修改xml标签

参考:https://blog.csdn.net/LOVE1055259415/article/details/79166754

xml 解析与生成：https://www.cnblogs.com/jonyq/p/5929073.html

import  xml.dom.minidom

dom=xml.dom.minidom.parse('/media/d_2/everyday/0727/xml/mingpai_0725_2.xml')
root=dom.documentElement
name=root.getElementsByTagName('name')

for i in range(len(name)):
    if 'CJHao' == name[i].firstChild.data:
        print "find___88"
        name[i].firstChild.data = 'chejiahao'
        
with open("/media/d_2/everyday/0727/xml/change.xml",'w') as fh:
    dom.writexml(fh)
    print('写入name/pose OK!')

晕死,还必须另外保存一个文件才能看到更改后的,不是就地修改的

批量改xml中的filename的代码如下：

import xml.dom.minidom as xmldom
import os   

root = "/media/d_2/2018biaozhushuju/20180904KeyMsg/add/20180912_hbd_jielun_qianming/xml/"
save_path = "/media/d_2/2018biaozhushuju/20180904KeyMsg/add/20180912_hbd_jielun_qianming/xml_nice/"

file_list = os.listdir(root)

for file in file_list:
    new_name = file.replace(".xml",".jpg")
    #print "srcname=",file
    #print "new_name=",new_name
    xmlfile = root + file
    domobj = xmldom.parse(xmlfile)
    elementobj = domobj.documentElement
    subElementObj = elementobj.getElementsByTagName("filename")
    subElementObj[0].firstChild.data =new_name
    save_file = save_path + file
    with open(save_file,"w") as fh:
        domobj.writexml(fh)
        print "write xml ok!"

   VOC2007
   mingpai_0725_2.jpg
   
       The VOC2007 Database
       PASCAL VOC2007
       flickr
       329145082
   
   >
       hiromori2
       Hiroyuki Mori
   >
   
       1227
       401
       3
   
   0
   
   
   
   

废了我九牛二虎之力啊，想自己写个脚本根据xml在图上画标记的框的，看到既有的脚本，obj.getElementsByTagName("bndbox")[0].childNodes[0].nodeValue都是用这个函数一层层的解析label的，可是解析坐标xmin，xmax就是报错：xmin = box.getElementsByTagName("xmin")[0].childNodes[0].nodeValue   AttributeError: 'unicode' object has no attribute 'getElementsByTagName'  看了，能解析的obj ，需要解析的bndbox ，无解，因为这个是我们这边流传的工具，应该是能运行的，可能版本问题吧，找了大半天都是不能解析bndbox，后来的后来：

在知乎上面找到一句话：推荐一个简单粗暴的组合：xml2dict+flatten 把xml转成orderdict，然后拍平成单层dict，管你三层还是四层，一次遍历无需递归~

然后我去下载xml2dict。真好用：

from xml2dict import XML2Dict
xml = XML2Dict()
r = xml.parse('/media/d_2/everyday/0915/HBD_JieLQianM_0914_2.xml')
#print r
for item in r.annotation:
    print item
print '------------'
for item in r.annotation.object:
    print item.name, item.bndbox.xmin, item.bndbox.xmax, item.bndbox.ymin, item.bndbox.ymax

##############################################################
Reloaded modules: xml2dict, object_dict
segmented
object
value
filename
source
owner
folder
size
------------
hb_qianming 1332 1491 252 384
hb_qianming 909 1040 277 370
hege 835 903 164 201
hege 1014 1079 86 122
hege 402 476 88 126

两个工具包下载地址：https://download.csdn.net/download/yang332233/10668460