Name:                tesseract
Version:             5.5.3
Release:             1
Summary:             Raw OCR Engine
License:             Apache-2.0 and BSD-2-Clause
URL:                 https://tesseract-ocr.github.io/
Source0:             https://github.com/tesseract-ocr/tesseract/archive/%{version}/%{name}-%{version}.tar.gz
BuildRequires:       gcc make
BuildRequires:       autoconf automake libtool
BuildRequires:       pkgconfig(cairo)
BuildRequires:       pkgconfig(icu-i18n) >= 52.1
BuildRequires:       pkgconfig(icu-uc) >= 52.1
BuildRequires:       pkgconfig(lept) >= 1.74
BuildRequires:       pkgconfig(libarchive)
BuildRequires:       pkgconfig(libcurl)
BuildRequires:       pkgconfig(libtiff-4)
BuildRequires:       pkgconfig(pango) >= 1.38.0
BuildRequires:       pkgconfig(pangocairo)
BuildRequires:       pkgconfig(pangoft2)
Requires:            tesseract-langpack-eng

%description
A commercial quality OCR engine originally developed at HP between 1985 and
1995. In 1995, this engine was among the top 3 evaluated by UNLV. It was
open-sourced by HP and UNLV in 2005.

%package devel
Summary:             Development files for %{name}
Requires:            %{name}%{?_isa} = %{version}-%{release}
Requires: pkgconfig(libarchive)
Requires: pkgconfig(libcurl)
%description devel
The %{name}-devel package contains header file for
developing applications that use %{name}.

%package tools
Summary:             Training tools for %{name}
Requires:            %{name}%{?_isa} = %{version}-%{release}
%description tools
The %{name}-tools package contains tools for training %{name}.

%prep
%autosetup -p1 -n %{name}-%{version}

%build
autoreconf -ifv
%configure --disable-static
%make_build
%make_build training

%install
%make_install
%make_install training-install
%{?delete_la}

mkdir -p %{buildroot}/%{_datadir}/%{name}/tessdata/

%files
%license LICENSE
%doc AUTHORS ChangeLog README.md
%{_bindir}/%{name}
%{_datadir}/tessdata/
%{_datadir}/%{name}/
%{_libdir}/lib%{name}.so.*

%files devel
%{_includedir}/%{name}
%{_libdir}/lib%{name}*.so
%{_libdir}/pkgconfig/%{name}.pc

%files tools
%{_bindir}/ambiguous_words
%{_bindir}/classifier_tester
%{_bindir}/cntraining
%{_bindir}/combine_lang_model
%{_bindir}/combine_tessdata
%{_bindir}/dawg2wordlist
%{_bindir}/lstmeval
%{_bindir}/lstmtraining
%{_bindir}/merge_unicharsets
%{_bindir}/mftraining
%{_bindir}/set_unicharset_properties
%{_bindir}/shapeclustering
%{_bindir}/text2image
%{_bindir}/unicharset_extractor
%{_bindir}/wordlist2dawg

%changelog
* Sat Jul 25 2026 Funda Wang <fundawang@yeah.net> - 5.5.3-1
- update to 5.5.3

* Mon Mar 09 2026 zhang_jian <zhang_jian7@hoperun.com> - 5.5.2-1
- update to 5.5.2
  * Fix and improve configuration for cmake builds
  * Move version info. to appropriate ALTO element
  * Bump mikepenz/action-junit-report from 4 to 6
 
* Tue Sep 09 2025 yaoxin <1024769339@qq.com> - 5.5.1-1
- Update to 5.5.1:
  * Fix linear congruential random number generator
  * Extend elf_aux_info() support for RISC-V on FreeBSD and OpenBSD
  * Fix building elf_aux_info() support on OpenBSD/arm
  * Fix invalid empty interval in punct_stripped() for all-punctuation words
  * Fix function addAvailableLanguages
  * Avoid error in pixSauvolaBinarizeTiled
  * Fix duplicated IDs in ALTO XML when multiple pages are present
  * Remove unused include statements for tprintf.h

* Fri Dec 20 2024 Funda Wang <fundawang@yeah.net> - 5.5.0-2
- add libarchive and curl into devel requires, as they are referenced by pkgconfig file

* Mon Nov 11 2024 Funda Wang <fundawang@yeah.net> - 5.5.0-1
- update to 5.5.0

* Wed Jul 03 2024 xu_ping <707078654@qq.com> - 5.4.1-1
- Update to 5.4.1

* Tue Aug 22 2023 liyanan <thistleslyn@163.com> - 5.3.2-2
- Remove unneeded compilation dependencies tesseract-langpack-eng

* Tue Jul 18 2023 yaoxin <yao_xin001@hoperun.com> - 5.3.2-1
- Update to 5.3.2

* Fri Jan 8 2021 chengzihan <chengzihan2@huawei.com> - 4.0.0-1
- Package init