[Bug] Strange behaviour with text above/below arrows

[gvdeynde]

Describe the bug
ASCII text above/below arrows changes formatting, no greek letters allowed.

To Reproduce

I'm using the bracket style #ce[blah] because I want to use unify code to typeset quantities above/below the arrows.

#import "@test/typsium:0.3.2": ce
#import "@preview/unify:0.8.0": num, numrange, qty, qtyrange, unit

#ce[A ->[test] B]

#ce[A ->[#num("2.3")] B]

#ce[A ->[#qty("2.3", "electronvolt")] B]

#ce[A ->[(n, test)] B]

#let tt = "(n, test)"
#ce[A ->[#tt] B]

#let tt = text("(n, test)")
#ce[A ->[#tt] B]

// #ce[A ->[(n, α)] B]

// #let tt = text("(n,α)")
// #ce[A ->[#tt] B]

Expected behavior

• The last three lines of the code example don't compile
• I would expect the "te" and "st" of the word "test" to be formatted the same way
• I would expect to be able to use greek letters (all of unicode?) in arrow labels

Screenshots
If applicable, add screenshots to help explain your problem.

Platform (please complete the following information):

• Editor: neovim
• LSP: tinymist
• Typst version: 0.14.2
• Typsium version: github clone of latest version/local install

Additional context
Please feel free to add any other context about the problem here.

[gvdeynde]

test.pdf

[tryptophawa]

The second issue concerns how Unicode in content is parsed. It is probably that Unicode takes up 2 index places, but typsium only counts as one letter, and causes an out-of-index issue. It can be easily fixed by adding
for _ in range(value.len()) { templates.push(()) } in parse-content... .typ file, line 670.

[tryptophawa]

The first one is because there was a mistake in the regex rule, which includes the lowercase letters in elements; it should be ok now.

[gvdeynde]

Thanks for the quick response. The greek letters are now ok, but the word "test" is still in two different fonts...

[tryptophawa]

oh it looks weird, I think I should reopen it.

[tryptophawa]

This issue conflicts with issue #37. Since typst doesn't support look ahead and the inputed content is fragmented into pieces like:

It is very hard to fix it.

[tryptophawa]

There is a way to fix it by using "([A-Zα-ωΑ-Ω][a-z]?|.[a-z])" so that you have to add a space in front of your lowercase elements, such as #ce[^1_1 p] instead of #ce[^1_1p]

[tryptophawa]

This issue should be considered with much more care than I thought.