Using rindex with a regexp would appear to undermine greediness.
irb(main):012:0> "a--b--c".rindex(/[-]+/)
=> 5
irb(main):013:0> p $~.begin(0)
5
=> nil
irb(main):014:0> p $~.end(0)
6
=> nil
I would expect the 5 to be a 4.
T.
Using rindex with a regexp would appear to undermine greediness.
irb(main):012:0> "a--b--c".rindex(/[-]+/)
=> 5
irb(main):013:0> p $~.begin(0)
5
=> nil
irb(main):014:0> p $~.end(0)
6
=> nil
I would expect the 5 to be a 4.
T.
Hi,
In message "Re: rindex with regexp" on Tue, 28 Dec 2004 13:09:58 +0900, "trans. (T. Onoma)" <transami@runbox.com> writes:
Using rindex with a regexp would appear to undermine greediness.
irb(main):012:0> "a--b--c".rindex(/[-]+/)
=> 5
irb(main):013:0> p $~.begin(0)
5
I would expect the 5 to be a 4.
rindex is as lazy as me not to seek longest match.
matz.
lol ![]()
T.
P.S. BTW, I recently saw a picture of you. A Most Excellent Beard!
On Monday 27 December 2004 11:22 pm, Yukihiro Matsumoto wrote:
rindex is as lazy as me not to seek longest match.
matz.