Samuele Pedroni wrote: > Nick Coghlan wrote: > >>What would a separate sig expression buy you over defining "->expr" as a >>special form of keyword argument that binds to the keyword name "return" >>in the dictionary for storing extra keyword arguments? > > > seems to me a quirky addition of sugar, also could not be limited; I > prefer going the full length and supporting argument name introduction > with : etc. as shown in the example. >
to be more precise, I find: @sig a: int, b: int -> int more readable and to the point than: @sig(a=int,b=int,->int). First-class sig expressions can have rules to leave out parens as genexp etc. Also it can be extended to support attaching annotations to * and ** args. It would be hard to devise separate sugar for those. > But it seems we agree that interspensing the annotation in the main > head of the function is not such a great idea after all. > _______________________________________________ > Python-3000 mailing list > Python-3000@python.org > http://mail.python.org/mailman/listinfo/python-3000 > Unsubscribe: > http://mail.python.org/mailman/options/python-3000/pedronis%40strakt.com _______________________________________________ Python-3000 mailing list Python-3000@python.org http://mail.python.org/mailman/listinfo/python-3000 Unsubscribe: http://mail.python.org/mailman/options/python-3000/archive%40mail-archive.com
