Idea [from Discord
conversation](https://discord.com/channels/1054443721975922748/1318670789150642246/1319338465757757562).
Let's users more precisely control how short-form and long-form terms
are used based on the mode ("short", "long", or "both").
Defaults to `__default-format-term()` as below:
```typst
// Formats a term string based on the specified mode.
//
// Given mode, long-form, and short-form, this function returns the formatted
// term based on the mode. When mode is ...
//
// - "short" -> returns "short-form"
// - "long" -> returns "long-form"
// - "both" -> returns "short-form (long-form)"
//
// Parameters:
// mode (string): The mode in which to format the term. Possible values are "short", "long", or "both".
// short-form (string): The short form of the term.
// long-form (string): The long form of the term.
//
// Returns:
// string: The formatted term based on the specified mode.
//
#let __default-format-term(mode, short-form, long-form) = {
if mode == "short" {
short-form
} else if mode == "long" {
long-form
} else {
// mode assumed to be "both"
long-form + " (" + short-form + ")"
}
}
```
- Uniqfies the duplicate term usage tracking via counter + state to just using the counter. Refreactor the code with the two methods "mark_term_used" and "is_term_used".
- Adds a label with the entry key for autocompletion
- either in the glossary (if it is visible and there is a corresponding entry)
- or invisible during init-glossary
- Use a single label for all usages of the entry in the text. This simplifies the back-referencing logic significantly and also avoids cluttering the auto-complete list
closes https://github.com/swaits/typst-collection/issues/3
swaits
Philipp Niedermayer