[NTG-pdftex] Re: backslashes in \pdfoutline

Hans Hagen pragma at wxs.nl
Thu Jul 15 09:08:48 CEST 2004


Heiko Oberdiek wrote:

>I think a clear no:
>
>  
>
it can be done by 'parsing strings' which is painful

>If my understanding of Hans's idea is correct then
>\escapedpdf can be used the following way, eg.
>
>  \def\foo{bar}
>  \pdfinfo{/Title (\escapedpdf{\foo})}
>  
>
indeed

>a) \escapedpdf has to be expandable.
>b) The argument is full expanded.
>The return value can be a string in the manner of \meaning
>(characters with catcode 12 and spaces with catcode 10).
>  
>
indeed

>b) can be achieved by \edef, but \edef cannot be used
>because of a).
>
>Even if we have a full expanded string, how to parse
>it in a safe way? Catcodes? Espacially spaces are
>problematic:
>  
>
indeed

>* \scantokens from e-TeX is AFAIK expandable, but
>  the catcode changes before are not.
>  
>
right, it's also slow to prepare settings (definitions, catcodes etc)

actually, it would be handy to have catcode vectors so that we could 
quickly swap 'm

>* For the \@sanitize at onelevel or \meaning method a
>  macro definition is necessary, violation of a).
>* \futurelet is not expandable.
>* Scan methods such as \scan#1(... are problematic and fail,
>  if there are spaces in the string.
>  
>
indeed, and 

  * why should we complicate live when it can be programmed (i already wasted too many days on such painful things) 

Hans 

-----------------------------------------------------------------
                                          Hans Hagen | PRAGMA ADE
              Ridderstraat 27 | 8061 GH Hasselt | The Netherlands
     tel: 038 477 53 69 | fax: 038 477 53 74 | www.pragma-ade.com
                                             | www.pragma-pod.nl
-----------------------------------------------------------------



More information about the ntg-pdftex mailing list