I'm considering moving to Xe(La)TeX for the main reason that it allows me to use unicode in my LaTeX code, making said code easier to read, especially the math.
But I'm confused on how XeTeX typesets symbols based on my input. I see three possibilities:
- Unicode chars are made active and XeTeX outputs the TeX symbols we know and trust.
- Unicode chars are piped (directly) through to the final (pdf, ps, dvi) document.
- A combination of the two.
With one of the advertised points of XeTeX being direct access to system UTF-8 fonts, I'm guessing (2) has a lot to do with it.
But TeX has always been about beautiful typesetting, and a lot of effort has gone into making symbols and their spacing look good. Do we still get the same benefits with unicode output? (Do the symbols look the same? Nicer? Worse?) I believe some symbols do not come directly from a font, but have been painstakingly crafted in TeX itself.
There are specific math constructs that get special treatment in TeX but also have unicode symbols. For example, for
\bigcap there are respectively ∩ and ⋂. Do they both behave accordingly? What about √? Or are there packages that implement this sort of thing?
Are most unicode math symbols interpreted correctly with regard to math spacing? (
Do math delimiters derived from unicode ⦃⦄ ⦅⦆ scale vertically as they should?
Are Combining Diacritical Marks handled appropriately?
Will the output be different when the code is compiled on different systems? Will my generated pdf/ps/dvi look different when viewed on different systems? Or are all relevant fonts automatically included?
Finally, what role does
unicode-math play in this story?