sporada/ocr
1
0
Fork 0
mirror of https://gitlab.ub.uni-bielefeld.de/sfb1288inf/ocr.git synced 2024-12-25 17:24:18 +00:00
Code Issues Projects Releases Wiki Activity

Use new Dockerfile structure

Browse Source
This commit is contained in:
Patrick Jentsch 2020-10-08 23:09:10 +02:00
parent e941f64ee4
commit c0069d5453
1 changed files with 16 additions and 8 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

24
Dockerfile
View File

@ -10,13 +10,16 @@ ENV LANG=C.UTF-8
RUN apt-get update
# Install pipeline dependencies #
## 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-get update \
&& apt-get install --no-install-recommends --yes \
python2.7 \
&& rm -r /var/lib/apt/lists/* \
&& python2.7 setup.py build install \
&& cd .. \
&& rm -r "pyflow-${PYFLOW_RELEASE}" "pyflow-${PYFLOW_RELEASE}.tar.gz"
@ -27,10 +30,12 @@ ENV OCROPY_RELEASE 1.3.3
ADD "https://github.com/tmbdev/ocropy/archive/v${OCROPY_RELEASE}.tar.gz" .
RUN tar -xzf "v${OCROPY_RELEASE}.tar.gz" \
&& cd "ocropy-${OCROPY_RELEASE}" \
&& apt-get install -y --no-install-recommends \
&& apt-get update \
&& apt-get install --no-install-recommends --yes \
python-pil \
python-tk \
$(cat PACKAGES) \
&& rm -r /var/lib/apt/lists/* \
&& python2.7 setup.py install \
&& cd .. \
&& rm -r "ocropy-${OCROPY_RELEASE}" "v${OCROPY_RELEASE}.tar.gz"
@ -41,7 +46,8 @@ ENV TESSERACT_RELEASE=4.1.1
ADD "https://github.com/tesseract-ocr/tesseract/archive/${TESSERACT_RELEASE}.tar.gz" .
RUN tar -xzf "${TESSERACT_RELEASE}.tar.gz" \
&& cd "tesseract-${TESSERACT_RELEASE}" \
&& apt-get install -y --no-install-recommends \
&& apt-get update \
&& apt-get install --no-install-recommends --yes \
autoconf \
automake \
g++ \
@ -53,6 +59,7 @@ RUN tar -xzf "${TESSERACT_RELEASE}.tar.gz" \
make \
pkg-config \
zlib1g-dev \
&& rm -r /var/lib/apt/lists/* \
&& ./autogen.sh \
&& ./configure \
&& make \
@ -73,18 +80,19 @@ ADD "https://github.com/tesseract-ocr/tessdata_best/raw/master/deu.traineddata"
RUN chmod 644 /usr/local/share/tessdata/*.traineddata
## Install Pipeline ##
RUN apt-get install -y --no-install-recommends \
## Further dependencies ##
RUN apt-get update \
&& apt-get install --no-install-recommends --yes \
ghostscript \
python-pip \
python3.7 \
zip \
&& rm -r /var/lib/apt/lists/* \
&& pip install natsort
COPY "hocrtotei" "ocr" "/usr/local/bin/"
## Cleanup ##
RUN rm -r /var/lib/apt/lists/*
## Install Pipeline ##
COPY hocrtotei ocr /usr/local/bin/
ENTRYPOINT ["ocr"]