Followup-To: comp.lang.scheme "Xah Lee" <[EMAIL PROTECTED]> writes:
> Since this is frequently used, Python provides a somewhat shorter > syntax for it, by specifying the column used as the ordering “key”. [...] > Because Python's implementation is not very refined , this specialized > syntax is actually much speedier than the general form “lambda x, y: > cmp(x[1],y[1])”. It is a burden on the programer to always use the > “key” syntax idiosyncrasy if he is sorting a large matrix. It's not only clearer for a human, but also faster in all good implementations of all languages which support that, except when the ordering function is very simple. It's called Schwartzian transform and I wish more language designers and programmers knew about it. http://en.wikipedia.org/wiki/Schwartzian_transform I urge future SRFI authors to include it. The withdrawn SRFI-32 for sorting didn't do that, and I can't find any other SRFI which deals with sorting. -- __("< Marcin Kowalczyk \__/ [EMAIL PROTECTED] ^^ http://qrnik.knm.org.pl/~qrczak/ -- http://mail.python.org/mailman/listinfo/python-list
