多线程爬虫批量下载pcgame图片url 保存为xml的实现代码
#coding=gbk
from xml.dom import minidom,node
import urllib2,re,os
def readsrc(src):
try:
url = urllib2.urlopen(src)
content = url.read()#.decode('utf-8')
return content
except:
print 'error'
return none
def pictype(content):
'''
通过抓取网站导航栏，获得网站的图片类型
返回列表，每个列表元素为一个字典，addr代表图片类型对于的链接，name代表图片类型的名称
错误会返回none
'''
p = re.compile(r'<ul>(.*)</ul>',re.s)
r=p.search(content)
if r:
content=r.group()
else:
print none
p = re.compile(r'<li\s*.*?>\s*<a href *= *"(?p<addr>.*?)">(?p<name>.*?)\s*</a>\s*</li>')
l = [i.groupdict() for i in p.finditer(content)]
l=l[1:]
if len(l):return l
else:return none
def pageinfo(src):
'''
获取一个页面的详细信息
返回对于的字典列表
name:图片的名字
cutaddr：缩小的浏览图
picaddr：实际图片的地址
'''
d=os.path.split(src)[0]
try:
url = urllib2.urlopen(src)
content = url.read()#.decode('utf-8')
except:
print 'error'
return none
#find all the pictures info in a page
p = re.compile(r'<ul.*?>(.*?)</ul>',re.s)
r = p.findall(content)
if not r: return none
r = r[1]
p = re.compile(r'<li><a href="(?p<picaddr>.*?)".*?><img.*?alt="(?p<name>.*?)" *src="(?p<cutaddr>.*?)" */></a>.*?</li>')
l = [ i.groupdict() for i in p.finditer(r)]
for i in l:
i['picaddr']=d+'/'+i['picaddr']
if len(l): return l
else: return none
def nextpageaddr(src):
'''
从页面的html源码中获取下一个页面地址的名称，最后一页返回none
'''
content=readsrc(src)
p = re.compile(r'<a class="next" href="(.*?)">.*?</a>')
r = p.search(content)
if r:
return os.path.dirname(src)+"/"+r.group(1)
else:
return none
def picinfoaddr(src):
'''
参数相册图集的html代码
返回全部图片的相对地址
'''
content=readsrc(src)
p = re.compile(r'<div class="picinfo">.*?<a href="(?p<addr>.*?)".*?>.*?</div>',re.s)
r = p.search(content)
if r:
return os.path.dirname(src)+"/"+r.group(1)
else:
return none
def parseinfo(content):
'''
读取全部图片html代码，获得一个相册的详细信息
kw：关键字
title：标题
type：类型
pic：各个图片的地址列表，末尾加上_220x165，_medium,_small 可以得到不同大小的图片
'''
info={}
temp=str()
#title
temp=''
r=re.search('<h1>(.*?)</h1>',content)#get the pic title
if r:
temp = r.group(1)
info['title']=temp
#keyword
temp=''
r=re.search('<meta name="keywords" content="(.*?)" />',content)
if r:
temp = r.group(1)
info['kw']=temp
#type
r=re.findall('<i><a.*?>(.*?)</a></i>.*?>',content)
if r:
info['type']=':'.join(r)
else:
info['type']=''
r=re.search('<ul class=".*?">(.*?)</ul>',content,re.s)
if not r:return none
content=r.group(1)#filter content
# print content
r=re.findall('<a href=".*?<img.*?src="(.*?)".*?</a>',content)
for index,i in enumerate(r):
r[index]=i[0:i.rfind('_')]
# print r[index]
info['pic']=r
return info
import threading
class mthread(threading.thread):
def __init__(self,tp,addr,lock):
threading.thread.__init__(self)
# self.doc = minidom.document()
self.doc=minidom.document()
self.tp=tp
self.lock=lock
self.addr=addr
self.thread_stop=false
self.picdoc=none
def run(self):
self.picdoc = self.doc.createelement('urlclass')
# print self.tp
self.picdoc.setattribute('type',self.tp)
# self.doc.appendchild(self.picdoc)
m=pageinfo(self.addr)
while self.addr:
for i in m:
# print i['picaddr']
picaddr=picinfoaddr(i['picaddr'])
# print picaddr
info=parseinfo(readsrc(picaddr))
name=info['title']
picture=doc.createelement('picture')
title = doc.createelement('title')
title.appendchild(doc.createtextnode(info['title']))
picture.appendchild(title)
keyword = doc.createelement('keywords')
keyword.appendchild(doc.createtextnode(info['kw']))
picture.appendchild(keyword)
tp = doc.createelement('pictype')
tp.appendchild(doc.createtextnode(info['type']))
picture.appendchild(tp)
cuturl = doc.createelement('piccut')
cuturl.appendchild(doc.createtextnode(i['cutaddr']))
picture.appendchild(cuturl)
urls = doc.createelement('urls')
self.lock.acquire()
print 'downloading ',name
self.lock.release()
for picurl in info['pic']:
singleurl=doc.createelement('url')
singleurl.appendchild(doc.createtextnode(picurl+'.jpg'))
urls.appendchild(singleurl)
picture.appendchild(urls)
self.picdoc.appendchild(picture)
m=pageinfo(self.addr)
self.addr=nextpageaddr(self.addr)
# f = open('c:\\'+self.tp+'.xml','w')
# f.write(doc.toprettyxml(indent = ''))
# f.close()
def stop(self):
self.thread_stop=true
path='c:\\pict\\'#下载的路径
#import sys
sys.exit(12)
content=readsrc('http://photos.pcgames.com.cn/cate/3/1.html')
r=pictype(content)
lt=[]
doc = minidom.document()
root=doc.createelement('url_resource')
root.setattribute('type','url')
root.setattribute('urltype','image')
root.setattribute('imgfmt','jpg')
doc.appendchild(root)
lock=threading.rlock()
for iaddr in r:
print 'downloading type: ',iaddr['name']
addr=iaddr['addr']
th=mthread(iaddr['name'],addr,lock)
lt.append(th)
th.start()
for t in lt:
t.join()
root.appendchild(t.picdoc)
print 'write'
f = open('c:\\'+'urls'+'.xml','w')
f.write(doc.toprettyxml(indent = ''))
f.close()
print doc.toprettyxml()
print 'end'