Diff for Android app translation xml files

translation_diff.py

# A script which takes a reference file (usually English), a translated file and outputs
# strings that need to be translated.
#
# Used for Slovak translation of Phoenix Wallet Android app
# Before using, run:
# pip install lxml

from lxml import etree
import sys

def extract_untranslated_strings(reference_file, translated_file, output_file):
    # Parse the reference and translated XML files
    parser = etree.XMLParser(remove_blank_text=True)
    reference_tree = etree.parse(reference_file, parser)
    reference_root = reference_tree.getroot()

    translated_tree = etree.parse(translated_file, parser)
    translated_root = translated_tree.getroot()

    # Dictionary to store translated strings
    translated_strings = {string.get('name'): string for string in translated_root.findall('.//string')}

    # Create a new XML tree for untranslated strings
    new_root = etree.Element('resources')
    
    # Iterate over strings in the reference file
    for string in reference_root.findall('.//string'):
        string_name = string.get('name')
        
        # Include only untranslated strings in the new XML file
        if string_name not in translated_strings:
            new_string_element = etree.Element('string')
            new_string_element.set('name', string_name)
            new_string_element.text = string.text
            new_root.append(new_string_element)

    # Build the new tree
    new_tree = etree.ElementTree(new_root)

    # Serialize the new XML tree to the output file, formatted for easy reading
    with open(output_file, 'wb') as file:
        new_tree.write(file, pretty_print=True, encoding='utf-8', xml_declaration=True)

if __name__ == "__main__":
    if len(sys.argv) != 4:
        print("Usage: python translation_diff.py <reference_file> <translated_file> <output_file>")
        sys.exit(1)
    
    reference_file = sys.argv[1]
    translated_file = sys.argv[2]
    output_file = sys.argv[3]

    extract_untranslated_strings(reference_file, translated_file, output_file)