Re: [Htmlparser-user] pre tag

[Derrick O. <Der...@Ro...>]

Dave,

PRE has not been added as a tag because it very often is not closed by 
the /PRE. You can create your own "PRE" tag class derived from 
CompositeTag, and register it with a PrototypicalNodeFactory you give to 
the parser.

To answer your previous question about filters for:

<div>Good Morning</div>
<h3>Description</h3>
<pre>
*Text to extract Line1*
*Text to extract Line2*
</pre>
<div>Good Morning</div>

... find the H3 node (with Description as the contents),
... get it's parent
... and extract all text from the parent's children (after the Heading)

so it would be something like
  ExtractTextFromChildrenOf  (HasSibling (And (TagName(H3), HasChild 
(String(Description)))))
This is a lot easier to construct with the FilterBuilder application.

... or alternatively I had thought of making a 'TriggerFilter' that 
would set a member flag when it's subordinate filter went true, and 
after that would always return true because the flag was set... but then 
this member would need to be reset or you would need to build the filter 
fresh for each parse.
 
Derrick

Dave wrote:

> <pre>
> text1
> </pre>
>  
> <table>
> <tr><td>text2</td><tr>
> </table>
>  
> >parse http://web-site table
>   show the whole table structure
> >parse http://web-site <http://web-site/>  pre
>   show the tag "pre" only, no text inside the pre tag.
>  
> It seems that pre is not treated as the parent node of "text1".
>  
> Is this a bug?
>  
> Thanks!
>  
>  
>
>