Access Unicode Script property in Python & find out which script(s) a string contains

unicode_scripts.py

#!/usr/bin/env python3

# Unicode characters are neatly categorized into different "scripts", as seen on
# the character code chart <http://www.unicode.org/charts/#scripts> and defined
# in Annex #24 <https://www.unicode.org/reports/tr24/>.
#
# Unfortunately, Python's unicodedata module doesn't provide access to this
# information. However, the fontTools library does include this.
# <https://github.com/fonttools/fonttools>
#
# Here's an example how to use it.
#
# Requires:
# - fonttools (tested with version 4.29.1)
# - rich (optional; just comment out the import)

from collections import Counter
from fontTools.unicodedata import script, script_name
from rich import print


def detect_scripts(text):
    """Returns a Counter mapping scripts to how often characters from that
    script appear in `text`.
    """
    characters = Counter(text)
    scripts = Counter()
    for char, count in characters.items():
        scripts[script(char)] += count
    return scripts


def detect_script_ratios(text):
    scripts = detect_scripts(text)
    total = sum(scripts.values())
    return {script_name(s, default="Unknown"): c / total for s, c in scripts.items()}


def get_majority_script(text):
    scripts = detect_scripts(text)
    for s, _ in scripts.most_common(3):
        name = script_name(s, default=None)
        if name not in ("Common", None):
            return name
    return None


if __name__ == "__main__":
    # From uk.wikipedia.org
    sample = """21 квітня 2021 р. Pfizer заявляє, що їй відомо про підроблені версіях своєї
                вакцини COVID-19, які виробник ліків розробив за допомогою BioNTech, оскільки
                злочинці прагнуть нажитися на світовому попиті на вакцини, який продовжує
                випереджати пропозицію."""
    print(detect_scripts(sample))
    # Counter({'Zyyy': 97, 'Cyrl': 183, 'Latn': 19})
    print(detect_script_ratios(sample))
    # {'Common': 0.32441471571906355, 'Cyrillic': 0.6120401337792643, 'Latin': 0.06354515050167224}
    print(get_majority_script(sample))
    # Cyrillic