tesseract 3.05.01, split tessaract with tesseract-data

fix #52

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')
+depends=('glibc' 'libwebp')
-makedepends=('giflib' 'gnuplot' 'libjpeg' 'libpng' 'libtiff' 'zlib' '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}

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

tesseract/PKGBUILD

@@ -1,58 +1,39 @@
-# contributor (x86_64): Giuseppe Calà <jiveaxe@gmail.com>
+# contributor (x86_64): Giuseppe Calà  <jiveaxe@gmail.com>
 
-_langs=(afr amh ara asm aze aze_cyrl bel ben bod bos bul cat ceb ces
+_langs=(afr ara aze bel ben bul cat ces chi_sim chi_tra chr dan deu ell eng enm
-chi_sim chi_tra chr cym dan dan_frak deu deu_frak dzo ell eng enm epo
+    epo equ est eus fin fra frk frm glg grc heb hin hrv hun ind isl ita ita_old
-equ est eus fas fin fra frk frm gle glg grc guj hat heb hin hrv hun
+    jpn kan kor lav lit mal mkd mlt msa nld nor pol por ron rus slk slv spa
-iku ind isl ita ita_old jav jpn kan kat kat_old kaz khm kir kor kur
+    spa_old sqi srp swa swe tam tel tgl tha tur ukr vie)
-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
+pkgname=tesseract
-pkgname=(tesseract $(for l in ${_langs[@]}; do echo tesseract-data-${l}; done))
+pkgver=3.05.01
-pkgver=3.04.00
+pkgrel=3
-pkgrel=2
+pkgdesc="An OCR program"
-pkgdesc="An OCR programm"
 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'
+sha256sums=('05898f93c5d057fada49b9a116fc86ad9310ff1726a0f499c3e5211b3af47ec1'
-         'b25e830d203af5c863081af3f684b53a'
+            'e19f2ae860792fdf372cf48d8ce70ae5da3c4052962fe22e9de1f680c374bb0e')
-         'd7c06843a771f30fb64b4109a1b059f9')
+
 build() {
-  cd $srcdir/$pkgname-${pkgver}
+  cd "$srcdir"/$pkgname-${pkgver}
+  [ -x configure ] || ./autogen.sh
   [ -f Makefile ] || ./configure --prefix=/usr
   make
   make training
 }
 
-package_tesseract() {
+package() {
-  optdepends=('icu' 'cairo' 'pango'
+  cd "$srcdir"/$pkgname-${pkgver}
-	    $(for l in ${_langs[@]}; do echo tesseract-data-${l}; done))
+  make DESTDIR="$pkgdir" install
-
+  make DESTDIR="$pkgdir" training-install
-  cd $srcdir/$pkgname-${pkgver}
+  mkdir -p "$pkgdir"/usr/share/tessdata
-  make DESTDIR=$pkgdir install
+  install -Dm0644 "$srcdir"/osd.traineddata "$pkgdir"/usr/share/tessdata/osd.traineddata
-  make DESTDIR=$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