Re: [boost] [yap] Re-announcing Yap

Subject: Re: [boost] [yap] Re-announcing Yap
From: Zach Laine (whatwasthataddress_at_[hidden])
Date: 2018-02-23 18:31:05

On Fri, Feb 23, 2018 at 11:14 AM, Jeffrey Graham via Boost <
boost_at_[hidden]> wrote:

>    Zack,
>
>    I have been lurking on this topic, and have interested because I use
>    muparserx (and before that exprtk) in a commercial application.
>
>    I am curious about 2 things:
>
>    1) Is it possible to evaluate a large number of expressions at
>    run-time, where each expression is provided dynamically as a
>    std::string (assuming all functions used are already available)?
>

No.  Yap is an expression *template* library, so the expressions you write
and evaluate with Yap are always in your source code, not runtime-parsable.

>    2) What is the performance compared to other expression evaluation
>    libraries, as compared here:
>

As I said before, it's a totally different problem domain.  But, if you're
still curious about Yap's performance, transformed and evaluated
expressions end up being identical to the equivalent hand-written code for
"reasonable" sizes of expressions.  In my experiments on Clang, expressions
of 35 terminals or less were "reasonable".

Zach