On Saturday 26 June 2004 00:27, Chris Karakas wrote:
>
> Hello Jose',
>
> >   regarding the CDATA or not CDATA inside equations if I try to insert
> > an inequality inside the equation the docbook output is:
> > ...
> >   < alt >\[
> > E< mc^{2} \]
> >
> >   </ alt >
> > ...
> >   without the CDATA jade complains:
> >
> > jade:f1.sgml:33:4:E: character "^" not allowed in attribute
> > specification list
> > jade:f1.sgml:33:4:E: element "MC" undefined
> > jade:f1.sgml:36:11:E: end tag for "EQUATION" which is not finished
> >
> >   don't you get this?
>
> Indeed, I get this. I had never tried an inequality, so I was caught by
> surprise!
>
> However, it turned out that the solution is simple: you should write the
> SGML entities of < and > instead of the brackets themselves! I adapted my
> scripts to do this:

  I will do that. The question is to show that something needs to be done. 
The introduction of the CDATA section was not arbitrary. :-)

  Also notice that since we have problems with some browser not 
understanding most entities I escape now only < > &

  This should simplify your scripts. ;-)

> First, the sedscr script
>
> http://www.karakas-online.de/mySGML/sedscr
>
> inserts a newline after the $ that follows an < alt > tag. Then, the
> awkscr_math script
>
> http://www.karakas-online.de/mySGML/awkscr_math
>
> transforms < and > inside alt tags (but excluding the lines where the
> tags themslves happen to be - the line of the starting alt tag containing
> only the alt tag itself due sedscr's inserted newline) to their SGML
> entities.
>
> See
>
> http://www.karakas-online.de/mySGML/explain-sgml-math-code-correction.htm
>l

  I think that you know this but to be more systematic I have to say that:
1) lyx no longer outputs MathML with the sgml version.
2) Now we no longer output equation, if it is a simple equation the lyx 
writes an inlineequation else it writes an informalequation. Is that what 
you want?

> Thus, the only thing you have to do is replace < and > with their
> entities instead of including  the alt part in CDATA!

  I will do that, and try to look to your other complains. ;-)

> Chris

-- 
José Abílio


Reply via email to