Re: Inconsistent libwww (bug?)

"Mihai T. LAZARESCU" writes:

> I tried to add the html-to-plain conversion facility to w3c (ComLine)
> but the linker complains:
> ld: Undefined symbol
>    _HText_beginAnchor
>    _styleSheet
>    _HText_new2
>    _HText_appendText
>    _HText_appendImage
>    _HText_endAppend
>    _HText_appendParagraph
>    _HText_appendCharacter
>    _HText_beginAppend
>    _HText_endAnchor
>    _HText_setStyle
> after declaring:
>     HTConversion_add(conv,"*/*","www/plain", HTMLToPlain, 1.0, 0.0, 0.0);
> The functions listed above as undefined are called mainly in file HTML.c,
> but also in other source files of the WWW library, but never defined.
> Am I missing something?  Are there other modules of the library that I
> should load and compile more than libwww_src.tar.gz?

The reason why the command line tool doesn't use any of the HTML parsing 
streams is that then the application is required to supply the functions 
above. The streams for parsing HTML looks like this:


The HText interface is defined in the HTText.h module but as you correctly 
mention - the functions are not defined. The reason for this is more historic 
than logical as the idea of the HText interface is to provide a platform 
independent interface. Using callback functions would be a better way of doing 
it. We are working on a new GUI browser called Amaya which handles HTML much 
better than the current Library version.

You can see how to use the HText interface in the Line Mode browser where all 
the functions are defined in the GridText module.


Henrik Frystyk Nielsen, <frystyk@w3.org>
World-Wide Web Consortium, MIT/LCS NE43-356
545 Technology Square, Cambridge MA 02139, USA