On Sun, 18 Mar 2007, nicola wrote:
In article
, Aditya Mahajan wrote: On Sun, 18 Mar 2007, nicola wrote:
My goal is to use an arbitrary text font together with the Euler math font. I am able to use a few combinations in ConTeXt, e.g. Palatino with Euler, without problems. But when I switch to XeConTeXt (to use my system's fonts), some Euler's math accents do not show up properly (\hat) or not at all (\acute, \grave, \bar, \breve, \dot, \ddot, \tilde, \check).
AFAIU, Euler does not have these accents. They are pulled from the text font. That is a bit tricky because different text encodings have accents at different locations. Right now, there is only suport for ec and texnansi. You can try to add support for uc encoding by telling ConTeXt about location of accents in uc encoding. Look at the code at the end of math-eul.tex
Ok, I see. Where is \definemathsymbol documented? The third argument is not very clear to me.
\definemathsymbol is defined in math-ini.tex. It has two versions, one with four arguments and one with six arguments. When using four arguments, #1 = name of symbol to be defined #2 = type of math atom (bin, rel, ord, punct, open, close, op, nothing) #3 = font family. Usually, mi = math italic mr = math roman sy = symbol The actual fonts are defined using \definebodyfont (e.g. see type-siz) #4 = hex code of the location of the symbol. In math-eul #3 = tf (text font), so the accent is pulled from the text font. This is usually ugly, but there is no other option since Euler is not a complete math font family. I am not sure when 6 argument version of definemathsymbol is used. Aditya