# We keep track manually of the library version
%global liblongver 0.0.0
%global libshortver 0

Name: cmpfit
Version: 1.2
Release: 1%{?dist}
Summary: A MINPACK-1 Least Squares Fitting Library in C

License: BSD
URL: http://cow.physics.wisc.edu/~craigm/idl/cmpfit.html
Source0: http://cow.physics.wisc.edu/~craigm/idl/down/%{name}-%{version}.zip

%description
CMPFIT uses the Levenberg-Marquardt technique to solve the least-squares 
problem. In its typical use, CMPFIT will be used to fit a user-supplied 
function (the "model") to user-supplied data points (the "data") by adjusting 
a set of parameters. CMPFIT is based upon MINPACK-1 (LMDIF.F) by More' and 
collaborators. 

%package devel
Summary: Headers for developing programs that will use %{name}
Group: Development/Libraries
Requires: %{name} = %{version}-%{release}

%description devel
These are the header files and libraries needed to develop a %{name} 
application.

%prep
%setup -q 

%build
# Manual build, makefile only does static library
gcc %{optflags} -fpic -c -o mpfit.o mpfit.c
gcc -shared -Wl,-soname,libmpfit.so.%{libshortver} -o libmpfit.so.%{liblongver} mpfit.o -lm

%install
mkdir -p  %{buildroot}/%{_includedir}
mkdir -p  %{buildroot}/%{_libdir}

cp mpfit.h %{buildroot}/%{_includedir}
cp libmpfit.so.%{liblongver} %{buildroot}/%{_libdir}
pushd %{buildroot}/%{_libdir}
ln -s libmpfit.so.%{liblongver} libmpfit.so.%{libshortver}
ln -s libmpfit.so.%{liblongver} libmpfit.so
popd

%post -p /sbin/ldconfig

%postun -p /sbin/ldconfig

%files
%doc README
%{_libdir}/libmpfit.so.%{liblongver}
%{_libdir}/libmpfit.so.%{libshortver}

%files devel
%{_includedir}/mpfit.h
%{_libdir}/libmpfit.so

%changelog
* Fri Aug 16 2013 Sergio Pascual <sergiopr at fedoraproject.org> 1.2-1
- Initial spec