TERMUX_PKG_HOMEPAGE=https://github.com/tesseract-ocr/tesseract
TERMUX_PKG_DESCRIPTION="Tesseract is probably the most accurate open source OCR engine available"
TERMUX_PKG_VERSION=3.04.01
-TERMUX_PKG_BUILD_REVISION=1
TERMUX_PKG_DEPENDS="libtool, libuuid, leptonica"
TERMUX_PKG_SRCURL=https://github.com/tesseract-ocr/tesseract/archive/${TERMUX_PKG_VERSION}.tar.gz
TERMUX_PKG_FOLDERNAME=tesseract-${TERMUX_PKG_VERSION}
# download english trained data
cd "${TERMUX_PREFIX}/share/tessdata"
rm -f eng.*
- wget https://raw.githubusercontent.com/tesseract-ocr/tessdata/master/eng.cube.bigrams
- wget https://raw.githubusercontent.com/tesseract-ocr/tessdata/master/eng.cube.fold
- wget https://raw.githubusercontent.com/tesseract-ocr/tessdata/master/eng.cube.lm
- wget https://raw.githubusercontent.com/tesseract-ocr/tessdata/master/eng.cube.nn
- wget https://raw.githubusercontent.com/tesseract-ocr/tessdata/master/eng.cube.params
- wget https://raw.githubusercontent.com/tesseract-ocr/tessdata/master/eng.cube.size
- wget https://raw.githubusercontent.com/tesseract-ocr/tessdata/master/eng.cube.word-freq
- wget https://raw.githubusercontent.com/tesseract-ocr/tessdata/master/eng.tesseract_cube.nn
- wget https://raw.githubusercontent.com/tesseract-ocr/tessdata/master/eng.traineddata
+ for f in cube.{bigrams,fold,lm,nn,params,size,word-freq} tesseract_cube.nn traineddata; do
+ f=eng.$f
+ # From the tessdata README: "These language data files only work with
+ # Tesseract 4. They are based on the sources in tesseract-ocr/langdata on GitHub.
+ # Get language data files for Tesseract 3.04 or 3.05 from the 3.04 tree."
+ termux_download https://raw.githubusercontent.com/tesseract-ocr/tessdata/3.04.00/$f $f
+ done
}