diff --git a/Dockerfile b/Dockerfile index b4382cf..3a463fc 100644 --- a/Dockerfile +++ b/Dockerfile @@ -6,8 +6,19 @@ ENV CGO_ENABLED=0 RUN go build -o pdftoolbox FROM ubuntu -RUN apt-get -y update && rm -rf /var/lib/apt/lists/* -RUN cp gs /usr/bin/gs +RUN apt-get -y update && apt-get install -y curl && rm -rf /var/lib/apt/lists/* +COPY gs /usr/bin/gs +RUN make -p /usr/local/share/tessdata +RUN curl https://github.com/tesseract-ocr/tessdata_best/raw/main/eng.traineddata -o /usr/local/share/tessdata/eng.traineddata +RUN curl https://github.com/tesseract-ocr/tessdata_best/raw/main/chi_sim.traineddata -o /usr/local/share/tessdata/chi_sim.traineddata +RUN curl https://github.com/tesseract-ocr/tessdata_best/raw/main/chi_sim_vert.traineddata -o /usr/local/share/tessdata/chi_sim_vert.traineddata +RUN curl https://github.com/tesseract-ocr/tessdata_best/raw/main/chi_tra.traineddata -o /usr/local/share/tessdata/chi_tra.traineddata +RUN curl https://github.com/tesseract-ocr/tessdata_best/raw/main/chi_tra_vert.traineddata -o /usr/local/share/tessdata/chi_tra_vert.traineddata +RUN curl https://github.com/tesseract-ocr/tessdata_best/raw/main/jpn.traineddata -o /usr/local/share/tessdata/jpn.traineddata +RUN curl https://github.com/tesseract-ocr/tessdata_best/raw/main/jpn_vert.traineddata -o /usr/local/share/tessdata/jpn_vert.traineddata +RUN curl https://github.com/tesseract-ocr/tessdata_best/raw/main/kor.traineddata -o /usr/local/share/tessdata/kor.traineddata +RUN curl https://github.com/tesseract-ocr/tessdata_best/raw/main/kor_vert.traineddata -o /usr/local/share/tessdata/kor_vert.traineddata +RUN apt-get -y remove curl RUN mkdir -p /opt/pdftoolbox RUN mkdir -p /opt/pdftoolbox/input RUN mkdir -p /opt/pdftoolbox/output