grammar_parser.GrammarParser¶

class tyrian.typarser.grammar_parser.GrammarParser(raw_grammar: dict=None, token_defs: dict=None, grammar_mapping: dict=None, settings: dict=None)¶

Does the grunt work of parsing the Grammar into a usable object; see grammar_nodes for more

Parameters:	raw_grammar – single string containing raw grammar definitions, see load_grammar token_defs – dictionary of token definitions, see load_token_definitions

load_grammar_mapping(nodes)¶

Load in a mapping between grammars and Nodes

Supply a dictionary with a mapping between subgrammar names and Node objects

Parameters:	grammar_mapping – dictionary mapping subgrammars to appropriate Nodes

load_grammar(content:str)¶

Load grammars from a string. All grammars need not be necessarily be loaded at once, but all must be loaded before parse_grammars() is called.

Parameters:	content – single string containing raw grammar definitions

a grammar can be defined like so:
name: <content>;

whereby within the following constructs are permissible;

OR, which can be nested, is denoted by a pipe character:
<token> | <token>

many of a particular token:
<token>+

a subgrammar or token is simply specified by name;
NAME

load_token_definitions(defs: dict)¶

Loads token definitions.

expected to be formatted as follows;

Parameters:	defs – dictionary containing token definitions

{
    'literal': {
        '<content>': '<name>',
        ...
    },
    'regex': {
        '<regex_expr>': '<name>',
        ...
    }
}

parse_grammars()¶

Parses loaded grammars into “check trees”.

These “check trees” consist of a root ContainerNode, where a list of tokens can be passed into the root GrammarNode‘s check() function and validated according to the loaded grammars.

parse_grammar(grammar: str, grammar_key: str, settings: dict)¶

See parse_grammars()

Parameters:	grammar – single string containing a single raw grammar definition, see parse_grammars grammar_key – key for grammar, aka name of grammar settings – dictionary of settings for Nodes