Re: [LaTeXML] Fwd: Special TeX/LaTeX fonts in MathML (calligraphy vs. script)

Frédéric WANG <[email protected]> Tue, 22 Jul 2014 07:24:37 +0200
Newsgroups gmane.comp.mozilla.devel.mathml
Message-ID <[email protected]>
Le 22/07/2014 01:39, William F Hammond a =E9crit :
> 1.  Can you explain quickly what 'ss01' means and where the =

> caligraphic set is to be found in the various *.otf ?
See http://www.microsoft.com/typography/otspec/features_pt.htm#ssxx
So font-feature-settings: "ss01"; maps to the second style (which is =

calligraphic for XITS).

I actually realized that Asana Math also provides alternate calligraphic =

via the "salt" feature =

(http://www.microsoft.com/typography/otspec/features_pt.htm#salt), but =

was not able to make it work, even with the "stylistic(...)" syntax below.

Also the "ssty" variants (defined in the OpenType MATH) for Latin Modern =

Math that I mentioned in a previous thread are supposed (and will be in =

Gecko 31) to be used to provide different glyph sizes for a same =

character. This is to fix the well-known "prime in superscript" issue. =

So as I said it has nothing to do with the calligraphic style, sorry.

As I indicated in a previous mail, you need to open fontforge, select a =

character do Element =3D> Glyph Info and check the substitution / variant =

choice menu to see possible remappings. Not very convenient, so that's =

why I said font designers should try to find a consistent way to do =

that. However, there is a CSS font spec =

http://dev.w3.org/csswg/css-fonts/ that I think it is well implemented =

in Gecko, but disabled by default for now =

(https://bugzilla.mozilla.org/show_bug.cgi?id=3D975744). This would =

provide cleaner and more consistent syntax like

- "font-variant-numeric: oldstyle-nums" (onum) for old style numbers
- "font-variant-alternates: stylistic(calligraphic)" (salt) or =

"font-variant-alternates: styleset(calligraphic)" (ssXX) for =

calligraphic. (with "calligraphic" defined via @font-feature-values on a =

per-font basis see =

http://dev.w3.org/csswg/css-fonts/#propdef-font-variant-alternates)

I need to check with someone at Mozilla about the status of this.

> 2.  Would it be possible to make a single .woff containing all of the =

> xits*.otf ?
I'm not sure that would be a good idea, the normal way is to have one =

"math" font and other fonts for the surrounding text with =

regular/italic/bold/bold-italic styles. Also, they have characters in =

the same Unicode blocks so you'll have to move many of them in a =

"private" part to avoid overlap. Why do you want to do that?

> P.S. 3.  It seems that MathJax supplies calligraphic glyphs when =

> \mathcal{} is used with tex input.
Yes, as discussed earlier, MathJax uses its own set of fonts and a =

special interpretation of the class=3D"MJX-.." attribute to provide some =

calligraphic glyphs. You can just use the MathJax calligraphic font (or =

other fonts) if you wish via a font-family (which was Bruce's original =

idea). But the point of this thread was to find a standard way that will =

work in all rendering engines and fonts. So either standardized as a =

MathML mathvariant for the "semantic" option or via CSS for the "style" =

option. BTW, the CSS font method is useless for MathJax, since it almost =

never uses the original math fonts but splits them into smaller files =

with some glyphs remapped into different positions. And MathJax does not =

known about user stylesheets so the style would still have to be =

computed internally by MathJax anyway.