|
Boost : |
From: Daryle Walker (darylew_at_[hidden])
Date: 2004-01-26 15:12:21
On 1/24/04 10:57 PM, "Eric Niebler" <eric_at_[hidden]> wrote:
> Daryle Walker wrote:
>>
>> Also, I wouldn't consider a list of characters to be equivalent to a pattern
>> matcher, even if the latter is implemented in terms of the former. (Can a
>> regex be defined by other ways than a string? Using "regex atom" objects
>> composed together could cut out the mandatory parsing engine the string
>> method requires.)
>>
>
> Indeed, such a regex engine already exists. It's called xpressive and it
> lives in the boost sandbox. It's under active development.
So Regex always uses string parsing during construction (besides copy
construction), for now? I guess that Regex would still use some sort of
"atom" structure internally, though.
> FWIW, I also agree that the Boost.Regex constructor should remain explicit.
> IMO, a regex is not a string, and a string is not a regex. A string is merely
> one (admittedly handy) way of initializing a regex.
-- Daryle Walker Mac, Internet, and Video Game Junkie darylew AT hotmail DOT com
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk