#15 feat: add PERF017 rule detecting correlated subqueries

[RAprogramm]

Closes #15

New performance rule PERF017 (Warning): flags subqueries that qualify columns with a source not declared inside the subquery body — i.e. references to the outer query, which force per-row re-execution. Uncorrelated IN/scalar subqueries are not flagged; string literals are masked so dots in values (a@b.com) never look like qualified references.

• README, Cargo.toml description, docs updated to 35 rules; version bumped to 0.16.0
• 4 new tests: EXISTS and scalar-projection correlation, uncorrelated negative, literal-dot negative

Local checks: fmt, clippy -D warnings, tests, cargo qual (0 issues), mdbook build — all green.

[github-actions]

Rust Diff Analysis

Tip

PR size is within limits. Good job keeping changes focused!

Limits — configured thresholds for this repository

Each metric is compared against its configured maximum. If any limit is exceeded, the PR check fails.

Metric	Value	Limit	Status
Production Units	9	30	✅
Weighted Score	14	100	✅
Lines Added	143	200	✅

Understanding the metrics:

• Production Units: Functions, structs, enums, traits, and other semantic code units in production code
• Weighted Score: Complexity score based on unit types (public APIs weigh more than private)
• Lines Added: Raw count of new lines in production code

Summary — breakdown of changes by category

Production code counts toward limits. Test code is tracked but doesn't affect limits.

Metric	Production	Test
Functions	7	-
Structs/Enums	1	-
Other	1	-
Lines added	+143	+27
Lines removed	-0	-0
Total units	9	4

Production Changes — 9 units modified

Semantic units (functions, structs, etc.) that were added or modified in production code.

File	Unit	Type	Changes
src/rules.rs:182-244	RuleRunner::with_config	function	+1 -0
src/rules/performance.rs:805-810	CorrelatedSubquery	struct	+6 -0
src/rules/performance.rs:812-839	subquery_bodies	function	+28 -0
src/rules/performance.rs:841-868	body_sources	function	+28 -0
src/rules/performance.rs:870-884	mask_string_literals	function	+15 -0
src/rules/performance.rs:886-915	references_outer_source	function	+30 -0
src/rules/performance.rs:917-951	Rule for CorrelatedSubquery	impl	+3 -0
src/rules/performance.rs:918-925	Rule for CorrelatedSubquery::info	function	+8 -0
src/rules/performance.rs:927-950	Rule for CorrelatedSubquery::check	function	+24 -0

Test Changes — 4 units modified

Test code changes don't count toward PR size limits.

File	Unit	Type	Changes
tests/rules_tests.rs:184-190	test_correlated_exists_flagged	function	+7 -0
tests/rules_tests.rs:192-198	test_correlated_scalar_projection_flagged	function	+7 -0
tests/rules_tests.rs:200-205	test_uncorrelated_in_subquery_ok	function	+6 -0
tests/rules_tests.rs:207-213	test_literal_dot_not_correlated	function	+7 -0

Analysis Scope

Analyzed: 3 Rust files

Skipped files:

• 6 non-Rust files

Skipped file list:

• Cargo.lock (non-Rust)
• Cargo.toml (non-Rust)
• README.md (non-Rust)
• docs/src/introduction.md (non-Rust)
• docs/src/rules/index.md (non-Rust)
• docs/src/rules/performance.md (non-Rust)

---

_{Rust Diff Analyzer}

[github-actions]

Cargo Quality Analysis

Tip

All quality checks passed. Code meets all standards!

Results — issues found by each analyzer

Each analyzer checks for specific code patterns. Zero issues means the check passed.

Analyzer	Issues	Status
path_import	0	✅
format_args	0	✅
empty_lines	0	✅
inline_comments	0	✅
mod_rs	0	✅
Total	0

Analyzers — what each check does

These analyzers enforce consistent code style and best practices.

Analyzer	Description	How to fix
path_import	Detects inline path imports like std::path::Path	Use use statements at the top
format_args	Detects format!("{}", x) patterns	Use format!("{x}") syntax
empty_lines	Detects empty lines inside function bodies	Remove extra blank lines
inline_comments	Detects // comment inside functions	Move to doc comments or remove
mod_rs	Detects mod.rs files (use modern module style)	Rename foo/mod.rs to foo.rs

Raw Output — detailed analysis log

Total issues: 0
Fixable: 0

---

_{Cargo Quality}

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ All tests successful. No failed tests found.

📢 Thoughts on this report? Let us know!