import csv
import sys
import codecs
reload(sys)
sys.setdefaultencoding('utf8')

with open('out.csv', 'w') as fw:
    csv_writer = csv.writer(fw,delimiter=' ')
    with codecs.open(input, 'r', 'utf-8') as f:
        for line in f:
        line = line.strip() #'\xe2\x80\x8b'
        line_parts = line.split('\t')
        csv_writer.writerow([line_parts[0].encode("gbk", "ignore")])