Grant Edwards wrote: > On 2005-04-21, Sergei Organov <[EMAIL PROTECTED]> wrote: > >>>>>> Well, I'm writing for embedded realtime systems in C/C++ and >>>>>> have never encountered a single need to use goto. >>>>> >>>>> I have encountered situations in C programs where the best >>>>> thing to use was a goto. Those situations have always been >>>>> handled beutifully by a "raise" in Python. >>>> >>>> setjmp/longjump? >>> >>> I've always found setjmp/longjmp much more confusing and hard >>> to maintain than a simple goto. It also requires library >>> support that goto doesn't. >> >> Agreed. The 'goto error' idiom is in fact the only goto usage >> I do agree with provided there is no support for exceptions, >> but that's not applicable to Python anyway. > > Exactly. I've been writing C code for 20+ years, and the only > problems where I found goto to be a good solution are the ones > where exceptions are even better solutions in Python. I've > never found myself wishing for a goto when writing Python code.
The Python C source code is full of such gotos, it must be good practise... Reinhold -- http://mail.python.org/mailman/listinfo/python-list