Get more frequent french city "name" as mapcustomizer.com visualizable data

stats.py

#!/usr/bin/env python3

# Based on https://public.opendatasoft.com/explore/dataset/correspondance-code-insee-code-postal/download/?format=json&timezone=Europe/Berlin&lang=fr
# The JSON file must be argv[1]

import json
import sys

class City():
    token_stats = {}

    @staticmethod
    def _sanitize_token(token):
        if token[:2] in ["D'", "L'"]:
            return token.replace("D'", "").replace("L'", "")
        return token

    @staticmethod
    def _trash_bad_token(token):
        trash = [
            "ARRONDISSEMENT",
            "AU",
            "AUX",
            "DE",
            "DES",
            "DU",
            "EN",
            "ET",
            "LA",
            "LE",
            "LES",
            "SAINT",
            "SAINTE",
            "SAINTES",
            "SOUS",
            "SUR",
            ]
        if token in trash:
            return None
        return token

    @staticmethod
    def _account_token(token):
        try:
            City.token_stats[token] += 1
        except:
            City.token_stats[token] = 1
        return token

    @staticmethod
    def _name_to_token(name, hooks):
        token_1 = name.split("-")
        token = []
        for tok_1 in token_1:
            token_2 = tok_1.split(" ")
            for tok_2 in token_2:
                for hook in hooks:
                    if tok_2 is None:
                        break
                    tok_2 = hook(tok_2)
                if tok_2 is not None:
                    token.append(tok_2)
        return token

    def __init__(self, name, coords):
        self.name = name
        self.token = City._name_to_token(name, [City._trash_bad_token, City._sanitize_token, City._account_token])
        self.coords = coords

    def __repr__(self):
        return f"{self.coords[0]},{self.coords[1]} {{{self.name}}}"

with open(sys.argv[1]) as f:
    data = json.loads(f.read())

cities = []
for ville in data:
    city = City(ville["nom_comm"], ville["geo_point_2d"])
    cities.append(city)

SORT = {k: v for k, v in sorted(City.token_stats.items(), key=lambda item: item[1], reverse=True)}
for (k, v), _ in zip(SORT.items(), range(15)):
    with open(f"{v}-{k}", "w") as f:
        for city in cities:
            if k in city.token:
                f.write(repr(city))
                f.write("\n")