def def_fsm_char_pattern(cls, ch, regex, *args):
"""Register new FSM character pattern within tokenizer.
:Parameters:
ch : str
the FSM character
regex
regular expression matching all characters from the class,
it may be either string describing regular expression or
a regular expression object compiled with ``re.compile()``.
See documentation of
`ezmlex.patterns.def_pattern()` for details.
:Note:
FSM character patterns must not overlap. If one character class
within tokenizer matches given item, no other pattern in the
same tokenizer may match this item.
:Note:
If it doesn't exist, this method will create the dictionary
``_fsm_char_patterns`` within ``cls`` subclass (as a class
attribute).
:Note:
This method may throw exceptions in case the `ch` was already
registered. For exception specification, see documentation of
`ezmlex.patterns.def_pattern()`.
"""
from ezmlex.patterns import def_pattern, ccat
if ch == '\0':
raise RuntimeError('FSM char %s is reserved' % ch)
def_pattern(cls.fsm_char_patterns(), ch, ccat('^', regex, '$'), *args)
def def_token_type(cls, _id, pattern, *args):
from ezmlex.tokens import def_token
from ezmlex.patterns import ccat
def_token(cls.token_types(), _id, ccat(pattern), *args)