Hi!

I have segmentation fault when i parsing contents of some sites (not all) by boost_html.

Program stoped in file element.hpp in method read on line 273:
ptr_element_t newElement = factory(elementName, this);

My fragment of code:

stringstream ost;
ost << s << endl ;
html.read(ost);

Variable s is string contains content of site downloaded by curl.

Please answer what is the problem.

P.S. Very good code/library (boost_html)

i'm actually in the process of doing some work in the boost.html library.

one of the things that will change will be the namespace (fyi) to boost.dom

Can you email me the links to the sites that are killing the library? i've tested the code on random sites, but am interested in sites that will kill the parser.

Example output from my program:


Reading: google.pl
Reading: images.google.pl/imghp?hl=pl&tab=wi
Reading: video.google.pl/?hl=pl&tab=wv
Reading: groups.google.pl/grphp?hl=pl&tab=wg
Reading: google.pl/dirhp?hl=pl&tab=wd
Reading: mail.google.com/mail/?hl=pl&tab=wm
Segmentation fault


other:


Reading: www.wp.pl
Segmentation fault

This is a good problem!
And, I might have a solution pretty quick.

Ok. I'm waiting for solution

I've been in the process of refactoring alot of the code, primarily the parser that reads the HTML/XML files. This problem is directly related to tag opening and tag closing.

The code chokes on the '<' and '>' in the javascript portions. I think in a few days the new code will be ready to download. If you find more examples of crashes, please post or email me the links, so I have a wider range to test against.