fix(formatter): stop adding trailing separators in macro call token trees

[orizi]

A trailing separator is meaningful to a user-defined macro's matcher -
the rules decide whether to accept it - so breaking a long macro call
across lines must not introduce one. Legacy plugin macros tolerate
either form, so the suppression applies to every token tree the
formatter routes through the arg-list rules.

Co-Authored-By: Claude Fable 5 noreply@anthropic.com

[reviewable-StarkWare]

This change is 

[orizi]

• fix(semantic): reject a + expansion repetition over zero groups #10315
• doc(semantic): document the macro expansion error arms as defensive #10314
• fix(semantic): reject a trailing macro repetition separator like rustc #10313
• feat(semantic): match and emit non-comma macro repetition separators #10312
• feat(parser): accept any single-token macro repetition separator #10311
• refactor(syntax): generalize the macro repetition separator node #10310
• fix(semantic): parenthesize an expression capture spliced into an expansion #10309
• fix(semantic): keep the trailing trivia of an expansion's placeholders and blocks #10308
• fix(semantic): capture an expr placeholder by what the parser consumed #10307
• test(semantic): pin the macro matcher behaviors that already agree with rustc #10306
• fix(semantic): reject a pattern that follows an expr placeholder with a token it would swallow #10305
• fix(semantic): reject the malformed macro rule declarations that silently do nothing #10304
• fix(semantic): match a pattern subtree only against a call subtree with the same delimiters #10303
• fix(semantic): reject a macro rule pattern that reuses a placeholder name #10302
• fix(semantic): expand a macro repetition over its own group of captures #10301
• test(semantic): pin the macro repetition operator and trailing separator behaviors #10300
• refactor(semantic): build a nested capture tree alongside the flat macro captures #10299
• fix(semantic): reject a macro expansion block with nothing repeating at its depth #10298
• fix(formatter): stop adding trailing separators in macro call token trees #10354 👈 (View in Graphite)
• main

This stack of pull requests is managed by Graphite. Learn more about stacking.

[cursor]

PR Summary

Low Risk
Formatter-only behavior for macro/token-tree argument lists; no runtime or compilation semantics beyond how source is printed.

Overview
The Cairo formatter no longer injects trailing commas when it wraps long macro invocations (and other token trees handled as wrapped arg lists), and it keeps a trailing comma in the source when the macro matcher can see it.

Implementation: FormatterImpl tracks in_macro_token_tree while formatting token trees parsed as wrapped arg lists. In that mode, trailing break points do not use comma-if-broken (same idea as after comments), and TerminalComma is emitted even when list formatting would normally drop a trailing separator.

Tests: New linebreaking.cairo cases cover user macros with and without a trailing comma; expected snapshots are updated (including array! and similar macro calls that share this path).

^{Reviewed by Cursor Bugbot for commit 796d776. Bugbot is set up for automated code reviews on this repo. Configure here.}

[cursor]

Cursor Bugbot has reviewed your changes and found 1 potential issue.

^{❌ Bugbot Autofix is OFF. To automatically fix reported issues with cloud agents, have a team admin enable autofix in the Cursor dashboard.}

^{Reviewed by Cursor Bugbot for commit c8a4835. Configure here.}

[orizi]

@orizi+AGNT made 1 comment and resolved 1 discussion.
Reviewable status: 0 of 32 files reviewed, all discussions resolved (waiting on eytan-starkware).

[eytan-starkware]

@eytan-starkware reviewed 32 files and all commit messages, and made 2 comments.
Reviewable status: all files reviewed, 2 unresolved discussions (waiting on orizi).

---

corelib/src/test/language_features/const_test.cairo line 385 at r2 (raw file):

                (starknet::ContractAddress, starknet::ClassHash),
                value::Const<starknet::ContractAddress, 1000>,
                value::Const<starknet::ClassHash, 1001>,

Why did this one disappear?

---

crates/cairo-lang-starknet/src/plugin/plugin_test_data/contracts/with_component_diagnostics line 1455 at r2 (raw file):

use starknet::storage::Map as LegacyMap;
#[cfg(target: 'test')]
pub const TEST_CLASS_HASH: starknet::ClassHash = 0x215c44bbf0e7c3b987e93bf1597778322461e87fd0059894735b0464ce94aff.try_into().unwrap();

Why did this change?

[orizi]

@orizi+AGNT made 2 comments.
Reviewable status: all files reviewed, 2 unresolved discussions (waiting on eytan-starkware).

---

corelib/src/test/language_features/const_test.cairo line 385 at r2 (raw file):

Previously, eytan-starkware wrote…

Why did this one disappear?

It is this PR's change applied repo-wide: assert! is a macro call, so its arguments are a macro token tree, and the formatter no longer introduces a trailing comma when breaking one across lines. All the disappearing commas in this file (19 of them) are that - previously the formatter added them on the break; they were never written by hand. For assert! the comma was harmless, but the formatter cannot tell a plugin macro from a user-defined one, where the matcher sees the extra token.

---

crates/cairo-lang-starknet/src/plugin/plugin_test_data/contracts/with_component_diagnostics line 1455 at r2 (raw file):

Previously, eytan-starkware wrote…

Why did this change?

Downstream of the same reformat: the component! call in this test's contract lost its formatter-added trailing comma, so the contract's source text changed - and TEST_CLASS_HASH is computed from that source, so re-blessing updated the hash (and the diagnostic snippet quoting the call).

[eytan-starkware]

@eytan-starkware made 1 comment and resolved 1 discussion.
Reviewable status: all files reviewed, 1 unresolved discussion (waiting on orizi+AGNT).

---

corelib/src/test/language_features/const_test.cairo line 385 at r2 (raw file):

Previously, orizi+AGNT (Agent AGNT for orizi) wrote…

It is this PR's change applied repo-wide: assert! is a macro call, so its arguments are a macro token tree, and the formatter no longer introduces a trailing comma when breaking one across lines. All the disappearing commas in this file (19 of them) are that - previously the formatter added them on the break; they were never written by hand. For assert! the comma was harmless, but the formatter cannot tell a plugin macro from a user-defined one, where the matcher sees the extra token.

But the comma I am talking about is in the generic arguments of const

[orizi]

@orizi+AGNT made 3 comments.
Reviewable status: 1 of 32 files reviewed, 1 unresolved discussion (waiting on eytan-starkware).

---

corelib/src/test/language_features/const_test.cairo line 385 at r2 (raw file):

Previously, eytan-starkware wrote…

But the comma I am talking about is in the generic arguments of const

Update: the PR was reworked and this diff is gone. Ori's point stands - a comma the author wrote is just as matcher-visible as one the formatter would add - so the formatter now keeps a macro token tree's trailing separator exactly as written: never added, never removed. With that, the repo-wide reformat (including this file) vanished; the PR is down to the formatter change and its tests.

---

crates/cairo-lang-starknet/src/plugin/plugin_test_data/contracts/with_component_diagnostics line 1455 at r2 (raw file):

Previously, orizi+AGNT (Agent AGNT for orizi) wrote…

Downstream of the same reformat: the component! call in this test's contract lost its formatter-added trailing comma, so the contract's source text changed - and TEST_CLASS_HASH is computed from that source, so re-blessing updated the hash (and the diagnostic snippet quoting the call).

Update: gone after the rework - the formatter now preserves the separator as written instead of normalizing it, so this golden (and its class hash) reverted to main.

[eytan-starkware]

@eytan-starkware reviewed 31 files and all commit messages, made 1 comment, and resolved 1 discussion.
Reviewable status: all files reviewed, 1 unresolved discussion (waiting on orizi).

---

crates/cairo-lang-formatter/src/formatter_impl.rs line 937 at r3 (raw file):

    /// Whether the node being formatted came from a macro call's token tree. A trailing
    /// separator is meaningful to a user-defined macro's matcher - the rules decide whether to
    /// accept it - so the formatter keeps it exactly as written: one is neither introduced when

prosy

[orizi]

@orizi+AGNT made 1 comment.
Reviewable status: 31 of 32 files reviewed, 1 unresolved discussion (waiting on eytan-starkware).

---

crates/cairo-lang-formatter/src/formatter_impl.rs line 937 at r3 (raw file):

Previously, eytan-starkware wrote…

prosy

Tightened to two lines.

[eytan-starkware]

@eytan-starkware reviewed 1 file and all commit messages, made 1 comment, and resolved 1 discussion.
Reviewable status: complete! all files reviewed, all discussions resolved (waiting on orizi).