llogick

It would avoid any subjectivity, see - #1680 - #1681 https://rust-analyzer.github.io/manual.html#user-snippet-completions

Maybe ```diff diff --git a/src/features/completions.zig b/src/features/completions.zig index ce92b9a..4460de3 100644 --- a/src/features/completions.zig +++ b/src/features/completions.zig @@ -1491,6 +1491,7 @@ fn collectFieldAccessContainerNodes( if (field_access_slice[0] == '@') break :blk 1; // assume `@import("..").some.Other{.}` var...

Yes, I was very [fond](https://github.com/zigtools/zls/issues/1301) of this feature and would like to see it come back or at least as an option. The original motivation as stated in that issue...

I keep repeating it, but before anyone dives into the completions logic — parser limitation. The workarounds (I know, not ideal): - structs => `S{1.};` - switch => `switch (e)...

I don't mean to nag, simply posted it as a heads-up to anyone that might look into it, and the workarounds as a proof/way to verify that the issue stems...

Notes: - Known (parser) limitation. - Workaround: Do `digit` `dot` as if this is a float number, eg `1.`

Would you happen to be using Copilot?

There's nothing wrong w/ Copilot, it's just that I knew ZLS doesn't provide completions for any of those but `list_of_list.` (there's https://github.com/zigtools/zls/pull/1373 but it is quite partial and pretty sure...

I did consider the scenario when implementing it and did and have had considered renaming the option to `highlight_root_level_var_decls`, but it was a messy PR as it were. Zig doesn't...

https://github.com/ziglang/zig/compare/hcs-win-poc