feat: ignore \\text command

[edeustace]

We have LaTex that intersperses text into mathematical expressions. We
want to be able to parse the LaTeX to math and just ignore thes
commands. This commit adds text as a whitespace rule.

I'm new to this lib so this may be a fairly naive approach - happy for feedback.

[dqnykamp]

Do you have any ideas how to modify the regular expression to match all of "\text{multiple {words} here}"? One probably would not encounter that frequently, but I did notice that it would stop at the first "}" in the above example. I'm wondering if one might need to actually parse the \text to match the braces.

@kisonecat, any opinion on dealing with the case of braces within the \text?

[edeustace]

any further thoughts on this @dqnykamp ?

[dqnykamp]

I played around with the idea of having \text{} be parsed as a regular command (with special logic so that its contents aren't parsed as math), but I couldn't come up with a satisfactory solution on how to handle the resulting \text{} object. For example, a common use of \text{} is to make sub/superscripts words. How should "a_\text{long} b" be handled? We couldn't ignore the \text{} in that case; one would need to put something in for the subscript. I would probably want the result to be ['*', ['_', 'a', 'long'], 'b']. Once we go down that road, we start departing from your idea to make \text{} be like a \comment{} command, so I didn't pursue that idea further.