Feature Release

• Support for unicode grammars where literals as well as idents can be unicode (e.g. Algol uses operators such as ÷, ≠, ≤, ⊃ which can only be represented as unicode chars)
• All file formats in use by EbnfStudio are now utf-8
• Similar to the .keywords file there is now a .tokmap file to define token names for chars or char sequences (format: definition per line, char whitespace name per definition)
• Coco code generation of LL:x prefixes now can handle nullable symbols
• fixes
• New syntax to manually specify LL prefixes is work in progress

Feature Release

• New pragma "%keywords" to dynamically define keywords directly within the ebnf file.
• New operator "+=" (in addition to "::=") to add values to a (potentially inexisting) pragma.
• Preprocessor operations #define, #undef, #ifdef, #ifndef, #else and #endif to handle versions of the grammar in the same file.
• Symbols defined with "#define" appear as C++ defines in the generated TokenType.h.

Features and Fixes

pragmas %module %namespace %suppress
generate terminal lexer; tokenTypeFromString; also accept nt in pragmas
fixes: find, mark nt idents, sort, etc.; features: check reachability, transform IEEE EBNF

v0.6.1: Initial Release

analyzer findpath; pathview