generate_strings.py

#!/usr/bin/env python3
# -*- coding: utf-8 -*-

import sys
import yaml
import argparse
from pathlib import Path

# The keys in the YAML file whose string values should be translated.
TRANSLATABLE_KEYS = {"title", "subtitle", "label"}

HEADER = """# -*- coding: utf-8 -*-
#
# DO NOT EDIT THIS FILE MANUALLY.
#
# This file is automatically generated by the 'generate_strings.py' script
# from the content of a YAML file. It is used by gettext to extract
# translatable strings from data files.
#
import gettext
_ = gettext.gettext

# --- Strings from data file ---
"""


def extract_strings_from_data(data):
    """
    Recursively finds all strings from the data structure
    that have a key present in TRANSLATABLE_KEYS.
    """
    strings = set()
    if isinstance(data, dict):
        for key, value in data.items():
            if key in TRANSLATABLE_KEYS and isinstance(value, str):
                strings.add(value)
            else:
                strings.update(extract_strings_from_data(value))
    elif isinstance(data, list):
        for item in data:
            strings.update(extract_strings_from_data(item))
    return strings


def main():
    """Main function to generate the translatable strings file."""
    parser = argparse.ArgumentParser(
        description="Extract translatable strings from a YAML file into a Python file for gettext."
    )
    parser.add_argument("input_yaml", type=Path, help="Path to the input YAML file.")
    parser.add_argument("output_py", type=Path, help="Path for the output Python file.")
    args = parser.parse_args()

    input_file: Path = args.input_yaml
    output_file: Path = args.output_py

    if not input_file.is_file():
        print(f"Error: Input file not found at {input_file}")
        sys.exit(1)

    print(f"Reading strings from: {input_file}")
    with open(input_file, "r", encoding="utf-8") as f:
        try:
            data = yaml.safe_load(f)
        except yaml.YAMLError as e:
            print(f"Error parsing YAML file: {e}")
            sys.exit(1)

    translatable_strings = extract_strings_from_data(data)

    if not translatable_strings:
        print("Warning: No translatable strings found.")
        # Still create the file so the build process doesn't fail
        with open(output_file, "w", encoding="utf-8") as f:
            f.write(HEADER)
        return

    # Sort for deterministic output, which is good for version control
    sorted_strings = sorted(list(translatable_strings))

    print(f"Writing {len(sorted_strings)} strings to: {output_file}")
    with open(output_file, "w", encoding="utf-8") as f:
        f.write(HEADER)
        for s in sorted_strings:
            # Use repr() to correctly handle quotes and special characters
            f.write(f"_({repr(s)})\n")

    print("Done.")


if __name__ == "__main__":
    main()