File tesseract.spec of Package tesseract-ocr
# Spec file for package tesseract
#
#
# norootforbuild
Name: tesseract
Version: 2.04
%define langv1 2.00
Release: 1
Summary: Open source OCR Engine developed by HP Labs - now improved by Google
Group: Applications/Multimedia
License: Apache License
URL: http://code.google.com/p/tesseract-ocr/
Source0: http://tesseract-ocr.googlecode.com/files/tesseract-%{version}.tar.bz2
Source1: http://tesseract-ocr.googlecode.com/files/tesseract-%{langv1}.eng.tar.bz2
Source2: http://tesseract-ocr.googlecode.com/files/tesseract-%{langv1}.fra.tar.bz2
Source3: http://tesseract-ocr.googlecode.com/files/tesseract-%{langv1}.ita.tar.bz2
Source4: http://tesseract-ocr.googlecode.com/files/tesseract-%{langv1}.deu.tar.bz2
Source5: http://tesseract-ocr.googlecode.com/files/tesseract-%{langv1}.spa.tar.bz2
Source6: http://tesseract-ocr.googlecode.com/files/tesseract-%{langv1}.nld.tar.bz2
Patch0: tesseract-%{version}.patch
BuildRoot: %{_tmppath}/build-%{name}-%{version}
BuildRequires: gcc-c++ libtiff-devel
%description
The Tesseract OCR engine was one of the top 3 engines in the 1995 UNLV
Accuracy test. Since then it has had little work done on it, but it is
probably one of the most accurate open source OCR engines available. The
source code will read a binary, grey or color image and output text. A tiff
reader is built in that will read uncompressed TIFF images, or libtiff can
be added to read compressed images.
%package devel
Summary: Development files for tesseract
Group: Development/Libraries
Requires: %name = %{version}
%description devel
The devel files for tesseract packages
%package eng
Group: Applications/Multimedia
Summary: English language pack for tesseract
%description eng
The %{name}-%{version}.eng package contains the data files required to recognize English
%package fra
Group: Applications/Multimedia
Summary: French language pack for tesseract
%description fra
The %{name}-%{version}.fra package contains the data files required to recognize French
%package ita
Group: Applications/Multimedia
Summary: Italian language pack for tesseract
%description ita
The %{name}-%{version}.ita package contains the data files required to recognize Italian
%package deu
Group: Applications/Multimedia
Summary: German language pack for tesseract
%description deu
The %{name}-%{version}.deu package contains the data files required to recognize German
%package spa
Group: Applications/Multimedia
Summary: Spanish language pack for tesseract
%description spa
The %{name}-%{version}.spa package contains the data files required to recognize Spanish
%package nld
Group: Applications/Multimedia
Summary: Dutch language pack for tesseract
%description nld
The %{name}-%{version}.nld package contains the data files required to recognize Dutch
%prep
%setup -q
%patch0 -p0
tar xjvf %{_sourcedir}/tesseract-%{langv1}.eng.tar.bz2
tar xjvf %{_sourcedir}/tesseract-%{langv1}.fra.tar.bz2
tar xjvf %{_sourcedir}/tesseract-%{langv1}.ita.tar.bz2
tar xjvf %{_sourcedir}/tesseract-%{langv1}.deu.tar.bz2
tar xjvf %{_sourcedir}/tesseract-%{langv1}.spa.tar.bz2
tar xjvf %{_sourcedir}/tesseract-%{langv1}.nld.tar.bz2
chmod 644 tessdata/eng*
chmod 644 tessdata/fra*
chmod 644 tessdata/ita*
chmod 644 tessdata/deu*
chmod 644 tessdata/spa*
chmod 644 tessdata/nld*
%build
CFLAGS="$RPM_OPT_FLAGS" ./configure --bindir=%{_bindir} --datadir=%{_datadir} --libdir=%{_libdir} --includedir=%{_includedir}
make
%install
rm -rf $RPM_BUILD_ROOT
make install DESTDIR=$RPM_BUILD_ROOT
%clean
rm -rf $RPM_BUILD_ROOT
%files
%defattr(-,root,root,-)
%doc AUTHORS COPYING ChangeLog README
%{_bindir}/cntraining
%{_bindir}/mftraining
%{_bindir}/tesseract
%{_bindir}/unicharset_extractor
%{_bindir}/wordlist2dawg
%dir %{_datadir}/tessdata
%{_datadir}/tessdata/configs
%{_datadir}/tessdata/confsets
%{_datadir}/tessdata/tessconfigs
%files devel
%defattr(-,root,root)
%{_includedir}/tesseract/
%{_libdir}/libtesseract*
%files eng
%defattr(-,root,root)
%{_datadir}/tessdata/eng.DangAmbigs
%{_datadir}/tessdata/eng.freq-dawg
%{_datadir}/tessdata/eng.inttemp
%{_datadir}/tessdata/eng.normproto
%{_datadir}/tessdata/eng.pffmtable
%{_datadir}/tessdata/eng.unicharset
%{_datadir}/tessdata/eng.user-words
%{_datadir}/tessdata/eng.word-dawg
%files fra
%defattr(-,root,root)
%{_datadir}/tessdata/fra.DangAmbigs
%{_datadir}/tessdata/fra.freq-dawg
%{_datadir}/tessdata/fra.inttemp
%{_datadir}/tessdata/fra.normproto
%{_datadir}/tessdata/fra.pffmtable
%{_datadir}/tessdata/fra.unicharset
%{_datadir}/tessdata/fra.user-words
%{_datadir}/tessdata/fra.word-dawg
%files ita
%defattr(-,root,root)
%{_datadir}/tessdata/ita.DangAmbigs
%{_datadir}/tessdata/ita.freq-dawg
%{_datadir}/tessdata/ita.inttemp
%{_datadir}/tessdata/ita.normproto
%{_datadir}/tessdata/ita.pffmtable
%{_datadir}/tessdata/ita.unicharset
%{_datadir}/tessdata/ita.user-words
%{_datadir}/tessdata/ita.word-dawg
%files deu
%defattr(-,root,root)
%{_datadir}/tessdata/deu.DangAmbigs
%{_datadir}/tessdata/deu.freq-dawg
%{_datadir}/tessdata/deu.inttemp
%{_datadir}/tessdata/deu.normproto
%{_datadir}/tessdata/deu.pffmtable
%{_datadir}/tessdata/deu.unicharset
%{_datadir}/tessdata/deu.user-words
%{_datadir}/tessdata/deu.word-dawg
%files spa
%defattr(-,root,root)
%{_datadir}/tessdata/spa.DangAmbigs
%{_datadir}/tessdata/spa.freq-dawg
%{_datadir}/tessdata/spa.inttemp
%{_datadir}/tessdata/spa.normproto
%{_datadir}/tessdata/spa.pffmtable
%{_datadir}/tessdata/spa.unicharset
%{_datadir}/tessdata/spa.user-words
%{_datadir}/tessdata/spa.word-dawg
%files nld
%defattr(-,root,root)
%{_datadir}/tessdata/nld.DangAmbigs
%{_datadir}/tessdata/nld.freq-dawg
%{_datadir}/tessdata/nld.inttemp
%{_datadir}/tessdata/nld.normproto
%{_datadir}/tessdata/nld.pffmtable
%{_datadir}/tessdata/nld.unicharset
%{_datadir}/tessdata/nld.user-words
%{_datadir}/tessdata/nld.word-dawg
%changelog
* Tue Nov 24 2009 - gbv@oxixares.com
- First version from Andrew Ziem and Ray Smith
