more edge values #14

youknowone · 2026-01-13T13:27:42Z

Summary by CodeRabbit

Tests
- Expanded integer operation tests with more edge and boundary values
- Increased floating-point test dataset with many additional edge cases
Improvements
- Improved numerical stability and overflow handling in complex math operations
- Adjusted complex magnitude and polar calculations for more robust behavior
Chores
- Updated project ignore rules (added "cpython")

_{✏️ Tip: You can customize this high-level summary in your review settings.}

coderabbitai · 2026-01-13T13:27:53Z

📝 Walkthrough

Walkthrough

Expanded test edge-case datasets for integers and floats, adjusted complex-math magnitude and logarithm computations for numeric stability, and added "cpython" to .gitignore. No public APIs were changed.

Changes

Cohort / File(s)	Summary
Configuration `\.gitignore`	Added `cpython` to ignore rules.
Test data — integers & floats `src/math/integer.rs`, `src/test.rs`	Enlarged EDGE_I64 from a fixed-size array to a slice and added many boundary values; expanded `EDGE_VALUES` from 30 to 64 f64 entries (subnormals, signaling NaN, trig-related and larger magnitudes). Adjusted test iteration accordingly.
Complex math — exponential `src/cmath/exponential.rs`	Introduced `mul_add` import and refactored ln(z) path to precompute a `log1p_arg` using `mul_add`, then use `log1p`/2.0 for near-unit magnitudes.
Complex math — magnitude helpers `src/cmath/misc.rs`	Added private helpers `c_abs_raw` and `c_abs_checked`; `abs` now uses `c_abs_raw`, and `polar` uses `c_abs_checked`, centralizing finite checks and overflow handling.

Estimated code review effort

🎯 3 (Moderate) | ⏱️ ~20 minutes

Poem

🐇 I nibble at edges, keen and neat,

Added numbers, tests, and a tiny treat.
cpython hides, magnitudes refined,
Logs tamed with mul_add, neatly aligned.
Hooray — more signals for this curious mind!

🚥 Pre-merge checks | ✅ 3

✅ Passed checks (3 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title 'more edge values' directly aligns with the primary changes: expanding edge case test datasets in EDGE_I64, EDGE_VALUES, and adjusting test iteration patterns.
Docstring Coverage	✅ Passed	Docstring coverage is 100.00% which is sufficient. The required threshold is 80.00%.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing touches

📝 Generate docstrings

🧹 Recent nitpick comments

src/cmath/misc.rs (1)
47-59: Consider removing the unnecessary set_errno(0) call.

The overflow detection relies on r.is_infinite() rather than checking errno, making the set_errno(0) on line 52 unnecessary. While valid per coding guidelines, it adds overhead without benefit.
♻️ Suggested simplification
 #[inline]
 fn c_abs_checked(z: Complex64) -> Result<f64> {
     if !z.re.is_finite() || !z.im.is_finite() {
         return Ok(c_abs_raw(z));
     }
-    crate::err::set_errno(0);
     let r = m::hypot(z.re, z.im);
     if r.is_infinite() {
         Err(Error::ERANGE)
     } else {
         Ok(r)
     }
 }

📜 Recent review details

Configuration used: defaults

Review profile: CHILL

Plan: Pro

📥 Commits

Reviewing files that changed from the base of the PR and between 8313a32 and 0c83ff1.

📒 Files selected for processing (3)

src/cmath/exponential.rs
src/cmath/misc.rs
src/math/integer.rs

🧰 Additional context used

📓 Path-based instructions (3)

src/**/*.rs

📄 CodeRabbit inference engine (AGENTS.md)

Error handling: Return Result directly with Err(crate::Error::EDOM) or Err(crate::Error::ERANGE) instead of using set_errno. Only valid set_errno call is set_errno(0) to clear errno before libm calls.

Files:

src/cmath/misc.rs
src/cmath/exponential.rs
src/math/integer.rs

src/cmath/**/*.rs

📄 CodeRabbit inference engine (AGENTS.md)

src/cmath/**/*.rs: For complex math (cmath module), copy special_type() enum and function, 7x7 special value tables (e.g., tanh_special_values), SPECIAL_VALUE macro usage, and complex-specific error handling from Modules/cmathmodule.c.
Use mul_add in cmath functions for expressions like 1.0 + x * x, cross-product calculations (mul_add(s1.re, s2.im, -(s2.re * s1.im))), and squared terms to achieve bit-exact matching with CPython.

Files:

src/cmath/misc.rs
src/cmath/exponential.rs

src/math/**/*.rs

📄 CodeRabbit inference engine (AGENTS.md)

src/math/**/*.rs: Use existing CPython helper equivalents: FUNC1 → math_1(x, func, can_overflow), FUNC2 → math_2(x, y, func), m_log/m_log10/m_log2 in exponential.rs. If CPython uses a helper, use the corresponding Rust helper.
Use crate::mul_add(a, b, c) instead of a * b + c for expressions of the form a * b + c or c + a * b to match CPython's FMA (fused multiply-add) behavior on platforms like macOS where it's enabled.

Files:

src/math/integer.rs

🧠 Learnings (11)

📓 Common learnings

Learnt from: CR
Repo: RustPython/pymath PR: 0
File: AGENTS.md:0-0
Timestamp: 2026-01-10T13:40:01.994Z
Learning: Applies to src/cmath/**/*.rs : Use mul_add in cmath functions for expressions like 1.0 + x * x, cross-product calculations (mul_add(s1.re, s2.im, -(s2.re * s1.im))), and squared terms to achieve bit-exact matching with CPython.

Learnt from: CR
Repo: RustPython/pymath PR: 0
File: AGENTS.md:0-0
Timestamp: 2026-01-10T13:40:01.994Z
Learning: Applies to src/**/*test*.rs : All float functions must be tested with crate::test::EDGE_VALUES which includes zeros, infinities, NaNs with different payloads, subnormals, boundary values (MIN_POSITIVE, MAX, MIN), large values near infinity, and trigonometric special values (PI, PI/2, PI/4, TAU).

Learnt from: CR
Repo: RustPython/pymath PR: 0
File: AGENTS.md:0-0
Timestamp: 2026-01-10T13:40:01.994Z
Learning: Applies to src/cmath/**/*.rs : For complex math (cmath module), copy special_type() enum and function, 7x7 special value tables (e.g., tanh_special_values), SPECIAL_VALUE macro usage, and complex-specific error handling from Modules/cmathmodule.c.