From dcb286ebcc875b76f4db34c986aa6e6867a6ee67 Mon Sep 17 00:00:00 2001 From: AlmAck Date: Thu, 22 Mar 2018 23:45:21 +0100 Subject: [PATCH] tesseract 3.05.01, split tessaract with tesseract-data fix #52 --- leptonica/PKGBUILD | 9 +++--- tesseract-data/PKGBUILD | 40 ++++++++++++++++++++++++ tesseract/PKGBUILD | 69 +++++++++++++++-------------------------- 3 files changed, 69 insertions(+), 49 deletions(-) create mode 100644 tesseract-data/PKGBUILD diff --git a/leptonica/PKGBUILD b/leptonica/PKGBUILD index 3b8984634..f48d2121c 100644 --- a/leptonica/PKGBUILD +++ b/leptonica/PKGBUILD @@ -1,12 +1,12 @@ pkgname=leptonica -pkgver=1.73 +pkgver=1.75.3 pkgrel=1 pkgdesc="Software that is broadly useful for image processing and image analysis applications" arch=('x86_64') url="http://www.leptonica.com/" license=('custom') -depends=('glibc') -makedepends=('giflib' 'gnuplot' 'libjpeg' 'libpng' 'libtiff' 'zlib' 'libwebp') +depends=('glibc' 'libwebp') +makedepends=('giflib' 'gnuplot' 'libjpeg' 'libpng' 'libtiff' 'zlib') categories=('graphics') optdepends=('giflib: for supporting gif files' 'gnuplot: gnuplot support' @@ -14,9 +14,8 @@ optdepends=('giflib: for supporting gif files' 'libpng: for supporting png files' 'libtiff: for supporting tiff files' 'zlib: for supporting compressed files') - 'libwebp: for supporting webp files' source=(http://www.leptonica.com/source/leptonica-${pkgver}.tar.gz) -sha512sums=('4808474d5941a91fbf542a22fed26ddc33f9c4baf2699096acf5e894b7e4e84889d165373b7d1df391a68cda8176113cf122f15203a9e38d3059e17140efc575') +sha512sums=('ba2586e76634afb9be6699e4261e55078dd176ee2d0674aedfe49f7f0736ddedadd6961bcce23391d47dc24511cf5ab4b451544f3349c2d060da5f789dc5ab3f') build() { cd ${srcdir}/leptonica-${pkgver} diff --git a/tesseract-data/PKGBUILD b/tesseract-data/PKGBUILD new file mode 100644 index 000000000..939d555d5 --- /dev/null +++ b/tesseract-data/PKGBUILD @@ -0,0 +1,40 @@ + +_langs=(afr amh ara asm aze aze_cyrl bel ben bod bos bul cat ceb ces +chi_sim chi_tra chr cym dan dan_frak deu deu_frak dzo ell eng enm epo +equ est eus fas fin fra frk frm gle glg grc guj hat heb hin hrv hun +iku ind isl ita ita_old jav jpn kan kat kat_old kaz khm kir kor kur +lao lat lav lit mal mar mkd mlt msa mya nep nld nor ori pan pol +por pus ron rus san sin slk slk_frak slv spa spa_old sqi srp srp_latn +swa swe syr tam tel tgk tgl tha tir tur uig ukr urd uzb uzb_cyrl vie yid) + +pkgbase=tesseract-data +pkgname=($(for l in ${_langs[@]}; do echo tesseract-data-${l}; done)) +epoch=1 +pkgver=3.04.00 +pkgrel=1 +pkgdesc="An OCR programm" +arch=(any) +url="https://github.com/tesseract-ocr/tessdata" +license=("APACHE") +depends=() +source=($pkgbase-$pkgver.tar.gz::https://github.com/tesseract-ocr/tessdata/archive/$pkgver.tar.gz) +sha256sums=('5dcb37198336b6953843b461ee535df1401b41008d550fc9e43d0edabca7adb1') + +build() { + true +} + +# Declare the package functions for data +for l in ${_langs[@]}; do + eval " +package_tesseract-data-${l}(){ + pkgdesc=\"Tesseract OCR data ($l)\" + depends=('tesseract') + groups=('tesseract-data') + + mkdir -p \$pkgdir/usr/share/tessdata + cp \$srcdir/tessdata-$pkgver/${l}.* \$pkgdir/usr/share/tessdata/ + find \$pkgdir/usr/share/tessdata -type f -exec chmod 0644 {} \; +} + " +done diff --git a/tesseract/PKGBUILD b/tesseract/PKGBUILD index 6be4a45d5..bed7d44e9 100644 --- a/tesseract/PKGBUILD +++ b/tesseract/PKGBUILD @@ -1,58 +1,39 @@ -# contributor (x86_64): Giuseppe Calà +# contributor (x86_64): Giuseppe Calà -_langs=(afr amh ara asm aze aze_cyrl bel ben bod bos bul cat ceb ces -chi_sim chi_tra chr cym dan dan_frak deu deu_frak dzo ell eng enm epo -equ est eus fas fin fra frk frm gle glg grc guj hat heb hin hrv hun -iku ind isl ita ita_old jav jpn kan kat kat_old kaz khm kir kor kur -lao lat lav lit mal mar mkd mlt msa mya nep nld nor ori pan pol -por pus ron rus san sin slk slk_frak slv spa spa_old sqi srp srp_latn -swa swe syr tam tel tgk tgl tha tir tur uig ukr urd uzb uzb_cyrl vie yid) +_langs=(afr ara aze bel ben bul cat ces chi_sim chi_tra chr dan deu ell eng enm + epo equ est eus fin fra frk frm glg grc heb hin hrv hun ind isl ita ita_old + jpn kan kor lav lit mal mkd mlt msa nld nor pol por ron rus slk slv spa + spa_old sqi srp swa swe tam tel tgl tha tur ukr vie) -pkgbase=tesseract -pkgname=(tesseract $(for l in ${_langs[@]}; do echo tesseract-data-${l}; done)) -pkgver=3.04.00 -pkgrel=2 -pkgdesc="An OCR programm" +pkgname=tesseract +pkgver=3.05.01 +pkgrel=3 +pkgdesc="An OCR program" arch=(x86_64) -url="http://code.google.com/p/tesseract-ocr" +url="https://github.com/tesseract-ocr/tesseract" license=("APACHE") -depends=(libwebp libpng libtiff libjpeg zlib giflib gcc-libs leptonica) +depends=(libpng libtiff libjpeg-turbo zlib giflib gcc-libs leptonica) makedepends=(icu cairo pango) +install=tesseract.install +optdepends=('icu' 'cairo' 'pango' + $(for l in ${_langs[@]}; do echo tesseract-data-${l}; done)) source=($pkgname-$pkgver.tar.gz::https://github.com/tesseract-ocr/tesseract/archive/$pkgver.tar.gz - $pkgbase-data-$pkgver.tar.gz::https://github.com/tesseract-ocr/tessdata/archive/$pkgver.tar.gz https://github.com/tesseract-ocr/tessdata/raw/master/osd.traineddata) -md5sums=('078130b9c7d28c558a0e49d432505864' - 'b25e830d203af5c863081af3f684b53a' - 'd7c06843a771f30fb64b4109a1b059f9') +sha256sums=('05898f93c5d057fada49b9a116fc86ad9310ff1726a0f499c3e5211b3af47ec1' + 'e19f2ae860792fdf372cf48d8ce70ae5da3c4052962fe22e9de1f680c374bb0e') + build() { - cd $srcdir/$pkgname-${pkgver} + cd "$srcdir"/$pkgname-${pkgver} + [ -x configure ] || ./autogen.sh [ -f Makefile ] || ./configure --prefix=/usr make make training } -package_tesseract() { - optdepends=('icu' 'cairo' 'pango' - $(for l in ${_langs[@]}; do echo tesseract-data-${l}; done)) - - cd $srcdir/$pkgname-${pkgver} - make DESTDIR=$pkgdir install - make DESTDIR=$pkgdir training-install - mkdir -p $pkgdir/usr/share/tessdata - install -Dm0644 $srcdir/osd.traineddata $pkgdir/usr/share/tessdata/osd.traineddata +package() { + cd "$srcdir"/$pkgname-${pkgver} + make DESTDIR="$pkgdir" install + make DESTDIR="$pkgdir" training-install + mkdir -p "$pkgdir"/usr/share/tessdata + install -Dm0644 "$srcdir"/osd.traineddata "$pkgdir"/usr/share/tessdata/osd.traineddata } - -# Declare the package functions for data -for l in ${_langs[@]}; do - eval " -package_tesseract-data-${l}(){ - pkgdesc=\"Tesseract OCR data ($l)\" - depends=('tesseract') - groups=('tesseract-data') - - mkdir -p \$pkgdir/usr/share/tessdata - cp \$srcdir/tessdata-$pkgver/${l}.* \$pkgdir/usr/share/tessdata/ - find \$pkgdir/usr/share/tessdata -type f -exec chmod 0644 {} \; -} - " -done