On 20/01/2012, at 1:08 PM, Motohiro KOSAKI wrote:
> Bug #5877: Poor performance of initial and final UTF-8 substrings
> https://bugs.ruby-lang.org/issues/5877
>=20
> Author: Nathan Weizenbaum
> Status: Feedback
> Priority: Normal
> Assignee:=20
> Category:=20
> Target version:=20
> ruby -v: ruby 1.9.3p0 (2011-10-30 revision 33570) [x86_64-linux]=20
>=20
>=20
> Currently, taking the substring of a UTF-8 string always takes time =
proportional to the length of the string. While this is necessary in the =
general case, there are many common cases where it could be much faster, =
in particular when the substring is near the beginning of the end of the =
superstring. In those cases, the operation should only be proportional =
to the length of the substring.

This discussion from ten years ago is relevant to performance =
workarounds for UTF-8:
=
<http://blade.nagaokaut.ac.jp/cgi-bin/vframe.rb/ruby/ruby-talk/45935?45702=
-47288>

Clifford Heath.=