import bz2
import sys


def unbz2(file):
    decomp = bz2.BZ2Decompressor()
    buf = b''
    for c in file:
        buf += decomp.decompress(c)

        while b'\n' in buf:
            i = buf.index(b'\n')
            if i + 1 < len(buf):
                ret = buf[:i + 1]
                buf = buf[i + 1:]
                yield ret.decode("utf-8")
            else:
                yield buf
                buf = b''


with open('./wiktionary_dump.xml.bz2', 'rb') as f:
    it = iter(lambda: f.read(32768), b'')
    for a in unbz2(it):
        print(a, end='')