v1.1.6

Performance

• Short-circuits invalid options with a nullish fast path to reduce runtime overhead when no config is passed
  (objectEquals(target, source) now avoids unnecessary optional chaining checks).

Types

• Refined objectEqualsCore and objectEquals signatures in .d.mts:
  • Replaced any with unknown for safer inputs.
  • Defaulted all boolean flags to false in type annotations for clarity.
  • Strengthened cache to Map<object, object> | WeakMap<object, object> in ObjectEqualsOptions interface and objectEqualsCore function.
• Explicitly exported ObjectEqualsOptions interface.

Documentation

• Updated README.md usage section to reflect new unknown parameter types instead of any.
• Clarified that V8 and JSC paths apply to runtimes.

v1.1.5

Fixed

• Fixed false positives for inline function props in React mode by comparing function bodies using .toString().
• Prevented RangeError in TypedArray comparison by validating byte alignment before using Int32Array views.
• Fallback to tag-based type check when constructors mismatch and may be shadowed.
• Corrected null equality shortcut to use strict comparison for semantic clarity.

Improved

• Reduced stack slot usage and instruction count in Set comparison by inlining .length checks.
• Renamed internal isTypedArray to isBufferView for clarity.

Refactored

• Removed browser-specific checks from runtime build.

Tests

• Added test case to verify fallback behavior for unaligned TypedArray offsets.

Documentation

• Added PayPal donation link via .github/FUNDING.yml to enable GitHub's Sponsor button.
• Added Support section and donation badge to README.md.

v1.1.4

Added

• Added dedicated benchmarks to evaluate deep equality performance for ArrayBuffer
• Added a test for asymmetric mutual circular references.

Documentation

• Included benchmark results for ArrayBuffer
• Clarified web-safe note and explained build output structure

v1.1.3

Fixed

• Added a null guard for recursive equality checks in Set comparison (cosmetic, non-functional change).
• Corrected globalIndex setting in Shuffled nested Set benchmarks to prevent early exit matches and ensure realistic evaluation.

Added

• Added node.isDeepStrictEqual to Shuffled nested Set group in advanced benchmarks.
• Added a note to the React benchmark to clarify the role of the react option and benchmark fairness.

Updated

• Refreshed benchmark results in README for Shuffled nested Set with corrected generator config and additional comparison.

v1.1.2

Benchmarks

• Fixed incorrect import of the web-safe (default) variant in benchmark suites.
• All runtime benchmarks (basic, react, and advanced) now use the correct Node-optimized build (.node.esm.js) to properly trigger runtime-specific optimizations like Buffer.compare, Bun.deepEquals and util.isDeepStrictEqual.
• Runtime benchmarks now reflect real-world performance more accurately, aligning with README.md results.

Documentation

• Fixed incorrect rendering of > [!NOTE] blocks on npm by appending proper spacing.
• Minor consistency fixes for better clarity in markdown.

v1.1.1

Removed

• Removed jsr.json configuration and badge due to current platform limitations.
• object-equals remains fully compatible with Deno via the npm registry, including runtime-specific optimizations.

Notes

• JSR currently lacks proper support for dual exports (node vs default), ESM-only packages and .d.ts mappings without TypeScript source.
• Package remains platform-agnostic and ready for modern runtimes (Node.js, Bun, Deno) via npm.

v1.1.0

Highlights

• Web implementation is now the default export (object-equals.esm.js).
• Node-specific optimizations (e.g. Buffer.compare, util.isDeepStrictEqual) moved to explicit subpath (object-equals.node.esm.js).
• Added support for JSR.

Build

• Make browser-safe version the default export
• Moved Node-optimized build to explicit node subpath in exports
• Added jsr.json for publishing via jsr.io

Benchmarks

• Added runtime-specific benchmark groups for TypedArray and DataView

Documentation

• Refactor and expand README.md with:
  • New feature descriptions
  • Separated benchmark results (web-safe vs runtime-specific) for TypedArray and DataView
  • Cleaned up badges and removed outdated sections

v1.0.3

Refactor

• Replaced assert.deepStrictEqual with util.isDeepStrictEqual, improving performance and eliminating the need for a custom wrapper.
• Cleaned up redundant wrapper logic across core comparison functions.

Benchmarks

• Refactored benchmark suite structure for better scalability and maintainability.

Tests

• Simplified DataView comparison test.
• Clarified React element test to verify behavior with inline function props.

Documentation

• Updated README.md with the latest benchmark results.

Fixes

• Prevented unnecessary import of node:util in Bun runtime to reduce overhead.

v1.0.2

Changed

• Migrated test framework from Jest to Vitest for native ESM support.
• Added compatibility test suite based on lodash.isEqual QUnit test cases.
• Updated README.md with new Build and updated Testing section.

Fixed

• Resolved asymmetric equality bug in circular reference mode.
• Corrected handling of objects with a shadowed constructor property.
• Added sideEffects: false flag to package.json for proper tree-shaking in ESM-aware bundlers.

Internal

• Removed all Jest-related dependencies and configuration.
• Introduced vitest.config.js for clean test management.

v1.0.1

v1.0.1 - minor update

• Added guard against null in constructor comparison
• Added regression tests for null comparison
• Improved engine and runtime detection
• Added browser field to package.json

Thanks for early community feedback and stars!