用DOM树分析代替正则搜索要处理的HTML节点
需求：分析从富文本编辑器传递到服务端的HTML源码，从HTML源码里检索形如 <img src="attachment/100" /> 或者 <a href="attachment/101" > ... </a>的标签，替换成为<tn-media hash = " ... ">的内部标签存入数据库
原方案：使用正则搜索替换上述标签，实际情况上还是比较容易出错的，因此考虑用DOM树节点分析节点代替正则搜索
环境：python2.7 , webpy , BeautifulSoup (DOM分析工具，第三方库)
#!/usr/bin/env python
# -*- coding: utf-8 -*-
import web
urls = (
'/', 'index',
'/parse','parse'
)
web.config.debug = False
app = web.application(urls, globals())
class index:
def GET(self):
s = '''
<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=utf-8" />
</head>
<body>
<form action="/parse" method="post">
<textarea id="s" name="s" style="width:500;height:400">
<font color="red">折腾人的中文</font>
<p><a href="http://www.baidu.com">baidu百度</a><img src="attachment/19" alt="" data-mce-src="attachment/19"></p><p><a href="attachment/20" data-mce-href="attachment/20"><img src="images/home/attType/unknown.png" alt="" data-mce-src="images/home/attType/unknown.png"></a></p><p><span>Bug.txt--0.31KB<br></span></p><p><a href="attachment/21" data-mce-href="attachment/21"><img src="images/home/attType/unknown.png" alt="" data-mce-src="images/home/attType/unknown.png"></a><br><span>BBS--0.23KB</span></p><p><span> </span></p><p> </p>
<p><img src="http://csdnimg.cn/www/images/csdnindex_logo.gif"></p>
</textarea>
<input type="submit" value="post">
</form>
</body>
</html>
'''
return s
import json
from BeautifulSoup import BeautifulSoup
class parse:
def POST(self):
wi = web.input()
s = wi.s
soup = BeautifulSoup(s)
links = soup.findAll('a')
a = []
a1 = []
for link in links:
a.append(link['href'])
if link['href'].find('attachment/') != -1:
a1.append(link['href'])
s1 ="ALL LINK:<br/>%s<br/><br/>WE NEED LINK:<br/>%s" %(json.dumps(a),json.dumps(a1))
imgs = soup.findAll('img')
imgsAll = []
imgsNeed = []
for img in imgs:
imgsAll.append(img['src'])
if img['src'].find('attachment/') != -1:
imgsNeed.append(img['src'])
s2 ="ALL IMG:<br/>%s<br/><br/>WE NEED IMG:<br/>%s" %(json.dumps(imgsAll),json.dumps(imgsNeed))
return '''<html><head><meta http-equiv="Content-Type" content="text/html; charset=utf-8" /></head><body> %s <br/><br/> %s <br/><br/> %s </body></html>''' %(s,s1,s2)
if __name__ == "__main__":
app.run()
不过因为担心BeautifulSoup不够稳定（字符编码会出错，解析糟糕的HTML标记符出错等因素）所以只采用BeautifulSoup功能的一个子集：让它去检索标签，而不修改DOM树，修改HTML的工作仍然由使用它的程序负责完成
def convertToMedia(self,db,s):
''' 新版代码 编写执行通过 等待测试 需要安装Soup'''
def getDigestOfAttchment(db,attId):
sql = "select digest from Attachment where attId=$attId"
list = db.query(sql,vars=locals()).list()
if len(list)>0:
return list[0].digest
else:
return None
def convertOtherToMedia(db,links,s):
for link in links:
#期望的字符串是'/attachment/12' 所以find必然等于0
if link['href'].find('/attachment/') == 0:
listPartOfHref = link['href'].split('/')
#字符串被拆分成['','attachment','12']
if len(listPartOfHref) != 3: continue
if len(listPartOfHref[2]) < 1: continue
attId = int(listPartOfHref[2])
digest = getDigestOfAttchment(db,attId)
if digest == None : continue
s = s.replace(str(link),'<tn-media hash="%s" />'%digest)
return s
def convertPictureToMedia(db,images,s):
for image in images:
#期望的字符串是'/attachment/12' 所以find必然等于0
if image['src'].find('/attachment/') == 0:
listPartOfSrc = image['src'].split('/')
#字符串被拆分成['','attachment','12']
if len(listPartOfSrc) != 3: continue
if len(listPartOfSrc[2]) < 1: continue
attId = int(listPartOfSrc[2])
digest = getDigestOfAttchment(db,attId)
if digest == None : continue
s = s.replace(str(image),'<tn-media hash="%s" />'%digest)
return s
from BeautifulSoup import BeautifulSoup
soup = BeautifulSoup(s)
links = soup.findAll('a')
s = convertOtherToMedia(db,links,s)
images = soup.findAll('img')
s = convertPictureToMedia(db,images,s)
return False,s #提供False是因为原来的实现里外部接口需要函数提供，修改外部代码即可取消
BeautifulSoup在DOM树上进行修改的例子
a = soup.findAll('img')
for i in a:
if i['src'].find('attachment/') > -1:
i.replaceWith('<tnmedia hash="123456">')
浙公网安备 33010602011771号