given a directory with YarnSpinner-Unity csv files, convert to godot localization .csv format (python 3)

unity_ys_csv_to_godot.py

import csv
import pathlib
import os
import re


def fix_locale_code(code):
    # https://docs.godotengine.org/en/stable/tutorials/i18n/locales.html
    return code.replace("-", "_")


'''
Read all non-metadata csv files from unity YarnSpinner and convert them to a godot
localization file
'''


def main():
    # adjust paths as necessary for your project
    project_root = pathlib.Path(os.path.abspath(os.path.dirname(__file__)))
    project_root = project_root.joinpath("..", "..", "..").resolve()
    locales_dir = pathlib.Path(project_root).joinpath('core/ui/dialogue/localization/locales/')
    converted_dir = locales_dir  # use a different dir if you want while testing: pathlib.Path(project_root).joinpath("temp")
    converted_dir.mkdir(parents=True, exist_ok=True)

    locale_to_line_list_dict = dict()
    locale_to_header_list_dict = dict()
    for csvfile in locales_dir.glob('*.csv'):
        if 'metadata.csv' in str(csvfile) or 'godot' in str(csvfile):
            print(f"Ignoring file {csvfile}")
            continue

        with open(csvfile, 'r', encoding='utf-8-sig') as locale_csv:
            print(f"Reading YS unity .csv {csvfile}")
            locale_code = None

            locale_csv_reader = csv.reader(locale_csv)
            row_num = 0
            current_headers = None
            for row in locale_csv_reader:
                if row_num == 0:
                    current_headers = row
                    print(f"Found csv headers {row}")
                    row_num += 1
                    continue
                if row_num == 1:
                    locale_code = fix_locale_code(row[0])
                    print(f"Using locale name '{locale_code}'")
                    locale_to_line_list_dict[locale_code] = dict()
                    locale_to_header_list_dict[locale_code] = row
                column = 0
                line = dict()
                line_id = None
                for header_value in row:
                    header = current_headers[column].strip()
                    if header == 'id':
                        line_id = header_value.replace('line:', '')
                        # replace myKey with MY_KEY
                        line_id = re.sub("([a-z])([A-Z])", r"\g<1>_\g<2>", line_id).upper()
                        header = 'keys'
                    line[header] = header_value
                    column += 1
                locale_to_line_list_dict[locale_code][line_id] = line
                print(f"row #{row_num}: {line}")
                row_num += 1

    converted_headers = ['keys']
    converted_headers.extend(locale_to_line_list_dict.keys())
    converted_rows = []
    line_id_to_line = dict()
    locales = locale_to_line_list_dict.keys()
    for locale in locale_to_line_list_dict.keys():
        print(f"Restructuring line dict for locale {locale}")
        for line_id in locale_to_line_list_dict[locale]:
            line = locale_to_line_list_dict[locale][line_id]

            if line_id not in line_id_to_line.keys():
                line_id_to_line[line_id] = dict()
            line_id_to_line[line_id][locale] = line['text']

    # fill in any missing lines for a locale with empty string
    for line_id in line_id_to_line.keys():
        line = line_id_to_line[line_id]
        line['keys'] = line_id
        found_non_blank = False
        for locale in locales:
            if locale not in line:
                line[locale] = ''
            if len(line[locale]) > 0:
                found_non_blank = True
        if found_non_blank:
            converted_rows.append(line)

    with open(locales_dir.joinpath('godot_locales.csv'), 'w', encoding='utf-8') as converted_csv:
        # make sure 'keys' is the first column
        fieldnames = ['keys']
        locales = [*converted_rows[0].keys()]
        locales.remove('keys')
        fieldnames.extend(locales)
        writer = csv.DictWriter(converted_csv, fieldnames=fieldnames, lineterminator='\n')

        writer.writeheader()
        writer.writerows(converted_rows)


if __name__ == "__main__":
    main()