type-ruby / t-ruby / 20560974963

92%

Build Type

push

github

Committed by web-flow

Commit Message

refactor: migrate parser from regex to token-based parser combinator (#29)

* refactor: migrate parser from regex to token-based parser combinator

- Replace monolithic parser_combinator.rb (2833 lines) with modular architecture
- Add Scanner for tokenization with regex literal support
- Create IR::InterpolatedString for string interpolation parsing
- Fix type inference for interpolated strings (returns String)
- Add TRuby::ParseError for unified error handling
- Organize parsers into primitives/, combinators/, and token/ directories
- Each file contains exactly one class (snake_case filename matches PascalCase class)

* fix: enhance parser to support ternary, splat args, and statement expressions

- Add ternary operator (? :) parsing in ExpressionParser
- Support double splat (**opts) and single splat (*args) in method calls
- Support keyword arguments (name: value) in method calls
- Allow case/if/unless/begin as assignment right-hand side values
- Improve generic type compatibility (Array[untyped] with Array[T])

Fixes type inference errors in keyword_args samples.

* style: fix RuboCop violations and adjust metrics limits

* fix: require set for Ruby 3.1 compatibility

Run Details

1849 of 2098 new or added lines in 53 files covered. (88.13%)

6 existing lines in 2 files now uncovered.

6644 of 8402 relevant lines covered (79.08%)

908.09 hits per line