Handling Elixir identifiers with characters that are not allowed in JavaScript identifiers

[elixirscript]

Handling Elixir identifiers with characters that are not allowed in JavaScript identifiers

[grabbou]

The most clever solution would've been to rename these methods to something like user.u1 following reasonable pattern and always use square brackets notation. I would avoid throwing any errors. The compiler should be clever enough to handle all the cases and translate the code in the way it works in the host environment.

[bryanjos]

It gets kind of messy. Like what to do with variables? For instance in the case below:

allowed? = true

would be translated into something like

let allowed? = true;

Which isn't valid. And then there are function parameters as well. If there is some reasonable pattern to use I'm all for it. Something I would like something that also looks decent if calling it from outside of ElixirScript (i.e. in plain javascript).

[grabbou]

Maybe just building a dictionary with all forbidden JS characters yet allowed in Elixir and replacing during conversion would do the trick?

Let's say

allowed_qmark = true

if calling it from outside of ElixirScript (i.e. in plain javascript).

Haven't seen such a use-case in already implemented e.g. ClojureScript. I think the main focus should be on implementing as many Elixir features as possible yet keeping the generated code as efficient yet small as possible ;)

[bryanjos]

That could definitely work. I'll give it a shot. Thanks!

[bryanjos]

Handling it as defined here. ? becomes __qmark__ and ! becomes __emark__. The scheme may change in the future but this works for now.