On Sat, Dec 06, 2003 at 09:38:08AM -0800, John Angel wrote:
> It could be some problem with 'locale' settings...

Maybe.  Did you try what I showed in the last message?

> Is it possible to use different function instead of UTF8Toisolat1()? Maybe 
> we can override it with our own function with character conversion table?

Sure, but I'd first check that indeed where the problem is happening.  
There's the iconv library that will do that.  But what would you convert 
to?  I'm not sure but maybe UTF8Toisolat1() already uses iconv -- need 
to look at the libxml2 source.

> How to use old built-in HTML parser? configure --without-libxml?

Not necessary.  Just use the "HTML" parser.

DefaultContents HTML

or 

IndexContents HTML .htm .html

HTML is the old parser
HTML2 is the libxml2 parser
HTML* is the libxml2 parser if built with libxml2, otherwise it's the 
HTML parser.

-- 
Bill Moseley
moseley@hank.org