diff --git a/Dockerfile b/Dockerfile index 1c785e9..a2a2017 100644 --- a/Dockerfile +++ b/Dockerfile @@ -60,11 +60,13 @@ RUN pip3 install --no-cache-dir --upgrade pip && \ RUN mkdir -p /var/lib/calamari-models COPY data/calamari-models/GT4HistOCR /var/lib/calamari-models/GT4HistOCR RUN mkdir -p $TESSDATA_PREFIX +RUN curl -sSL -O https://qurator-data.de/mirror/github.com/tesseract-ocr/tessdata_best/archive/$TESSDATA_BEST_VERSION.tar.gz && \ + tar xzf $TESSDATA_BEST_VERSION.tar.gz && \ + mv tessdata_best-$TESSDATA_BEST_VERSION/* $TESSDATA_PREFIX/ && \ + rm -rf $TESSDATA_BEST_VERSION.tar.gz COPY data/tesseract-models/GT4HistOCR/GT4HistOCR_2000000.traineddata $TESSDATA_PREFIX/ COPY data/textline_detection /var/lib/textline_detection -RUN tesseract --list-langs - # Install requirements COPY requirements.txt /tmp/ diff --git a/requirements.txt b/requirements.txt index 514472e..df1f69a 100644 --- a/requirements.txt +++ b/requirements.txt @@ -6,6 +6,6 @@ ocrd_tesserocr >= 0.8.1 ocrd_calamari >= 0.0.6 -https://github.com/qurator-spk/sbb_textline_detector/archive/689822c.tar.gz +https://github.com/qurator-spk/sbb_textline_detector/archive/8618be2.tar.gz -https://github.com/qurator-spk/dinglehopper/archive/c305539.tar.gz +https://github.com/qurator-spk/dinglehopper/archive/745095e.tar.gz