Only pure unicode is supported. It is not about typesetting formulae, but about quickly accessing the several mathematical unicode symbols. Therefore, there are several limitations.
• This site can also be found under math.xn--xck.cc
• Unicode is a big 💩, but it is the one thing that is supported on most environments. If you want something more sophisticated but also less supported, better look at MathML or MathJax.
• Sometimes characters that should look the same look differently, because they are from different unicode blocks, and your renderer will use different fonts with different metrics for them.
• Both superscript and subscript are limited. For example, the superscript small letters do not contain q. Because Unicode does (to my knowledge) not have a superscript q. Therefore, there is no general support for curly braces _{…} at the moment. You have to write \sum_(_k_=_1_ _t_o_ _n_)\frac1_k for ∑₍ₖ₌₁ ₜₒ ₙ₎⅟ₖ. Notice how _ and ^ followed by a space becomes a small space.
• Also for italics, fraktur, etc., there is no support for more than one letter in the curly brace: {\it A} works, but instead of {\it AB}, you will have to write {\it A}{\it B}. Maybe I will change this in a future version, if there is great demand.
• The parser parses greedy. For example, \leq could be interpreted as ≤q and just ≤, and will thus be interpreted as ≤. If you want ≤q, you need to write \le{}q.
• Only few \frac directives are supported. There is the special \frac1 symbol ⅟, and Unicode has a special sign ⁄ which some renderers recognize as fractional slash. It can be set with \fraction, for example, d^n\fraction_x=^x^d^n^-^n^d^x\fraction_x_x becomes dⁿ⁄ₓ=ˣᵈⁿ⁻ⁿᵈˣ⁄ₓₓ
• List of all currently supported sequences

Back.