Re: [Plplot-devel] Proposed Hershey to unicode table changes.

[Andrew R. <aro...@ya...>]

At 09:20 AM 4/01/2005 +0100, you wrote:
>* Andrew Roach <aro...@ya...> [2005-01-04 12:32]:
>
> > I wonder if things like the boldface and italics would not be better 
> handed
> > with their own escape ? For example, #u and #d shift into super-script and
> > sub-script now. Maybe
> >
> > #i for italics
> > #m for medium or #n for normal
> > #b for bold
> >
> > might not be a better way to do it ?
>
>Both schemes are fine with me.  Just a question: which escape sequence would
>we use for upright, since #u is already taken?

Good point... #n for normal ?

> > That way we avoid an incompatibility bomb with the #fi. Presently #fi
> > always maps to a "face" (sans-serif-italic) rather than a styling. If we
> > keep #f for face or family changes, backward compatibility can be
> > maintained. Then #i would become italic, so:
> >
> >     "#fr This is serif, #b now serif-bold, #m and back to serif #i and
> > italicised"
> >
> > Another possibility, which does not have to be mutually exclusive I might
> > point out, could be using a limited set of html-like tags:
> >
> >     "#fr This is serif, <b>now serif-bold</b>, and back to serif <i>and
> > italicised</i>"
>
>I like the XML syntax above because it has explicit closing tags.  In SGML,
>closing tags are not mandatory, what makes the parser quite complicated.
>
>At any rate, your idea is clever.  We have just to have the manpower to
>implement it.

Maybe it's because it's 2AM and I'm not really thinking too clearly, but I 
wonder how difficult it really would be to write a parser ? I am thinking 
that could be easier than other parts of this API expansion. I don't think 
we want a full parser, with 100% strict syntax - it would be too painful to 
loose > and < sigs, and have to replace them with &gt; and &lt;. But a 
simple parser which "tokenised" things we recognise like <b> <i> etc..., by 
my way of thinking would be no harder than the current escape sequences.

-Andrew