package init

tesseract.spec

@@ -0,0 +1,86 @@
+Name:                tesseract
+Version:             4.0.0
+Release:             1
+Summary:             Raw OCR Engine
+License:             Apache 2.0 and BSD-2-Clause
+URL:                 https://github.com/tesseract-ocr/%{name}
+Source0:             https://github.com/tesseract-ocr/tesseract/archive/%{version}.tar.gz
+Patch0:              tesseract_datadir.patch
+BuildRequires:       automake autoconf-archive gcc-c++ libtool libtiff-devel leptonica-devel
+BuildRequires:       cairo-devel libicu-devel pango-devel
+BuildRequires:       tesseract-langpack-eng
+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}
+%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
+find %{buildroot}%{_libdir} -type f -name '*.la' -delete
+mkdir -p %{buildroot}/%{_datadir}/%{name}/tessdata/
+%ldconfig_scriptlets
+
+%files
+%license LICENSE
+%doc AUTHORS ChangeLog README.md
+%{_bindir}/%{name}
+%dir %{_datadir}/%{name}
+%dir %{_datadir}/%{name}/tessdata
+%{_datadir}/%{name}/tessdata/configs/
+%{_datadir}/%{name}/tessdata/tessconfigs/
+%{_datadir}/%{name}/tessdata/pdf.ttf
+%{_libdir}/lib%{name}*.so.4*
+
+%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}/language-specific.sh
+%{_bindir}/lstmeval
+%{_bindir}/lstmtraining
+%{_bindir}/merge_unicharsets
+%{_bindir}/mftraining
+%{_bindir}/set_unicharset_properties
+%{_bindir}/shapeclustering
+%{_bindir}/tesstrain.sh
+%{_bindir}/tesstrain_utils.sh
+%{_bindir}/text2image
+%{_bindir}/unicharset_extractor
+%{_bindir}/wordlist2dawg
+
+%changelog
+* Fri Jan 8 2021 chengzihan <chengzihan2@huawei.com> - 4.0.0-1
+- Package init

tesseract_datadir.patch

@@ -0,0 +1,49 @@
+diff -rupN tesseract-4.0.0/src/ccutil/Makefile.am tesseract-4.0.0-new/src/ccutil/Makefile.am
+--- tesseract-4.0.0/src/ccutil/Makefile.am	2018-10-29 09:53:12.000000000 +0100
++++ tesseract-4.0.0-new/src/ccutil/Makefile.am	2018-11-13 15:23:08.331318335 +0100
+@@ -3,7 +3,7 @@ AM_CXXFLAGS =
+ AM_CPPFLAGS =
+ 
+ if !NO_TESSDATA_PREFIX
+-AM_CXXFLAGS += -DTESSDATA_PREFIX=@datadir@
++AM_CXXFLAGS += -DTESSDATA_PREFIX=@datadir@/tesseract/
+ endif
+ 
+ if VISIBILITY
+diff -rupN tesseract-4.0.0/src/lstm/Makefile.am tesseract-4.0.0-new/src/lstm/Makefile.am
+--- tesseract-4.0.0/src/lstm/Makefile.am	2018-10-29 09:53:12.000000000 +0100
++++ tesseract-4.0.0-new/src/lstm/Makefile.am	2018-11-13 15:23:08.332317564 +0100
+@@ -12,7 +12,7 @@ SUBDIRS =
+ AM_CXXFLAGS = $(OPENMP_CXXFLAGS)
+ 
+ if !NO_TESSDATA_PREFIX
+-AM_CXXFLAGS += -DTESSDATA_PREFIX=@datadir@
++AM_CXXFLAGS += -DTESSDATA_PREFIX=@datadir@/tesseract/
+ endif
+ 
+ if VISIBILITY
+diff -rupN tesseract-4.0.0/tessdata/configs/Makefile.am tesseract-4.0.0-new/tessdata/configs/Makefile.am
+--- tesseract-4.0.0/tessdata/configs/Makefile.am	2018-10-29 09:53:12.000000000 +0100
++++ tesseract-4.0.0-new/tessdata/configs/Makefile.am	2018-11-13 15:23:08.332317564 +0100
+@@ -1,3 +1,3 @@
+-datadir = @datadir@/tessdata/configs
++datadir = @datadir@/tesseract/tessdata/configs
+ data_DATA = inter makebox box.train unlv ambigs.train lstm.train lstmdebug api_config kannada box.train.stderr quiet logfile digits hocr tsv linebox pdf rebox strokewidth bigram txt
+ EXTRA_DIST = inter makebox box.train unlv ambigs.train lstm.train lstmdebug api_config kannada box.train.stderr quiet logfile digits hocr tsv linebox pdf rebox strokewidth bigram txt
+diff -rupN tesseract-4.0.0/tessdata/Makefile.am tesseract-4.0.0-new/tessdata/Makefile.am
+--- tesseract-4.0.0/tessdata/Makefile.am	2018-10-29 09:53:12.000000000 +0100
++++ tesseract-4.0.0-new/tessdata/Makefile.am	2018-11-13 15:23:08.332317564 +0100
+@@ -1,4 +1,4 @@
+-datadir = @datadir@/tessdata
++datadir = @datadir@/tesseract/tessdata
+ 
+ data_DATA = pdf.ttf
+ EXTRA_DIST = $(data_DATA)
+diff -rupN tesseract-4.0.0/tessdata/tessconfigs/Makefile.am tesseract-4.0.0-new/tessdata/tessconfigs/Makefile.am
+--- tesseract-4.0.0/tessdata/tessconfigs/Makefile.am	2018-10-29 09:53:12.000000000 +0100
++++ tesseract-4.0.0-new/tessdata/tessconfigs/Makefile.am	2018-11-13 15:23:08.332317564 +0100
+@@ -1,3 +1,3 @@
+-datadir = @datadir@/tessdata/tessconfigs
++datadir = @datadir@/tesseract/tessdata/tessconfigs
+ data_DATA = batch batch.nochop nobatch matdemo segdemo msdemo
+ EXTRA_DIST = batch batch.nochop nobatch matdemo segdemo msdemo