Re: [GiNaC-devel] [SCM] GiNaC -- a C++ library for symbolic computations branch, ginac_1-5, updated. release_1-4-0-198-g8bf0597

On Fri, Jul 31, 2009 at 03:30:45PM +0200, Jens Vollinga wrote:

oops, yes, thanks. What was I thinking?!?

71 GiNaC::function* f; 72 try { 73 f = new GiNaC::function(reinterpret_cast<unsigned>(reader->second), args); 74 } 75 catch ( std::runtime_error ) { I think if (f) delete f; should be inserted here, otherwise we might leak memory if ->second is a pointer to a (C++) function. Best regards, Alexei