Generates a JSON file of all unicode characters, their name, and their aliases. This is a hack but it works.

my_unicodedata_db.py

"""
Tool that names each unicode character based on their name or their aliases

Must be run from the cpython repo root directory:
https://github.com/python/cpython/tree/3.10

Relies on Tools.unicode being an importable path due to implicit packages
"""
from contextlib import redirect_stderr, redirect_stdout
from io import StringIO
import json
from pathlib import Path
from unicodedata import unidata_version
from Tools.unicode.makeunicodedata import UnicodeData, makeunicodename


def get_characters():
    unicode_data_path = Path(__file__).parent / "unicode_data.json"
    if unicode_data_path.exists():
        # If unicode_data.json path exists, load data from there
        with unicode_data_path.open(mode="rt") as f:
            characters = json.load(f)
    else:
        # Use CPython's makeunicodedata tool which downloads needed data
        with redirect_stdout(StringIO()), redirect_stderr(StringIO()):
            unicode = UnicodeData(unidata_version)
            makeunicodename(unicode, 1)
        characters = {
            chr(int(r.codepoint, 16)): {"name": r.name, "aliases": []}
            for r in unicode.table
            if r
        }
        for alias, codepoint in unicode.aliases:
            characters[chr(codepoint)]["aliases"].append(alias)
        with unicode_data_path.open(mode="wt") as f:
            json.dump(characters, f)
    return characters


characters = get_characters()


def name(character):
    r"""Like unicodedata.name, but works for unnamed characters like \n."""
    record = characters[character]
    if record["name"].startswith("<"):  # If no name, use the first alias
        return record["aliases"][0]
    return record["name"]