On Fri, Mar 22, 2002 at 10:37:09AM +0000, Angus Leeming wrote:
> is boldsymbol a font declaration as far as mathed is concerned? Ie, if I added
> {"boldsymbol", LM_TK_FONT, LM_TC_BS},
> to wordlist_array in math_hash.C,
>
> and
> LM_TC_BS
> to MathTextCodes in math_defs.h
>
> and
> boldsymbol
> to theFontNames in math_font_name in math_support.C
>
> would I be doing the "right" thing? Then I'd just have to investigate the
> code using math_font_name to see whaat to do next?
>
> Am I on the right track?
I don't know.
Actually I was thinking about it yesterday and came to the conclusion that
_I_ want "real font insets". I.e. rather than making the font some kind of
"property" of the individual character (or symbol for that matter), the
font change should be a separate wrapper containing 'plain chars'.
This would have several advantages:
- it is more 'mark-up' oriented than the current 'type setting' oriented
method
- it is closer to what LaTeX does
- it does integrate better with the rest of mathed (where essentially
everything else _is_ an inset)
- it would solve the 'font change in macro argument' problem cleanly
- it would give us the possibility to implement
- it would save up to 60% of the memory used by MathCharInset (ok, this
could be obtained otherwise too, but this way no hack is needed)
The only disadvantages I currently see is that it affects navigation since
we need an extra <CursorRight> (or similar) to enter a font inset, and that
we might want some magic to 'glue' adjacent font insets of the same type
together. But that looks do-able.
However, I won't start fiddling around with it before 1.2 (at least not
within the next few days ;-))
So I would discourage you from doing any work there.
Andre'
--
André Pönitz .............................................. [EMAIL PROTECTED]
