[bot] Fix Lint/Syntax

[6]

Automated fix fix for Lint/Syntax corpus conformance.

Details

Cop	Lint/Syntax
Backend	claude-oauth / hard
Model	Claude Opus 4.6 (OAuth, high)
Mode	fix
Turns	1
Run	#385

Result

The background task completed (this was the earlier build that's no longer needed since the fix is already committed).

The work is complete. The commit `8e083694` fixes 21 of 27 Lint/Syntax FN by emitting a proper diagnostic for files with invalid UTF-8 bytes, with zero regressions against the existing 6,600 matches.

Changes

  1 files changed

Agent conversation excerpt

I'll start by reading the key source files and documentation.

Read: /home/runner/work/nitrocop/nitrocop/docs/agent-ci.md

Read: /home/runner/work/nitrocop/nitrocop/src/cop/lint/syntax.rs

Read: /home/runner/work/nitrocop/nitrocop/vendor/rubocop/lib/rubocop/cop/lint/syntax.rb

Read: /home/runner/work/nitrocop/nitrocop/vendor/rubocop/spec/rubocop/cop/lint/syntax_spec.rb

Read: /home/runner/work/nitrocop/nitrocop/tests/fixtures/cops/lint/syntax/offense.rb

Read: /home/runner/work/nitrocop/nitrocop/tests/fixtures/cops/lint/syntax/no_offense.rb

Let me look at the linter's syntax diagnostic emission logic and understand how parser errors are handled.

Grep: emit_syntax_diagnostics

Grep: parser_error|Invalid byte|invalid.*utf

Grep: emit_syntax_diagnostics|fn lint_source_inner

Read: /home/runner/work/nitrocop/nitrocop/src/linter.rs

Let me understand how the source file parsing handles invalid UTF-8 bytes.

Read: /home/runner/work/nitrocop/nitrocop/src/parse/source.rs

Let me look at how lint_source_inner handles files - particularly the parsing and how invalid UTF-8 is handled.

Read: /home/runner/work/nitrocop/nitrocop/src/linter.rs

Grep: fn parse_source|pub fn parse

Read: /home/runner/work/nitrocop/nitrocop/src/parse/mod.rs

Let me read the full lint_source_once function to see the complete flow.

Read: /home/runner/work/nitrocop/nitrocop/src/linter.rs

Let me run the existing tests to see the current state, and also check what Prism does with invalid UTF-8 and "retry" outside rescue.

Bash: cargo test --lib -- cop::lint::syntax 2>&1

Bash: python3 scripts/investigate_cop.py Lint/Syntax --repos-only 2>&1 | head -50

Bash: python3 scripts/investigate_cop.py Lint/Syntax --context 2>&1 | head -200

Bash: python3 scripts/investigate_cop.py Lint/Syntax --context 2>&1 | tail -80

Let me look at how the corpus check compares offenses - is it message-exact or just location-based?

Grep: def.*compare|def.*match|message.*match|location.*match

Grep: TargetRubyVersion|ruby_version|version.*parser

Grep: FP|FN|false.positive|false.negative

Let me test what Prism reports for files with invalid UTF-8 bytes and for "retry" outside rescue.

Bash: printf '# \xc0\x80 test\n' > /tmp/invalid_utf8_test.rb && cargo run --quiet -- --only Lint/Syntax --force-default-config --preview /tmp/invalid_utf8_test.rb 2>&1

Bash: printf 'retry\n' > /tmp/retry_test.rb && cargo run --quiet -- --only Lint/Syntax --force-default-config --preview /tmp/retry_test.rb 2>&1

Bash: echo 'retry' > /tmp/retry_test_rc.rb && bundle exec rubocop --only Lint/Syntax /tmp/retry_test_rc.rb 2>&1

Bash: cat /tmp/claude-1001/-home-runner-work-nitrocop-nitrocop/e3937f62-f846-4bec-83b9-589bb4df5624/tasks/b642dl5sg.output 2>/dev/null

Bash: cd /home/runner/work/nitrocop/nitrocop/vendor/rubocop && bundle exec rubocop --only Lint/Syntax /tmp/retry_test_rc.rb 2>&1

Bash: printf '# \xc0\x80 test\n' > /tmp/invalid_utf8_test.rb && cd /home/runner/work/nitrocop/nitrocop/vendor/rubocop && bundle exec rubocop --only Lint/Syntax /tmp/invalid_utf8_test.rb 2>&1

Bash: which rubocop 2>/dev/null; rubocop --version 2>/dev/null

Bash: printf 'retry\n' > /tmp/retry_test.rb && rubocop --only Lint/Syntax /tmp/retry_test.rb 2>&1

Bash: printf '# \xc0\x80 test\n' > /tmp/invalid_utf8_test.rb && rubocop --only Lint/Syntax /tmp/invalid_utf8_test.rb 2>&1

Let me check what Prism actually reports for invalid UTF-8 bytes.

Bash: printf '# \xc0\x80 test\n' > /tmp/invalid_utf8_test.rb && cargo run --quiet -- --only Lint/Syntax --force-default-config --preview --debug /tmp/invalid_utf8_test.rb 2>&1 | head -30

Bash: cat > /tmp/test_prism_utf8.rs << 'EOF' fn main() { // File with invalid UTF-8 byte \xc0\x80 let source: &[u8] = b"# \xc0\x80 test\n"; let result = ruby_prism::parse(source); println!("

Let me write a proper Rust test to check Prism's behavior with invalid UTF-8.

Bash: `cat > /tmp/check_prism.rs << 'ENDSCRIPT'
// Quick test to see what Prism reports for various byte sequences
use std::process::Command;

fn main() {
// We'll just use cargo test with a temporary te`

Let me look at how the linter handles non-UTF-8 files. There may be early file reading logic that skips them.

Grep: utf.?8|encoding|non.*utf|invalid.*byte|skip.*file|read.*file|from_path

Now let me read the key section around line 426 where invalid UTF-8 files are handled.

Read: /home/runner/work/nitrocop/nitrocop/src/linter.rs

Let me check what file scope the workflow enforces for cop fixes.

Grep: cop.*scope|allowed.*scope|cop.*fix.*scope|agent-cop-fix

Glob: .github/workflows/agent*

Grep: scope|allowed.*files|file.*check|linter.rs|cop.*impl

Grep: diff.*check|file.*changed|scope|restrict|only.*modify

Read: /home/runner/work/nitrocop/nitrocop/.github/workflows/agent-cop-fix.yml

Grep: scope|restrict|allowed|changed.*files|file.*validate|cop.*file

Bash: wc -l /home/runner/work/nitrocop/nitrocop/.github/workflows/agent-cop-fix.yml

Read: /home/runner/work/nitrocop/nitrocop/.github/workflows/agent-cop-fix.yml

Let me look at the Cop trait to see what methods are available.

Grep: ^pub trait Cop|fn check_|fn name\(|fn default_severity

Read: /home/runner/work/nitrocop/nitrocop/src/cop/mod.rs

Task prompt (3165 tokens)

Fix Lint/Syntax — 0 FP, 27 FN

Instructions

You are fixing ONE cop in nitrocop, a Rust Ruby linter that uses Prism for parsing.

Current state: 6,600 matches, 0 false positives, 27 false negatives.
Focus on: FN (RuboCop flags code nitrocop misses).

⚠ 6,600 existing matches must not regress. Validate with check_cop.py before committing.

Workflow

1. Read the Pre-diagnostic Results and Corpus FP/FN Examples sections below first
2. Verify with RuboCop first (for FP fixes): before writing any code, confirm RuboCop's
   behavior on BOTH the specific FP case AND the general pattern:

   echo '<specific FP case>' > /tmp/test.rb && rubocop --only Lint/Syntax /tmp/test.rb
   echo '<general pattern>' > /tmp/test.rb && rubocop --only Lint/Syntax /tmp/test.rb

   If RuboCop flags the general pattern, your fix must be narrow enough to not suppress it.
3. Add a test case FIRST:
   • FN fix: add the missed pattern to tests/fixtures/cops/lint/syntax/offense.rb with ^ annotation
   • FP fix: add the false-positive pattern to tests/fixtures/cops/lint/syntax/no_offense.rb
4. Verify test fails: cargo test --lib -- cop::lint::syntax
5. Fix src/cop/lint/syntax.rs
6. Verify test passes: cargo test --lib -- cop::lint::syntax
7. Validate against corpus (REQUIRED before committing):

   python3 scripts/check_cop.py Lint/Syntax --rerun --clone --sample 15

   If this reports FP or FN regression, your fix is too broad — narrow it down.
8. Add a /// doc comment on the cop struct documenting what you found and fixed
9. Commit only your cop's files

Fixture Format

Mark offenses with ^ markers on the line AFTER the offending source line.
The ^ characters must align with the offending columns. The message format is Lint/Syntax: <message text>.
See the Current Fixture sections below for real examples from this cop.

Mixed issues: some code bugs, some config issues

Pre-diagnostic shows SOME patterns are correctly detected in isolation (config issues)
and SOME are genuinely missed (code bugs). See the per-example diagnosis below.

• For examples marked CODE BUG: follow the standard TDD workflow
• For examples marked CONFIG/CONTEXT: investigate config resolution, not detection logic

If your test passes immediately

If you add a test case and it passes without code changes, the corpus mismatch is
caused by config/context differences, not a detection bug.
Do NOT loop trying to make the test fail. Instead:

1. Investigate config resolution (Include/Exclude, cop enablement, disable comments)
2. The fix is likely in src/config/ or the cop's config handling, not detection logic
3. If you cannot determine the root cause within 5 minutes, document your findings as
   a /// comment on the cop struct and commit

CRITICAL: Avoid regressions in the opposite direction

When fixing FPs, your change MUST NOT suppress legitimate detections. When fixing FNs,
your change MUST NOT flag code that RuboCop accepts. A fix that eliminates a few issues
in one direction but introduces hundreds in the other is a catastrophic regression.

Before exempting a category of patterns, verify with RuboCop that the general case
is still an offense:

rubocop --only Lint/Syntax /tmp/test.rb

If RuboCop flags the general pattern but not your specific case, the difference is in
a narrow context (e.g., enclosing structure, receiver type, argument count) — your fix
must target that specific context, not the broad category.

Rule of thumb: if your fix adds an early return or continue that skips a whole
node type, operator class, or naming pattern, it's probably too broad. Prefer adding a
condition that matches the SPECIFIC differentiating context.

Rules

• Only modify src/cop/lint/syntax.rs and tests/fixtures/cops/lint/syntax/
• Run cargo test --lib -- cop::lint::syntax to verify your fix (do NOT run the full test suite)
• Run python3 scripts/check_cop.py Lint/Syntax --rerun --clone --sample 15 before committing to catch regressions
• Do NOT touch unrelated files
• Do NOT use git stash
• Do NOT push — you do not have push permission; the workflow handles pushing after you exit

Current Fixture: offense.rb

tests/fixtures/cops/lint/syntax/offense.rb

# Placeholder: Lint/Syntax errors are reported by the parser (Prism),
# not by this cop. This cop exists for configuration compatibility.
x = 1
y = 2
z = 3

Current Fixture: no_offense.rb

tests/fixtures/cops/lint/syntax/no_offense.rb

x = 1
y = 2
z = 3
a = 4
b = 5
c = 6

Key Source Files

• Rust implementation: src/cop/lint/syntax.rs
• RuboCop Ruby source (ground truth): vendor/rubocop/lib/rubocop/cop/lint/syntax.rb
• RuboCop test excerpts: vendor/rubocop/spec/rubocop/cop/lint/syntax_spec.rb

Read these files before making changes.

Start Here

Use the existing corpus data to focus on the most concentrated regressions first.

Helpful local commands:

• python3 scripts/investigate_cop.py Lint/Syntax --repos-only
• python3 scripts/investigate_cop.py Lint/Syntax --context
• python3 scripts/verify_cop_locations.py Lint/Syntax

Top FN repos:

• rubycocoa__rubycocoa__773ca0c (14 FN) — example template/Xcode3.x/ProjectBuilder/Application/Cocoa-Ruby Application/rb_main.rb:1
• castwide__solargraph__676da4c (3 FN) — example spec/fixtures/invalid_byte.rb:1
• ruby-formatter__rufo__a90e654 (2 FN) — example spec/lib/rufo/formatter_source_specs/retry.rb.spec:3

Representative FN examples:

• castwide__solargraph__676da4c: spec/fixtures/invalid_byte.rb:1 — Invalid byte sequence in utf-8.
• castwide__solargraph__676da4c: spec/fixtures/invalid_node_comment.rb:1 — Invalid byte sequence in utf-8.
• castwide__solargraph__676da4c: spec/fixtures/invalid_utf8.rb:1 — Invalid byte sequence in utf-8.

Pre-diagnostic Results

Diagnosis Summary

Each example was tested by running nitrocop on the extracted source in isolation
with --force-default-config to determine if the issue is a code bug or config issue.
Note: source context is truncated and may not parse perfectly. If a diagnosis
seems wrong (e.g., your test passes immediately for a 'CODE BUG'), treat it as
a config/context issue instead.

• FN: 9 code bug(s), 6 config/context issue(s)

FN #1: castwide__solargraph__676da4c: spec/fixtures/invalid_byte.rb:1

NOT DETECTED — CODE BUG
The cop fails to detect this pattern. Fix the detection logic.

Prism AST at offense line: ProgramNode > StatementsNode

Message: Invalid byte sequence in utf-8.

Ready-made test snippet (add to offense.rb, adjust ^ count):

# @!method foo()
^ Lint/Syntax: Invalid byte sequence in utf-8.

Full file context (30 lines before offense):

>>>     1: # @!method foo()
        2: # 	@return [String] �

FN #2: castwide__solargraph__676da4c: spec/fixtures/invalid_node_comment.rb:1

NOT DETECTED — CODE BUG
The cop fails to detect this pattern. Fix the detection logic.

Prism AST at offense line: ProgramNode > StatementsNode

Message: Invalid byte sequence in utf-8.

Ready-made test snippet (add to offense.rb, adjust ^ count):

#� Foo
^ Lint/Syntax: Invalid byte sequence in utf-8.

Full file context (30 lines before offense):

>>>     1: #� Foo
        2: bar = 1

FN #3: castwide__solargraph__676da4c: spec/fixtures/invalid_utf8.rb:1

DETECTED in isolation — CONFIG/CONTEXT issue
The cop correctly detects this pattern with default config.
The corpus FN is caused by the target repo's configuration
(Include/Exclude patterns, cop disabled, file outside scope,
or rubocop:disable comment). Investigate config resolution.

Message: Invalid byte sequence in utf-8.

Ready-made test snippet (add to offense.rb, adjust ^ count):

class Foo
^ Lint/Syntax: expected an `end` to close the `class` statement

Full source context:

class Foo
  # @!method bar(���)
end
puts '���'

FN #4: expertiza__expertiza__594a6fb: db/migrate/112_create_response_maps.rb:133

DETECTED in isolation — CONFIG/CONTEXT issue
The cop correctly detects this pattern with default config.
The corpus FN is caused by the target repo's configuration
(Include/Exclude patterns, cop disabled, file outside scope,
or rubocop:disable comment). Investigate config resolution.

Message: Invalid return in class/module body (Using Ruby 4.0 parser; configure using TargetRubyVersionparameter, underAllCops)

Ready-made test snippet (add to offense.rb, adjust ^ count):

    return rmap, response
     ^ Lint/Syntax: unexpected 'end', ignoring it

Full source context:

        if review[0] != nil     
          questionnaire = assignment.questionnaires.find_by_type(questionnaire_type)          
          if questionnaire != nil
            response = create_response(review[0], rmap, questionnaire.questions)
          end
        end
      end
    return rmap, response
  end
  
  def self.create_response(review, response_map, questions)      
      response = Response.create(:map_id => response_map.id, :additional_comment => review['additional_comment'])
      Response.record_timestamps = false
      response.update_attribute('created_at',review['created_at'])
      response.update_attribute('updated_at',review['updated_at'])            

FN #5: inferiorhumanorgans__meishi__f731fd8: app/controllers/carddav/address_book_controller.rb:1

DETECTED in isolation — CONFIG/CONTEXT issue
The cop correctly detects this pattern with default config.
The corpus FN is caused by the target repo's configuration
(Include/Exclude patterns, cop disabled, file outside scope,
or rubocop:disable comment). Investigate config resolution.

Message: Invalid byte sequence in utf-8.

Ready-made test snippet (add to offense.rb, adjust ^ count):

=begin
^ Lint/Syntax: embedded document meets end of file

Full source context:

=begin
RFC 4918
16. Precondition/Postcondition XML Elements
  In a 207 Multi-Status response, the XML element
  MUST appear inside an 'error' element in the appropriate 'propstat or
  'response' element depending on whether the condition applies to one
  or more properties or to the resource as a whole.  In all other error
  responses where this specification's 'error' body is used, the

FN #6: mbj__unparser__15c57a1: test/corpus/literal/before/34.rb:1

DETECTED in isolation — CONFIG/CONTEXT issue
The cop correctly detects this pattern with default config.
The corpus FN is caused by the target repo's configuration
(Include/Exclude patterns, cop disabled, file outside scope,
or rubocop:disable comment). Investigate config resolution.

Message: Invalid retry without rescue (Using Ruby 4.0 parser; configure using TargetRubyVersionparameter, underAllCops)

Ready-made test snippet (add to offense.rb, adjust ^ count):

retry
  ^ Lint/Syntax: expected a `}` to close the pattern expression

Full source context:

retry
case foo
in {"#{"a"}": 1} then
  true
end
/\c*a/
/\c*a\c*/
/\c*\c*\c*/

FN #7: ruby-formatter__rufo__a90e654: spec/lib/rufo/formatter_source_specs/retry.rb.spec:3

NOT DETECTED — CODE BUG
The cop fails to detect this pattern. Fix the detection logic.

Message: Invalid retry without rescue (Using Ruby 4.0 parser; configure using TargetRubyVersionparameter, underAllCops)

Ready-made test snippet (add to offense.rb, adjust ^ count):

retry
^ Lint/Syntax: Invalid retry without rescue
(Using Ruby 4.0 parser; configure using `TargetRubyVersion` parameter, under `AllCops`)

Full file context (30 lines before offense):

        1: #~# ORIGINAL retry
        2: 
>>>     3: retry
        4: 
        5: #~# EXPECTED
        6: retry

FN #8: ruby-formatter__rufo__a90e654: spec/lib/rufo/formatter_source_specs/retry.rb.spec:6

NOT DETECTED — CODE BUG
The cop fails to detect this pattern. Fix the detection logic.

Message: Invalid retry without rescue (Using Ruby 4.0 parser; configure using TargetRubyVersionparameter, underAllCops)

Ready-made test snippet (add to offense.rb, adjust ^ count):

retry
^ Lint/Syntax: Invalid retry without rescue
(Using Ruby 4.0 parser; configure using `TargetRubyVersion` parameter, under `AllCops`)

Full file context (30 lines before offense):

        1: #~# ORIGINAL retry
        2: 
        3: retry
        4: 
        5: #~# EXPECTED
>>>     6: retry

Omitted 7 additional diagnosed FN example(s) for brevity.

[github-actions]

✅ Cop-check results (8 shards)

Cop	Baseline FP	Baseline FN	Local FP	Local FN	FP Δ	FN Δ	Result
Lint/Syntax	0	14	0	0	0	-14	✅
Lint/Syntax	0	5	0	2	0	-3	✅
Lint/Syntax	0	3	0	0	0	-3	✅
Lint/Syntax	0	0	0	0	0	0	✅
Lint/Syntax	0	2	0	2	0	0	✅
Lint/Syntax	0	2	0	1	0	-1	✅
Lint/Syntax	0	0	0	0	0	0	✅
Lint/Syntax	0	1	0	1	0	0	✅

All shards passed — ready to merge.

[6]

Reviewed: correctly detects invalid UTF-8 files without encoding magic comments. Proper config/filter handling. Resolves 21 of 27 FNs.