[antlr-interest] BENCHMARK. ANTLR. Bad results.

[Braden McDaniel]

iank at bearcave.com wrote:

>   I wrote a reference counted string container originally for ANTLR.
>   However I never integrated it into ANTLR.  Obviously it is written
>   in C++ and is available under something like the BSD license (e.g.,
>   you can do what ever you want with it, but you can't blame me for
>   any problems you encounter).  See
> 
>      http://www.bearcave.com/software/string/index.html
> 
>   I have not benchmarked this against the GNU std::string, so I don't
>   know if it is faster or not.  It does properly implement reference
>   counted copy-on-write semantics (which is more difficult than I
>   thought in C++).  So copying should not be an issue, since it just
>   copies pointers.

It's worth noting here that COW has fallen out of favor in many 
std::string implementations because it performs poorly when strings are 
shared between threads.

I believe, however, that the GNU libstdc++ std::string still uses COW.

Braden


 
Yahoo! Groups Links

<*> To visit your group on the web, go to:
    http://groups.yahoo.com/group/antlr-interest/

<*> To unsubscribe from this group, send an email to:
    antlr-interest-unsubscribe at yahoogroups.com

<*> Your use of Yahoo! Groups is subject to:
    http://docs.yahoo.com/info/terms/