tmx_parser.py

import re
from pathlib import Path
from bs4 import BeautifulSoup

def parse_segment(multilingual_segment, src_lang, target_lang):
    src_seg = ''
    target_seg = ''
    segments = multilingual_segment.find_all('tuv')
    for segment in segments:
        seg_text = segment.find("seg").text if segment.find("seg").text else '-----'
        if segment.get("xml:lang", "") == src_lang:
            src_seg = seg_text.replace('\n', '')
            src_seg = seg_text.strip() + '\n'
        elif segment.get("xml:lang", "") == target_lang:
            target_seg = seg_text.replace('\n', '')
            target_seg = seg_text.strip() + '\n'
    return [src_seg, target_seg]

def parse_tmx(tmx_content, src_lang='bo', target_lang='en'):
    bi_text = {
        src_lang: '',
        target_lang: ''
    }
    soup = BeautifulSoup(tmx_content, "xml")
    multilingual_segments = soup.find_all("tu")
    for multilingual_segment in multilingual_segments:
        src_seg, tar_seg = parse_segment(multilingual_segment, src_lang, target_lang)
        bi_text[src_lang] += src_seg
        bi_text[target_lang] += tar_seg
    return bi_text


if __name__ == "__main__":
    tmx_paths = list(Path('./tmx').iterdir())
    for tmx_path in tmx_paths:
        tmx = tmx_path.read_text(encoding='utf-8')
        bi_text = parse_tmx(tmx)
        tib_text = bi_text['bo']
        eng_text = bi_text['en']
        eng_text = re.sub('\d+', '', eng_text)
        eng_text = eng_text.replace('[]', '')
        eng_text = eng_text.replace('()', '')
        eng_text = eng_text.replace(r'{}', '')
        Path(f'./bo_text/{tmx_path.stem}.txt').write_text(tib_text)
        Path(f'./en_text/{tmx_path.stem}.txt').write_text(eng_text)