Guess a delimiter being used to split a string using term frequencies of non-word characters.

guess_string_separator.py

import re
from collections import Counter

def guess_string_separator(
    string: str,
    to_ignore: Iterable[str] = tuple(),
    ignore_empty_strings: bool = True,
) -> Optional[str]:
    """Guess a delimiter being used to split a string using term frequencies.

    E.g. "mike, john, bob" -> ", "
    """
    non_word_parts = re.split(r"\w+", string)
    if ignore_empty_strings:
        # get rid of spaces and empty string
        non_word_parts = (part for part in non_word_parts if part.strip())
    if to_ignore:
        non_word_parts = (part for part in non_word_parts if part not in to_ignore)
    term_frequencies = Counter(non_word_parts)
    most_common = [
        (term, count) for (term, count) in term_frequencies.most_common() if count > 1
    ]
    if most_common:
        return most_common[0][0]
    return None