Newline regex considered as zero length

Describe the bug I’m using the Earley parser for parsing with lark. When using a regular expression consisting of a single newline \n (or together with a lookahead in the regex) - e.g. /\n(?=a)/x - lark complains that the regular expression is of zero length - which it is not. Also, the usage of other flags mliux does not change this behaviour.

To Reproduce

import re
from lark import Lark
GRAMMAR = """start: /\n(?=a)/x"""
parser = Lark(GRAMMAR, ambiguity="resolve", debug=True)
print(parser.parse("\na").pretty())

Error Message

Traceback (most recent call last):                                                                                
  File "min_work_ex.py", line 8, in <module>                                                                      
    parser = Lark(GRAMMAR, ambiguity="resolve", debug=True)                                                       
  File "/home/lol/.local/lib/python3.8/site-packages/lark/lark.py", line 377, in __init__                      
    self.parser = self._build_parser()                                                                            
  File "/home/lol/.local/lib/python3.8/site-packages/lark/lark.py", line 419, in _build_parser                 
    return parser_class(self.lexer_conf, parser_conf, options=self.options)                                       
  File "/home/lol/.local/lib/python3.8/site-packages/lark/parser_frontends.py", line 40, in __call__           
    return ParsingFrontend(lexer_conf, parser_conf, options)                                                      
  File "/home/lol/.local/lib/python3.8/site-packages/lark/parser_frontends.py", line 69, in __init__           
    self.parser = create_parser(lexer_conf, parser_conf, options)                                                 
  File "/home/lol/.local/lib/python3.8/site-packages/lark/parser_frontends.py", line 216, in create_earley_pars
er                                                                                                                
    return f(lexer_conf, parser_conf, options, resolve_ambiguity=resolve_ambiguity, debug=debug, tree_class=tree_c
lass, **extra)                                                                                                    
  File "/home/lol/.local/lib/python3.8/site-packages/lark/parser_frontends.py", line 193, in create_earley_pars
er__dynamic                                                                                                       
    earley_matcher = EarleyRegexpMatcher(lexer_conf)                                                              
  File "/home/lol/.local/lib/python3.8/site-packages/lark/parser_frontends.py", line 182, in __init__          
    raise GrammarError("Dynamic Earley doesn't allow zero-width regexps", t)                                      
lark.exceptions.GrammarError: ("Dynamic Earley doesn't allow zero-width regexps", TerminalDef('__ANON_0', '(?x:\n(
?=a))'))                                                                                                          

Lark version lark-parser==0.11.3

Question Is this in any way intended? Or is there a way around this problem? Thanks in advance for your response 😃

[MegaIng]

The regex is Zero length. You have the x flag, which means that all Whitespace characters are ignored. The escape sequence \n will get translate to a newline before being passed to the regex engine. That means that it is ignored. There are two fixes: Don’t use the x flag, or add another backslash: \\n.

[erezsh]

I don’t think so, but you’re welcome to make the argument. Preferably in a separate issue.