Boost logo

Boost :

From: Chris Russell (cdr_at_[hidden])
Date: 2003-05-30 17:24:26


For what it's worth, I've had great success using James Clark's expat to
populate directed graphs abstracted by a BGL adjacency list. I'm too busy to
document and submit to the sandbox at the moment but that should give anyone
interested enough information to go do it. Parenthetically, I've found that
BGL's vistor concept can be easily adapted to to provide extremely flexible
schema validation, and some nifty deserialization tricks using function
objects installed in the vertex property map. I've been mulling over the
idea of generalizing this work to make it easier to run the process in
reverse (that is re-serialize the data to XML) in order to create a
generalized validating XML-based persistence mechanism. The code I have
today has not been benchmarked against any other validating parser but I'm
willing to go out on a limb and speculate that the BGL-based approach is at
least as fast or faster than other validating parsers. And besides it
aesthetically pleasing and you effectively get a DOM for free. It would be
fairly straight forward for example to put an XPath interface on an
adjacency list representation of a parsed XML document. Just a few ideas.
Anyone who wants more details, e-mail me off the list.

- Regards
Chris

"Stefan Seefeld" <seefeld_at_[hidden]> wrote in message
news:0441aa86ffefdb165349500c0d94b4503ed638f9_at_Orthosoft.ca...
> hi there,
>
> I'v been wondering for quite a while whether a generic XML C++ library
> would be an interesting project (providing DOM- and SAX- like as well as
> other APIs), in particular because all the standard interfaces that
> are released by the W3C (and others) are tailored towards java, and
> there is often either no C++ mapping or one that I consider rather poor
> as it doesn't make any use of C++ specific idioms.
>
> I wrote a C++ wrapper library around libxml2 (http://xmlsoft.org/),
> and would like to submit it to boost. It currently provides a DOM-like
> and a SAX-like interface, is parametrized for the (possibly unicode
> enabled) string type, and has already proven its value for my own
> projects.
>
> Is there any interest in such a (boost) project ?
>
> Kind regards,
> Stefan
>
> _______________________________________________
> Unsubscribe & other changes:
http://lists.boost.org/mailman/listinfo.cgi/boost
>


Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk