summaryrefslogtreecommitdiff
path: root/pod/perltodo.pod
diff options
context:
space:
mode:
authorNicholas Clark <nick@ccl4.org>2004-07-13 21:10:21 +0000
committerNicholas Clark <nick@ccl4.org>2004-07-13 21:10:21 +0000
commita5c75c1d181494e06a58988f267a8b3d081df483 (patch)
tree341b41398681a3b3681ff22304da7bd08fd5a7ff /pod/perltodo.pod
parent60cb11a8de37fb2d6b164364e3bb7cdc6724a2c4 (diff)
downloadperl-a5c75c1d181494e06a58988f267a8b3d081df483.tar.gz
Some thoughts on foreach reverse
p4raw-id: //depot/perl@23104
Diffstat (limited to 'pod/perltodo.pod')
-rw-r--r--pod/perltodo.pod4
1 files changed, 4 insertions, 0 deletions
diff --git a/pod/perltodo.pod b/pod/perltodo.pod
index db7274c23f..156095635c 100644
--- a/pod/perltodo.pod
+++ b/pod/perltodo.pod
@@ -183,6 +183,10 @@ code review. Also fchdir is available in some platforms."
The old perltodo notes that we could optimise foreach to iterate in reverse.
(instead of making a reversed copy on the stack)
+It seems that cx->blk_loop.itermax could be hijacked to act as a go up/go
+down flag, with cx->blk_loop.iterix primed in pp_enteriter for the go down
+case (ie reverse). However, it looks slightly tricky identifying the shape of
+foreach reverse in the optree in Perl_peep.
=head1 Tests