记录 || Python | 提取xml/tmx文件中的文本内容

# -*- coding:utf-8 -*-
import codecs
import xml.etree.ElementTree as ET
import sys


defaultencoding = 'utf-8'
if sys.getdefaultencoding() != defaultencoding:
    reload(sys)
    sys.setdefaultencoding(defaultencoding)
# 默认情况下，Python采用的是ascii编码方式，将Python的默认编码方式修改为utf-8

tree = ET.parse('test.tmx')
root = tree.getroot()
body = root[1]
txtfile = codecs.open('target.txt', 'w', 'utf-8')
for item in body:
    try:
        bs = item[0][0].text
        zh = item[1][0].text
        bs = bs.decode('utf-8')
        zh = zh.decode('utf-8')
        txtfile.write(bs + "\t" + zh + "\n")
    except Exception as ex:
        print ex.message
        continue

txtfile.close()

 

以上方法能够很快遍历小型的xml和tmx，对于大型文件，直接读入建文档树，内存可能会爆炸，查了一些资料，像是分块处理之类的能够解决，小白没看懂，直接用最简单粗暴的方法，当作普通文件一行一行读入之后正则匹配提取内容

xmlfile = codecs.open("test.xml", 'r', 'utf-8')
txtfile = codecs.open("target.txt", 'w', 'utf-8')
line = xmlfile.readline()
while line:
    result = re.search('', line)
    if result is not None:
        bs = re.sub('<.*?>', "", line)
        bs = bs.strip()
        bs = bs.decode('utf-8')
        txtfile.write(bs+"\t")
    else:
        result = re.search('', line)
        if result is not None:
            zh = re.sub('<.*?>', "", line)
            zh = zh.strip()
            zh = zh.decode('utf-8')
            txtfile.write(zh+"\n")
    line = xmlfile.readline()

xmlfile.close()
txtfile.close()

以上方法太过粗暴，会受很多因素影响，比较合理的办法是先识别tuv标签头作为一对语句的开始，匹配到tuv关闭标签代表这对语句的结束