fix: Skip cfg attributes in macro input attribute stripping

[eihqnh]

Description:

This fixes a regression introduced in commit f0e372c ("Rewrite attribute handling"), where cfg attributes
were incorrectly counted during macro input stripping.

Problem

When using attribute macros like maybe-async-cfg, if cfg attributes are present on the item, rust-analyzer may
fail to strip the macro attribute itself from the input. This results in the macro being expanded recursively
(e.g., TestStruct -> TestStructAsync(Correct) -> TestStructAsyncAsync(Wrong)).

Cause

The macro_input_callback function in crates/hir-expand/src/cfg_process.rs uses a counter (item_tree_attr_id)
to identify which attribute to strip. This counter was being incremented for cfg attributes.

However, collect_item_tree_attrs (which assigns the attribute IDs) filters out cfg attributes. This caused the
counter in macro_input_callback to desynchronize with the actual attribute IDs, leading to the macro attribute
not being identified and stripped.

Fix

Explicitly skip cfg attributes when determining whether to call should_strip_attr() in macro_input_callback.
This ensures the attribute counter remains in sync with the item tree.

Repro

#[maybe_async_cfg::maybe(sync(keep_self), async(feature = "async"))]
#[derive(Debug)]
pub struct TestStruct {
    pub value: u32,
}

#[cfg(feature = "async")]
pub fn test_async_only() {
    let _x = TestStructAsync { value: 42 };
}
#[cfg(not(feature = "async"))]
pub fn test_sync_only() {
    let _y = TestStruct { value: 1 };
}
pub fn test_both_exist() {
    let _a = TestStruct { value: 1 };

    let _b = TestStructAsync { value: 2 };
}

[package]
name = "test-maybe-async"
version = "0.1.0"
edition = "2024"

[dependencies]
maybe-async-cfg = "=0.2.4"

[features]
async = []

[ChayimFriedman2]

Thanks, that's my mistake.

LGTM, but can you construct a test? (I understand if it's too hard).

[eihqnh]

Thanks, that's my mistake.

LGTM, but can you construct a test? (I understand if it's too hard).

I added a simple test. Could you give me some suggestions?
cargo test -p hir-def macro_expansion_tests::proc_macros::attribute_macro_stripping_with_cfg

without this change

cargo test -p hir-def macro_expansion_tests::proc_macros::attribute_macro_stripping_with_cfg

running 1 test
test macro_expansion_tests::proc_macros::attribute_macro_stripping_with_cfg ... FAILED

failures:

---- macro_expansion_tests::proc_macros::attribute_macro_stripping_with_cfg stdout ----


error: expect test failed
   --> crates/hir-def/src/macro_expansion_tests/proc_macros.rs:354:9

You can update all `expect!` tests by running:

    env UPDATE_EXPECT=1 cargo test

To update a single test, place the cursor on `expect` token and use `run` feature of rust-analyzer.

Expect:
----
#[cfg(all())]
#[proc_macros::generate_suffixed_type]
struct S;

struct S;
struct SSuffix;
----

Actual:
----
#[cfg(all())]
#[proc_macros::generate_suffixed_type]
struct S;

----

Diff:
----
#[cfg(all())]
#[proc_macros::generate_suffixed_type]
struct S;

struct S;
struct SSuffix;
----

failures:
    macro_expansion_tests::proc_macros::attribute_macro_stripping_with_cfg

test result: FAILED. 0 passed; 1 failed; 0 ignored; 0 measured; 466 filtered out; finished in 0.02s

error: test failed, to rerun pass `-p hir-def --lib`

with this change

test macro_expansion_tests::proc_macros::attribute_macro_stripping_with_cfg ... ok
test result: ok. 1 passed; 0 failed; 0 ignored; 0 measured; 466 filtered out; fin

[ChayimFriedman2]

Thanks!