2022年 11月 4日

使用python写xml

工作中需要把PPOCRLabel标注格式转化成需要的xml格式。

遇到的问题：1、之前只使用过使用minidom读xml，还没有使用过minidom写xml。2、minidom写xml顺序与写入的顺序不一致问题。

minidom源码在我本地路径：D:\Anaconda3\Lib\xml\dom\minidom.py

参考：Python3.7写入xml文件 保持节点属性顺序不变

首先导入模块 from xml.dom.minidom import Document,然后按下ctrl 使用鼠标点击minidom进入源码

ctrl+F 搜索 a_names = sorted(attrs.keys())。然后注释掉

添加 a_names = attrs.keys()  不使用排序

 

下面是我的PPOCRLabel标注格式转化成需要的xml格式的代码。

  1. # -*- coding : UTF-8 -*-
  2. # @file   : conver_json_icdar.py
  3. # @Time   : 2021/4/9 11:24
  4. # @Author : wmz
  5.  
  6. import os
  7. import json
  8. import xml.dom.minidom as minidom
  9.  
  10.  
  11. def json_2_icdar(js_path, ic_path):
  12.     with open(js_path, 'r', encoding='utf-8') as f:
  13.         for line in f.readlines():
  14.             print(line)
  15.             content = line.split('\t')
  16.             print(content[0])
  17.             txt_file = str(content[0]).replace('jpg', 'txt')
  18.             dst_file = os.path.join(ic_path, txt_file)
  19.             # write file
  20.             file_lineinfo = open(dst_file, 'w', encoding='utf-8')
  21.             list_dict = json.loads(content[1])
  22.             nsize = len(list_dict)
  23.             print(nsize)
  24.             for i in range(nsize):
  25.                 print(list_dict[i])
  26.                 lin = list_dict[i]
  27.                 info = lin['transcription']
  28.                 points = lin['points']
  29.                 points = [int(y) for x in points for y in x]
  30.                 pts = ','.join(map(str, points))
  31.                 lineinfo = pts + ',' + info + '\n'
  32.                 file_lineinfo.write(lineinfo)
  33.             file_lineinfo.close()
  34.  
  35.  
  36. def json_2_xml(js_path, xml_path):
  37.     with open(js_path, 'r', encoding='utf-8') as f:
  38.         for line in f.readlines():
  39.             print(line)
  40.             content = line.split('\t')
  41.             print(content[0])
  42.             xml_file = str(content[0]).replace('jpg', 'xml')
  43.             dst_xml_file = os.path.join(xml_path, xml_file)
  44.             # txt_file = str(content[0]).replace('jpg', 'txt')
  45.             # dst_file = os.path.join(xml_path, txt_file)
  46.             # write file
  47.             # 1.创建DOM树对象
  48.             dom = minidom.Document()
  49.             # 2.创建根节点。每次都要用DOM对象来创建任何节点。
  50.             root_node = dom.createElement('ImageInfo')
  51.             # 3.用DOM对象添加根节点
  52.             dom.appendChild(root_node)
  53.             # 设置该节点的属性
  54.             root_node.setAttribute('bModify', '3')
  55.  
  56.             # file_lineinfo = open(dst_file, 'w', encoding='utf-8')
  57.             list_dict = json.loads(content[1])
  58.             nsize = len(list_dict)
  59.             print(nsize)
  60.             for i in range(nsize):
  61.                 print(list_dict[i])
  62.                 lin = list_dict[i]
  63.                 info = lin['transcription']
  64.                 points = lin['points']
  65.                 points = [int(y) for x in points for y in x]
  66.                 # 用DOM对象创建元素子节点
  67.                 info_node = dom.createElement('LineInfo')
  68.                 # 用父节点对象添加元素子节点
  69.                 root_node.appendChild(info_node)
  70.                 # 设置该节点的属性
  71.                 info_node.setAttribute('ptLTX', str(points[0]))
  72.                 info_node.setAttribute('ptLTY', str(points[1]))
  73.                 info_node.setAttribute('ptRTX', str(points[2]))
  74.                 info_node.setAttribute('ptRTY', str(points[3]))
  75.                 info_node.setAttribute('ptRBX', str(points[4]))
  76.                 info_node.setAttribute('ptRBY', str(points[5]))
  77.                 info_node.setAttribute('ptLBX', str(points[6]))
  78.                 info_node.setAttribute('ptLBY', str(points[7]))
  79.                 info_node.setAttribute('Chars', info)
  80.                 info_node.setAttribute('bModify', '3')
  81.  
  82.                 pts = ','.join(map(str, points))
  83.                 # lineinfo = pts + ',' + info + '\n'
  84.                 # file_lineinfo.write(lineinfo)
  85.             # file_lineinfo.close()
  86.             with open(dst_xml_file, 'w', encoding='UTF-8') as fh:
  87.                 dom.writexml(fh, indent='', addindent='\t', newl='\n', encoding='UTF-8')
  88.  
  89.  
  90. if __name__ == "__main__":
  91.     # src_path = r"C:\Users\WT\Desktop\hkb-bz\Label.txt"
  92.     src_path = r"C:\Users\WT\Desktop\hkb\Cache.cach"
  93.     dst_path = r"C:\Users\WT\Desktop"
  94.     # src_path = r"C:\Users\WT\Desktop\户口本\Cache.cach"
  95.     # dst_path = r"C:\Users\WT\Desktop"
  96.     json_2_xml(src_path, dst_path)