY_VERSION}.tar.gz" .
+ENV OCROPY_RELEASE 1.3.3
-RUN tar -xzf "v${OCROPY_VERSION}.tar.gz" \
+ADD "https://github.com/tmbdev/ocropy/archive/v${OCROPY_RELEASE}.tar.gz" .
- && cd "ocropy-${OCROPY_VERSION}" \
+RUN tar -xzf "v${OCROPY_RELEASE}.tar.gz" \
- && apt-get update \
+ && cd "ocropy-${OCROPY_RELEASE}" \
  && apt-get install -y --no-install-recommends \
       python-pil \
       python-tk \
       $(cat PACKAGES) \
  && python2.7 setup.py install \
  && cd .. \
- && rm -rf \
+ && rm -r "ocropy-${OCROPY_RELEASE}" "v${OCROPY_RELEASE}.tar.gz"
-      "ocropy-${OCROPY_VERSION}" \
-      "v${OCROPY_VERSION}.tar.gz"
 
 
-ENV PYFLOW_VERSION=1.1.20
+## Install Tesseract OCR ##
-ADD "https://github.com/Illumina/pyflow/releases/download/v${PYFLOW_VERSION}/pyflow-${PYFLOW_VERSION}.tar.gz" .
+ENV TESSERACT_RELEASE=4.1.1
-RUN tar -xzf "pyflow-${PYFLOW_VERSION}.tar.gz" \
+ADD "https://github.com/tesseract-ocr/tesseract/archive/${TESSERACT_RELEASE}.tar.gz" .
- && cd "pyflow-${PYFLOW_VERSION}" \
+RUN tar -xzf "${TESSERACT_RELEASE}.tar.gz" \
- && python2.7 setup.py build install \
+ && cd "tesseract-${TESSERACT_RELEASE}" \
- && cd .. \
- && rm -rf \
-      "pyflow-${PYFLOW_VERSION}" \
-      "pyflow-${PYFLOW_VERSION}.tar.gz"
-
-
-RUN echo "deb https://notesalexp.org/tesseract-ocr/buster/ buster main" >> /etc/apt/sources.list \
- && wget -O - https://notesalexp.org/debian/alexp_key.asc | apt-key add - \
- && apt-get update \
  && apt-get install -y --no-install-recommends \
-      tesseract-ocr  \
+      autoconf \
-      tesseract-ocr-deu \
+      automake \
-      tesseract-ocr-eng \
+      g++ \
-      tesseract-ocr-enm \
+      libjpeg62-turbo-dev \
-      tesseract-ocr-fra \
+      libleptonica-dev \
-      tesseract-ocr-frk \
+      libtiff5-dev \
-      tesseract-ocr-frm \
+      libtool \
-      tesseract-ocr-ita \
+      libpng-dev \
-      tesseract-ocr-por \
+      make \
-      tesseract-ocr-spa
+      pkg-config \
+      zlib1g-dev \
+ && ./autogen.sh \
+ && ./configure \
+ && make \
+ && make install \
+ && ldconfig \
+ && cd - > /dev/null \
+ && rm -r "tesseract-${TESSERACT_RELEASE}" "${TESSERACT_RELEASE}.tar.gz"
+ADD "https://github.com/tesseract-ocr/tessdata_best/raw/master/deu.traineddata" \
+    "https://github.com/tesseract-ocr/tessdata_best/raw/master/eng.traineddata" \
+    "https://github.com/tesseract-ocr/tessdata_best/raw/master/enm.traineddata" \
+    "https://github.com/tesseract-ocr/tessdata_best/raw/master/fra.traineddata" \
+    "https://github.com/tesseract-ocr/tessdata_best/raw/master/frk.traineddata" \
+    "https://github.com/tesseract-ocr/tessdata_best/raw/master/frm.traineddata" \
+    "https://github.com/tesseract-ocr/tessdata_best/raw/master/ita.traineddata" \
+    "https://github.com/tesseract-ocr/tessdata_best/raw/master/por.traineddata" \
+    "https://github.com/tesseract-ocr/tessdata_best/raw/master/spa.traineddata" \
+    "/usr/local/share/tessdata/"
+RUN chmod 644 /usr/local/share/tessdata/*.traineddata
 
 
-RUN rm -rf /var/lib/apt/lists/*
+## Install Pipeline ##
+RUN apt-get install -y --no-install-recommends \
+      ghostscript \
+      python3.7 \
+      zip
+COPY "hocrtotei" "ocr" "/usr/local/bin/"
 
 
-COPY hocrtotei /usr/local/bin
+## Cleanup ##
-COPY ocr /usr/local/bin
+RUN rm -r /var/lib/apt/lists/*
 
 
 ENTRYPOINT ["ocr"]

ocr

@@ -6,7 +6,8 @@
 ocr
 
 Usage:  For usage instructions run with option --help
-Author: Patrick Jentsch <p.jentsch@uni-bielefeld.de>
+Authors: Patrick Jentsch <p.jentsch@uni-bielefeld.de
+         Stephan Porada <sporada@uni-bielefeld.de>
 """