Re: BUG REPORT: Website META Language 2.0.1 (23-May-2000)

[Howard Alexander LaHurreau <nospam@thanx>]

On Wed, May 24, 2000 at 12:46:10AM +0200, Denis Barbier wrote:
> On Tue, 23 May 2000, Howard Alexander LaHurreau wrote:
> > PROBLEM DESCRIPTION:
> > I'm having problems with the migration from 2.0.0 to 2.0.1.
> > It appears that when mp4h reaches an unknown tag, it treats
> > it as if it had been defined with whitespace=delete.
> No, it parses attributes and writes them in a space separated list.

Ahh.  That explains why it does that.  :-)
 
> > The problem I'm having with this is that this code:
> > ---------------8< cut here 8<---------------
> >         print <<"END IMGLINK";
> > <a href="$rss->{'image'}->{'link'}"><img src="$rss->{'image'}->{'url'}"
> >  alt="$rss->{'image'}->{'title'}" border="0" align="right"
> > END IMGLINK
> >         print " width=\"$rss->{'image'}->{'width'}\""
> >                 if $rss->{'image'}->{'width'};
> >         print " height=\"$rss->{'image'}->{'height'}\""
> >                 if $rss->{'image'}->{'height'};
> >         print " /></a>\n";
> > ---------------8< cut here 8<---------------
> Let me try to explain what happens:
>   Input     Interpretation
>   print           string
>    <<             these < delimiters can not begin a tag, because they
>                   are followed by invalid characters (resp. < and ")
>   "END IMGLINK";  string
>   <a              beginning of an undefined tag.  We are looking for
>                   attributes and a right angle bracket.
>     href="$rss->{'image'}->{'link'}"  first attribute
>   >               end of unknown tag
>   <img            beginning of an undefined tag.  Below is the list of
>                   attributes:
> 1    src="$rss->{'image'}->{'url'}"
> 2    alt="$rss->{'image'}->{'title'}"
> 3    border="0"
> 4    align="right"
> 5    END IMGLINK
> 6    print
> 7    " width=\"$rss->{'image'}->{'width'}\""
> 8    if
> 9    $rss-
>   >               end of unknown tag (img)
>   
>   {'image'}->{'width'};   string, etc.

The question that I have is should MP4H be playing with the whitespace
inside the unknown tag?  (Or is there any way to prevent it from doing
so?)
 
> > gets turned into this code:
> > ---------------8< cut here 8<---------------
> >         print <<"END IMGLINK";
> > <a href="$rss->{'image'}->{'link'}"><img src="$rss->{'image'}->{'url'}" alt="$rss->{'image'}->{'title'}" border="0" align="right" END IMGLINK print 
" width="$rss->{'image'}->{'width'}"" if $rss->{'image'}->{'width'};
> >         print " height=\"$rss->{'image'}->{'height'}\""
> >                 if $rss->{'image'}->{'height'};
> >         print " /></a>\n";
> > ---------------8< cut here 8<---------------
> Indeed concatenate attributes of the img tag and you obtain the code
> above.  The solution is to prevent parsing <img> tag, e.g. with
>         print <<"END IMGLINK";
> <a href="$rss->{'image'}->{'link'}"><<suck>img src="$rss->{'image'}->{'url'}"
>  alt="$rss->{'image'}->{'title'}" border="0" align="right"
> END IMGLINK
> The <suck> tag prevents scanning by mp4h, and is automatically removed
> in pass 8.
> Another solution is to enclose your ePerl code with
> <protect pass=2>...</protect> whenever possible.

OK, thanks for your help.  :-)

-- 
AlexDW <alexdw@locl.net> -- KB9SYM -- DOCTOR WHO FOREVER
Pecor's Health-Food Principle:
        Never eat rutabaga on any day of the week that has a "y" in it.
Visit The Home of AlexDW at http://www.locl.net/homes/alexdw/
______________________________________________________________________
Website META Language (WML)                www.engelschall.com/sw/wml/
Official Support Mailing List                   sw-wml@engelschall.com
Automated List Manager                       majordomo@engelschall.com