[standalone.py] NameError: name 'RULE' is not defined

I am using:

pip3 show lark-parser
Name: lark-parser
Version: 0.6.3

Python 3.6.4

When I generate the standalone parser and try to run it, it throws:

Traceback (most recent call last):
  File "lexer.py", line 601, in <module>
    1: Rule(NonTerminal('language_syntax'), [NonTerminal(Token(RULE, '__anon_star_0')), NonTerminal('preamble_statements')], None, RuleOptions(False, False, None)),
NameError: name 'RULE' is not defined

The generated parse is this: lexer.zip

The line 601 is this:

File: lexer.py
599: RULES = {
600:   0: Rule(NonTerminal('language_syntax'), [NonTerminal('preamble_statements')], None, RuleOptions(False, False, None)),
601:   1: Rule(NonTerminal('language_syntax'), [NonTerminal(Token(RULE, '__anon_star_0')), NonTerminal('preamble_statements')], None, RuleOptions(False, False, None)),

The problem is this RULE constant on the python code which is not defined anywhere by the standalone.py tool.

This is the input grammar used:

language_syntax: _NEWLINE* preamble_statements
preamble_statements: target_language_name_statement | master_scope_name_statement

target_language_name_statement: "name" ":" free_input_string
master_scope_name_statement: "scope" ":" free_input_string

// General terminal constructs
CR: "/r"
LF: "/n"
SPACES: /[\t \f]+/

// General non terminal constructs
free_input_string: /[^\n]/
SINGLE_LINE_COMMENT: /(#|\/\/)[^\n]*/

// _NEWLINE: (CR? LF)+
_NEWLINE: ( /\r?\n[\t ]*/ | SINGLE_LINE_COMMENT )+
INTEGER: "0".."9"

%ignore SPACES
%ignore SINGLE_LINE_COMMENT

This was the command line used to build the standalone parser:

python -m lark.tools.standalone ../simple.lark > lexer.py

[erezsh]

Pushed some code to master that should fix this. Let me know if it works (or if it doesn’t)

[erezsh]

Sorry for not replying sooner. The reason that the generated parser doesn’t work, is that right now it only supports the “standard” lexer, while your grammar requires the “contextual” lexer.

It shouldn’t be too difficult to add, but it will take a few hours of work. Hopefully I’ll get to it soon.