- Add new rule `prefer-logical-properties` that enforces logical CSS properties over physical directional properties
- Detects 140+ physical property mappings across margin, padding, border, inset, size, overflow, and scroll properties
- Supports value-based detection for `text-align`, `float`, `clear`, and `resize` properties
- Provides automatic fixes for all detected violations
- Preserves original formatting (camelCase/kebab-case and quote style)
- Configurable allowlist via `allow` option to skip specific properties
- Comprehensive test coverage
- Add new rule `no-px-unit` that disallows `px` units in vanilla-extract styles with an allowlist option
- Provides fix suggestions for string literals and simple template literals (no expressions)
- New rule that flags and fixes unnecessary trailing zeros in numeric values
- Handles various CSS units, negative numbers, and decimal values
- Preserves non-trailing zeros in numbers like 11.01rem and 2.05em
- Includes comprehensive test coverage for edge cases
- add reference tracking for wrapper functions in vanilla-extract style objects
- implement ReferenceTracker class for detecting vanilla-extract imports
- add createReferenceBasedNodeVisitors for automatic function detection
- support wrapper functions with parameter mapping enable all lint rules to work with custom wrapper functions
This commit introduces robust reference tracking and wrapper function support, enabling all lint rules to work seamlessly with custom vanilla-extract style patterns while preserving compatibility with existing usage and improving rule extensibility.
Adds a rule to disallow unknown or invalid CSS units in vanilla-extract style objects.
- Reports any usage of unrecognized units in property values
- Handles all vanilla-extract APIs (style, styleVariants, recipe, etc.)
- Ignores valid units in special contexts (e.g., CSS functions, custom properties)
No autofix is provided because replacing or removing unknown units may result in unintended or invalid CSS. Manual developer review is required to ensure correctness.
This rule enforces unitless zero values in vanilla-extract style objects:
- Automatically removes unnecessary units from zero values
- Handles both positive and negative zero values
- Preserves units where required (time properties, CSS functions)
- Works with all vanilla-extract APIs
Add comprehensive rule to detect and prevent empty CSS style blocks:
- Identify style objects with no properties
- Flag empty style blocks as potential code quality issues
- Provide auto-fix capability to remove empty blocks
- Handle edge cases like comments-only blocks
This rule helps maintain cleaner codebases by eliminating empty style definitions that often result from incomplete refactoring or forgotten implementations, reducing confusion and unnecessary code.
- Fix handling of missing groupOrder configuration
- Refactor negative conditions to positive ones with optional chaining
- Add comprehensive tests to achieve total coverage
Add comprehensive tests for shared utility modules to improve code coverage:
- Test property name extraction edge cases
- Test CSS property priority map with invalid groups
- Test order strategy visitor creator edge cases
- Test font face property order enforcer early returns
- Test style node processor with arrays and null values
These tests ensure all code paths in shared utilities are properly exercised,
including error handling and edge cases.
Add tests for all three CSS property ordering rules:
alphabetical-order,
concentric-order,
custom-order,
Tests cover all implemented vanilla-extract APIs, fontFace, globalFontFace, globalKeyframes, globalStyle, keyframes, style, and styleVariants.. Each test verifies both valid and invalid cases, along with proper auto-fixing functionality.
- Ensure 'src' property always appears first
- Sort remaining properties alphabetically
- Handle both APIs correctly despite different argument structures
- Handles font faces ordering the same in all 3 available CSS rules
- Update documentation with fontFace ordering details
