graphics/tesseract: Updated for version 3.04.00.

Signed-off-by: Willy Sudiarto Raharjo <willysr@slackbuilds.org>

4 files changed, 21 insertions, 27 deletions

diff --git a/graphics/tesseract/README b/graphics/tesseract/README
index 5feadb549d46..f159608df6ea 100644
--- a/graphics/tesseract/README
+++ b/graphics/tesseract/README
@@ -5,14 +5,8 @@ by UNLV. It was open-sourced by HP and UNLV in 2005.
 You will need to get one of the language packs in order to do anything
 useful with tesseract, and that language pack tarball should be present
 in the same directory as the SlackBuild script when the package is created.
-See http://code.google.com/p/tesseract-ocr/downloads/list for a list of
-all available language packs.  Note that you can install more than one
-(or even all) of the language packs, as they do not conflict with each
-other.  The build script defaults to use English, but this is easily 
-changed by passing an alternate value on the command line.
-
-Here is the relevant code from the build script:
-  # Language pack(s) to use
-  # We'll install English by default, but you can pass another one.
-  # Edit the LANGNAM variable to switch to (or add) another language
-  # see https://code.google.com/p/tesseract-ocr/downloads/list for the list
+See the website for a list of all available language packs.  Note that you
+can install more than one (or even all) of the language packs, as they do
+not conflict with each other. The build script defaults to use English, but
+this is easily changed by passing an alternate value on the command line:
+edit the LANGNAM variable to switch to (or add) another language.
diff --git a/graphics/tesseract/slack-desc b/graphics/tesseract/slack-desc
index addc34c30ad7..5a90f2564cca 100644
--- a/graphics/tesseract/slack-desc
+++ b/graphics/tesseract/slack-desc
@@ -13,7 +13,7 @@ tesseract: at HP between 1985 and 1995. In 1995, this engine was among the
 tesseract: top 3 evaluated by UNLV. It was open-sourced by HP and UNLV in
 tesseract: 2005.
 tesseract:
-tesseract: http://code.google.com/p/tesseract-ocr/
+tesseract:
 tesseract:
 tesseract:
 tesseract:
diff --git a/graphics/tesseract/tesseract.SlackBuild b/graphics/tesseract/tesseract.SlackBuild
index 90ba6f357786..20be19b84529 100644
--- a/graphics/tesseract/tesseract.SlackBuild
+++ b/graphics/tesseract/tesseract.SlackBuild
@@ -2,7 +2,7 @@
 # Slackware build script for tesseract
 
 # Copyright 2009-2010 Pierre Cazenave <pwcazenave {at} gmail [dot] com>
-# Copyright 2013 LukenShiro, Italy
+# Copyright 2013-2015 LukenShiro, Italy
 # All rights reserved.
 #
 # Redistribution and use of this script, with or without modification, is
@@ -31,7 +31,7 @@
 # Modified by Willy Sudiarto Raharjo <willysr@slackware-id.org>
 
 PRGNAM=tesseract
-VERSION=${VERSION:-3.02.02}
+VERSION=${VERSION:-3.04.00}
 BUILD=${BUILD:-1}
 TAG=${TAG:-_SBo}
 
@@ -71,8 +71,7 @@ DOCFILES="AUTHORS COPYING ChangeLog INSTALL NEWS README ReleaseNotes"
 # slk ron por pol nor nld msa mlt mkd mal lit lav kor kan ita_old ita isl
 # ind chr hun hrv hin heb glg frm frk fra fin eus est equ epo enm ell due
 # dan ces cat bul ben bel aze ara afr jpn chi_sim chi_tra rus vie
-# see https://code.google.com/p/tesseract-ocr/downloads/list for complete list
-LANGVER=${LANGVER:-3.02}
+# see https://github.com/tesseract-ocr/tessdata/ for a list.
 LANGNAM=${LANGNAM:-"eng"}
 
 set -e
@@ -80,7 +79,7 @@ set -e
 rm -rf $TMP/$PRGNAM-$VERSION $PKG
 mkdir -p $TMP $PKG $OUTPUT
 cd $TMP
-tar xvf $CWD/$PRGNAM-$VERSION.tar.gz
+tar xvf $CWD/$PRGNAM-$VERSION.tar.gz || tar xvf $CWD/$VERSION.tar.gz
 cd $PRGNAM-$VERSION
 chown -R root:root .
 find -L . \
@@ -102,13 +101,14 @@ CFLAGS="$SLKCFLAGS" \
 
 make
 make install DESTDIR=$PKG
+make training
+make training-install DESTDIR=$PKG
 
-# Extract language name described above;
+# Install language name described above;
 # language files are not automatically installed.
 mkdir -p $PKG/usr/share/tessdata
-for LNG in $LANGNAM; \
- do tar xvf $CWD/tesseract-ocr-$LANGVER.$LNG.tar.gz ; \
- mv tesseract-ocr/tessdata/* $PKG/usr/share/tessdata/
+for LNG in $LANGNAM ; \
+  do cp $CWD/$LNG.traineddata $PKG/usr/share/tessdata/ ; \
 done
 
 find $PKG -print0 | xargs -0 file | grep -e "executable" -e "shared object" | grep ELF \
diff --git a/graphics/tesseract/tesseract.info b/graphics/tesseract/tesseract.info
index 9d0e42ae6966..1a4862f2a0fc 100644
--- a/graphics/tesseract/tesseract.info
+++ b/graphics/tesseract/tesseract.info
@@ -1,10 +1,10 @@
 PRGNAM="tesseract"
-VERSION="3.02.02"
-HOMEPAGE="http://code.google.com/p/tesseract-ocr/"
-DOWNLOAD="https://tesseract-ocr.googlecode.com/files/tesseract-3.02.02.tar.gz \
-          https://tesseract-ocr.googlecode.com/files/tesseract-ocr-3.02.eng.tar.gz"
-MD5SUM="3d57ee5777fa998632ad0693c13a0e9e \
-        3562250fe6f4e76229a329166b8ae853"
+VERSION="3.04.00"
+HOMEPAGE="https://github.com/tesseract-ocr"
+DOWNLOAD="https://github.com/tesseract-ocr/tesseract/archive/3.04.00.tar.gz \
+          https://github.com/tesseract-ocr/tessdata/raw/3.04.00/eng.traineddata"
+MD5SUM="078130b9c7d28c558a0e49d432505864 \
+        59a99c829aa385ae8cde35775e32e57f"
 DOWNLOAD_x86_64=""
 MD5SUM_x86_64=""
 REQUIRES="leptonica"