Regex for lex

Question

I am developing a simple translator from MathML to Latex, using Lex and Yacc. In my lex file containing the regex rules I have one defined for arithmetic operators [-+*=/]. I want to extended so that it would recognize plus-minus (+-) and invisible times ('&InvisibleTimes'), but I'm unfamiliar with regex and I need some help.

Answer 1

Try this:

([-+*=/]|\+-|&InvisibleTimes)

Note that you need to escape the + in +- because it's an operator outside of character classes. You can do this with backslash (as I've done here) or with double quotes. (The double-quote syntax is pretty unusual -- most other regex implementations only use backslash for escaping, so I'd be inclined to use backslashes as it makes the regex more "conventional".)

Answer 2

Would something like this work?

(?:[-+*=/]|\+-|&InvisibleTimes)

Answer 3

I'm not very familiar with MathML, so I have the opposite problem of you. As others have said, you can do this all in one regex, like this:

[-+*=/]|\+-|&InvisibleTimes

However, if you want to have different actions associated with each of these, you need to do it like this:

[-+*=/]            {/* action 1 here */}
\+-                {/* action 2 here */}
&InvisibleTimes    {/* action 3 here */}