Front page | perl.perl6.language |
Postings from April 2003
On Thu, Apr 10, 2003 at 02:56:56PM -0700, Austin Hastings wrote: > Sure. But if I write my patterns expecting a series of Colorific > objects, that's MY problem. Just like if you pass \u2105\u2276 to a > pattern like /[0-9]+/ -- it fails. That's YOUR problem -- your data and > your pattern didn't agree: fix your code. > > I think the point is that if I've invested the MIPS in generating a > massive array of Colorific objects, then I should be able to tell the > compiler, "This is an array of Colorific objects. Do pattern matching > using that knowledge" and reap the benefits of my foresight. > > Really, this is just the same argument as signatures: If I adhere to > some basic constraints, I should be able to tell the compiler about the > constraints, and benefit from my own self-discipline. Ah, you're right, I just had blinders on there for a little bit. -Scott -- Jonathan Scott Duff duff@cbi.tamucc.eduThread Previous | Thread Next