exposing uniqueIdent and toAsciiText to lua api in filters

[massifrg]

This is a feature request.

uniqueIdent and toAsciiText are functions used internally by pandoc to generate identifiers, but only for formats that support the auto_identifiers and ascii_identifiers extensions.

They could be useful in lua filters too.

In my particular case, I'm looking for a function that takes a list of Inlines and transforms it to a normalized (low case, ascii chars only) string. I'm going to use it to automatically generate a sort-key attribute for index terms, to be manually refined by the user to tune the sorting of an index.

They could go under pandoc.utils or pandoc.text packages.

[tarleb]

Exposing uniqueIdent makes perfect sense, feel free to open an issue for that. In case you also want to provide a PR: the necessary code would go into pandoc-lua-engine/src/Text/Pandoc/Lua/Module/Utils.hs. It should be just a couple of lines.

I'm less certain about toAsciiText, but I guess adding that to pandoc.text is reasonable.

[massifrg]

Thank you @tarleb for your guidance, I'm going to make a PR.

Just yesterday I recompiled pandoc from source for the first time.
Months ago, I already tried to read and understand the code under pandoc-lua-engine.
Though I understood the logic, I was not able to add any code.
This task seems to be feasible; let's see...

[tarleb]

The make_sections function is probably closest to what's needed. Feel free to ping me directly if you run into problems.