Added - Added grako.compile(grammar, name=None, **kwargs) as a substitute for grako.genmodel (which remains for backwards compatibility). - Added a grako.parse(grammar, input, **kwargs) that will compile a grammar and uset to parse the given input. For efficiency, parsed grammars are cached. - Added a grako.to_python_sourcecode(grammar, name=None, filename=None, **kwargs) that compiles the grammar to the Python sourcecode that implements the parser. - Rename the existing *join* (.{}) expression to *gather*, and use *join* for a new expression (%{}) that keeps the separator in the resulting AST. - Added left-join (<{}) and right-join (>{}) expressions to the grammar. Changed - Enable python setup.py test with pytest. - Remove the deprecated prefix= argument to ParseContext.closure(). - Document that *Grako* may be used as a library (with no code generation) by compiling grammars to Grammar objects that can be used to parse any given input, much like Python's re does with regular expressions. - Refactored grammar so double naming (name1:name2:exp) is disallowed. Grako 3.20.0 <https://pypi.python.org/pypi/grako/3.20.0> *Grako* (for *grammar compiler*) is a tool that takes grammars in a variation of EBNF as input, and outputs memoizing (Packrat) PEG parsers in Python. *Grako* can also compile a grammar stored in a string into a grako.grammars.Grammar object that can be used to parse any given input, much like the *re* module does with regular expressions. https://pypi.python.org/pypi/grako/3.20.0 https://bitbucket.org/neogeny/grako/ -- Juancarlo *Añez*