[boost] Re: Formal Review: The Spirit Parser Framework.

From: David B. Held (dheld_at_[hidden])
Date: 2002-10-14 16:03:51

"John Maddock" <john_maddock_at_[hidden]> wrote in message
news:017001c27112$d1ce3f60$b93d87d9_at_1016031671...
> The formal review for the Spirit Parser Framework begins today, and runs
> until the end of Sunday 20th October.
> [...]

I regret that I don't have the time or the experience to write a lengthy
and in-depth review.  However, I feel it is important to offer the
perspective
of a casual user with no particularly deep parsing experience.

I vote to accept the library.

This vote is based on a past evaluation of the library (around version 1.2,
I believe).  The main reasons for inclusion are: simplicity, ease-of-use,
and documentation for small-scale projects.  While I did not use the micro-
parsers extensively, I did use the library in a notably simple context: to
parse an XML-like language.  While there is example code to more
elegantly parse even full XML, it is a testament to the library that I was
able to hack together an adequate parser in just a few days, from scratch.

The fact that the documentation was more than adequate for this task, and
that the library itself is easy enough to use for such small projects makes
it
immediately useful, in my opinion.  The resulting source code was easy to
read, understand, and maintain.  Yet the library also made the parser
extensible in many ways, and I found it was possible to design it
incrementally, although a larger project would be well-advised to design it
properly ahead of time.

So, at the very least, I can testify that Spirit is more than capable of
providing an elegant, easy-to-use parser for simple tasks without having
to learn a complicated parser generator and perform multiple phases to
build an actual parser.  The fact that it can parse arbitrarily complex
grammars gives it a distinct advantage over tokenizers or even regex
parsers (though those also have their place).  And yet, it is almost as
simple to use.

Dave