On Oct 14, 2008, at 12:12 AM, Dag Sverre Seljebotn wrote: > Robert Bradshaw wrote: >> On Oct 11, 2008, at 7:31 AM, Dag Sverre Seljebotn wrote: >>> This absolutely shouldn't be more than half an hour, it is just >>> about >>> throwing up a compiler error in the right transform. >> >> Now that you put it that way, it makes it sound much easier--I love >> transformations :). But I actually realize how much more restrictive >> it is, as you can't do >> >> cdef foo(int n, int* a): >> ... >> >> foo(5, [1,2,3,4,5]) >> >> which I think could be very useful. So I would maybe say disallow >> array assignment except in declarations, but the above would still >> work. > > As you pointed out in another post this was more hairy than I expected > because of cdef transformation. > > But anyway a patch is up doing this now. It is not tied to > declaration, > but what can Cython prove is the first assignment, but that works just > as well (or better) for this purpose! (And is the same currently).
Thanks. A bit more restrictive, but certainly harder for the user to invoke confusing behavior. - Robert _______________________________________________ Cython-dev mailing list [email protected] http://codespeak.net/mailman/listinfo/cython-dev
