Improved Rust attributes

[kaivol]

Rust attributes with nested brackets are currently not correctly highlighted.

For example, in

#[foo(bar = [baz, qux])]

the attribute is only highlighted until the first closing square bracket.

I think this should fixed with these changes.

[Anteru]

Can you please add a test for this? There's a rust example file, it should be sufficient to add one line with an attribute like that there.

[kaivol]

You mean like that?

Unfortunately I'm not familiar with pygments so I'm not exactly sure what I'm doing here...

[Anteru]

Yeah, adding it to the example file is sufficient, thanks!

[dario23]

i just stumbled over this problem as well. as far as i can see, the diff needed to make the rust tests pass is

diff --git a/tests/examplefiles/rust/eval.rs.output b/tests/examplefiles/rust/eval.rs.output
index 186a7c0c..dd0f7997 100644
--- a/tests/examplefiles/rust/eval.rs.output
+++ b/tests/examplefiles/rust/eval.rs.output
@@ -218,7 +218,10 @@
 '/// Represents a value (either 16-bit or 32-bit) at runtime.\n' Literal.String.Doc
 
 '#['          Comment.Preproc
-'derive(Clone, PartialEq, Eq, Debug)' Comment.Preproc
+'derive'      Comment.Preproc
+'('           Comment.Preproc
+'Clone, PartialEq, Eq, Debug' Comment.Preproc
+')'           Comment.Preproc
 ']'           Comment.Preproc
 '\n'          Text.Whitespace

aka the #[derive(..)] line is now looked at in more detail as well, just like the initial example.

[birkenfeld]

Actually, why are we tracking ( and ) pairs? The only thing that's relevant is nested [ and ]...

[kaivol]

For highlighting purposes it should indeed be sufficient to look at square brackets.
Should I change the code accordingly?

[birkenfeld]

That would be nice, and it will also reduce the number of individual tokens.

[birkenfeld]

Thanks!