Release Notes: This release adds UTF-16BE and UTF-16LE support. The parse-tree can be used for debugging Ruby's internal Regexp class.
Release Notes: Most of the parser has been rewritten in order to deal with multiple encodings. UTF-8 support has been tested and implemented. For full compatibility with Ruby's native Regexp engine, the String class has been extended with: #sub, #gsub, #match, #scan, #split. The syntax has been extended with octal, hex, widehex, and atomic grouping. A problem with word-boundary anchors has been fixed.
Release Notes: This release added lookbehind positive/negative of variable width. Other engines usually only supports fixed width, so this is a major achievement.
Release Notes: The allocation model was refactored so that the Context class now has a role of a caretaker (Memento pattern). This has paved the way for more advanced patterns where lookahead is being used inside of repeats. Now 96.66% of the 1560 rubicon tests passes OK. This is an improvement of 0.66% since the last release.