On Thu, Jul 22, 2010 at 2:14 AM, Stefan Behnel <[email protected]> wrote: > Stefan Behnel, 22.07.2010 11:08: >> Kay Hayen, 22.07.2010 11:02: >>> Well yes, I am adequately proud that my compiler does not only do >>> closures, but generator expressions and generator functions as closures, >>> classes with closures, and meta classes. >> >> Yep, class closures are another long-standing open issue in Cython, and >> meta-classes aren't there, either. > > BTW, I think the main reason why these are missing is that Cython > development is largely driven by personal needs, and these features are > rarely used in real-world Python code. Also, as a descendent of Pyrex, > Cython was originally not intended to compile regular Python code, and most > code that it compiles even today was written either for Pyrex or Cython > specifically. That reduces the urge to get unimplemented language features > done. I think generators are the most frequently requested one, though.
This is surely the reason. I would go so far as to say the majority of the development (and complexity) is due to the fact that we want to operate will with C types and libraries (and cdef classes), which is where most of our users see the most value. Of course we are not as tied to this as we used to be, and look forward to the day when most of this happens transparently under the hood. Generators are hot on our list, and that's the last big item I can think of (though there's countless little ones.) - Robert _______________________________________________ Cython-dev mailing list [email protected] http://codespeak.net/mailman/listinfo/cython-dev
