#!/usr/bin/env python
# -*- coding: utf-8 -*-
# File              : gist:2f2acad6fce7ebc8baff1288633309d7/xliff_translate.py
# Author            : Nester Chung <twntwn3838@gmail>
# Date              : 09.08.2018
# Last Modified Date: 10.08.2018

# Please install module before use this script:
#   lxml
#   googletrans


import argparse
import codecs
import logging as log
from lxml import etree
from lxml.etree import tostring
from googletrans import Translator

NS = {'x':'urn:oasis:names:tc:xliff:document:1.2'}

TRANSLATOR = Translator()

#xliff language code mapping to google language code
LM = {"zh-Hans" : "zh-Cn",
      "zh-Hans-CN":"zh-Cn",
      "yue-Hans-CN" : "zh-Cn",
      "zh-HK": "zh-Tw",
      "zh-Hant":"zh-Tw",
      "zh-Hant-Cn":"zh-Tw",
      "zh-Hant-MO" : "zh-Tw",
      "zh-Hant-TW" : "zh-Tw",
      "zh-Hant-HK" : "zh-Tw",
      "yue-Hant-CN" : "zh-Tw"}

def translate(filein, fileout):
    """ Translate the dom """
    log.debug("read input file: %s", filein)
    locale_tree = etree.parse(filein)
    locale_root = locale_tree.getroot()

    for trans_node in locale_root.xpath('//x:trans-unit', namespaces=NS):
        for child in trans_node.xpath('./x:target', namespaces=NS):
            src_lang = trans_node.getparent().getparent().get('source-language')
            dest_lang = trans_node.getparent().getparent().get('target-language')
            source_string = trans_node.xpath('./x:source', namespaces=NS)[0].text
            tobj = TRANSLATOR.translate(
                source_string,
                src=LM.get(src_lang, src_lang),
                dest=LM.get(dest_lang, dest_lang))
            log.debug("%s => %s", source_string, tobj.text)
            child.text = tobj.text

    tmp = tostring(locale_root, encoding='unicode')
    if fileout is None:
        print tmp
    else:
        with codecs.open(fileout, encoding='utf-8', mode='w+') as out:
            log.debug("write to %s", fileout)
            out.write(tmp)

if __name__ == "__main__":
    PARSER = argparse.ArgumentParser(description='Translate xliff file')
    PARSER.add_argument('filein', metavar='INPUT', help='input filename')
    PARSER.add_argument('-o', '--output', dest='fileout' help='output filename')
    PARSER.add_argument('-v', '--verbose', action='store_true')
    ARGS = PARSER.parse_args()

    if ARGS.verbose:
        log.basicConfig(format="%(levelname)s: %(message)s", level=log.DEBUG)
    else:
        log.basicConfig(format="%(levelname)s: %(message)s")

    translate(ARGS.filein, ARGS.fileout)