Re^5: Turning foreach into map?

Replies are listed 'Best First'.
Re^6: Turning foreach into map? by ysth (Canon) on Apr 05, 2005 at 02:27 UTC
If so, I apologize for the fud. I remember it coming up at least twice, with patches to curb the worst case behaviour, but thought that each time there was criticism, delay, and finally silence. But my memory is not what it could be.	[reply]
Re^7: Turning foreach into map? by tilly (Archbishop) on Apr 06, 2005 at 01:52 UTC
I know that I fixed map, and I know that the patch was accepted. But when I tested it again map is slower than foreach for long lists. (Not as badly as it was, however.) Looking at the current code, my patch to mapwhile is still in pp_ctl.c. Therefore the current performance problem in map has to be something else than what I was thinking. Do you know any details of what the current issue is?	[reply]
Re^8: Turning foreach into map? by ysth (Canon) on Apr 06, 2005 at 02:24 UTC
No, I don't. I'd think (without having actually looked at the code in a while) that there's the possibility of optimizing the assignment away in @foo = map ...	[reply]