almet/dangerzone
1
0
Fork 0
mirror of https://github.com/freedomofpress/dangerzone.git synced 2025-04-28 18:02:38 +02:00
Code Issues Projects Releases Packages Wiki Activity Actions 11

Support english-language OCR

Browse source
This commit is contained in:
Micah Lee 2020-01-07 09:52:04 -08:00
parent db23ced6c5
commit ab425feef5
No known key found for this signature in database
GPG key ID: 403C2657CD994F73
2 changed files with 9 additions and 3 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

4
share/container/Containerfile
View file

@ -1,7 +1,9 @@
FROM ubuntu:18.04 FROM ubuntu:18.04
RUN apt-get update && \ RUN apt-get update && \
apt-get install -y file poppler-utils imagemagick ghostscript tesseract-ocr tesseract-ocr-all apt-get install -y file poppler-utils imagemagick ghostscript tesseract-ocr
# TODO: when we support OCR in other languages, we need tesseract-ocr-all
# Fix imagemagick policy to allow writing PDFs # Fix imagemagick policy to allow writing PDFs
RUN sed -i '/rights="none" pattern="PDF"/c\<policy domain="coder" rights="read|write" pattern="PDF" />' /etc/ImageMagick-6/policy.xml RUN sed -i '/rights="none" pattern="PDF"/c\<policy domain="coder" rights="read|write" pattern="PDF" />' /etc/ImageMagick-6/policy.xml

8
share/container/pixels-to-pdf
View file

@ -19,6 +19,7 @@ for RGB_FILENAME in $(find /dangerzone/page-*.rgb); do
WIDTH_FILENAME=${FILENAME_BASE}.width WIDTH_FILENAME=${FILENAME_BASE}.width
HEIGHT_FILENAME=${FILENAME_BASE}.height HEIGHT_FILENAME=${FILENAME_BASE}.height
PNG_FILENAME=/tmp/$(basename $FILENAME_BASE).png PNG_FILENAME=/tmp/$(basename $FILENAME_BASE).png
OCR_FILENAME=/tmp/$(basename $FILENAME_BASE)
PDF_FILENAME=/tmp/$(basename $FILENAME_BASE).pdf PDF_FILENAME=/tmp/$(basename $FILENAME_BASE).pdf
echo "Converting page $PAGE from pixels to PNG" echo "Converting page $PAGE from pixels to PNG"
@ -31,13 +32,16 @@ for RGB_FILENAME in $(find /dangerzone/page-*.rgb); do
die "Page $PAGE conversion failed: $CONVERT_MSGS" die "Page $PAGE conversion failed: $CONVERT_MSGS"
fi fi
echo "Converting page $PAGE from PNG to PDF" echo "Converting page $PAGE from PNG to searchable PDF"
CONVERT_MSGS=$(convert "$PNG_FILENAME" "$PDF_FILENAME" 2>&1) CONVERT_MSGS=$(tesseract $PNG_FILENAME $OCR_FILENAME pdf 2>&1)
if [ $? -ne 0 ]; then if [ $? -ne 0 ]; then
die "Page $PAGE conversion failed: $CONVERT_MSGS" die "Page $PAGE conversion failed: $CONVERT_MSGS"
fi fi
echo
done done
echo
echo "Merging $NUM_PAGES pages into a single PDF" echo "Merging $NUM_PAGES pages into a single PDF"
# Put PDF filenames into an array # Put PDF filenames into an array