python 爬取指定url的ICP備案資訊（結構化抓取）

pythontab發表於2013-04-19
Python
#coding=gbk  
import os 
import sys 
import re 
import time 
import urllib2 
  
def perror_and_exit(message, status = -1): 
    sys.stderr.write(message + '\n') 
    sys.exit(status) 
  
def get_text_from_html_tag(html): 
    pattern_text =  re.compile(r">.*?    return pattern_text.findall(html)[0][1:-2].strip() 
  
def parse_alexa(url): 
    url_alexa = "http://icp.alexa.cn/index.php?q=%s" % url 
    print url_alexa 
    #handle exception  
    times = 0
    while times < 5000: #等待有一定次數限制  
        try: 
            alexa = urllib2.urlopen(url_alexa).read() 
  
            pattern_table = re.compile(r".*?", re.DOTALL | re.MULTILINE) 
            match_table = pattern_table.search(alexa) 
            if not match_table: 
                raise BaseException("No table in HTML") 
            break
        except: 
            print "try %s times:sleep %s seconds" % (times, 2**times) 
            times += 1 
            time.sleep(2**times) 
            continue
  
    table = match_table.group() 
    pattern_tr = re.compile(r".*?", re.DOTALL | re.MULTILINE) 
    match_tr = pattern_tr.findall(table) 
    if len(match_tr) != 2: 
        perror_and_exit("table format is incorrect") 
      
    icp_tr = match_tr[1] 
    pattern_td = re.compile(r".*?", re.DOTALL | re.MULTILINE) 
    match_td = pattern_td.findall(icp_tr) 
      
    #print match_td  
    company_name = get_text_from_html_tag(match_td[1]) 
    company_properties = get_text_from_html_tag(match_td[2]) 
    company_icp = get_text_from_html_tag(match_td[3]) 
    company_icp = company_icp[company_icp.find(">") + 1:] 
    company_website_name = get_text_from_html_tag(match_td[4]) 
    company_website_home_page = get_text_from_html_tag(match_td[5]) 
    company_website_home_page = company_website_home_page[company_website_home_page.rfind(">") + 1:] 
    company_detail_url = get_text_from_html_tag(match_td[7]) 
    pattern_href = re.compile(r"href=\".*?\"", re.DOTALL | re.MULTILINE) 
    match_href = pattern_href.findall(company_detail_url) 
    if len(match_href) == 0: 
        company_detail_url = "" 
    else: 
        company_detail_url = match_href[0][len("href=\""):-1] 
    return [url, company_name, company_properties, company_icp, company_website_name, company_website_home_page, company_detail_url] 
    pass
  
if __name__ == "__main__": 
    fw = file("out.txt", "w") 
    for url in sys.stdin: 
        fw.write("\t".join(parse_alexa(url)) + "\n") 
 
#coding=gbk
import os
import sys
import re
import time
import urllib2
 
def perror_and_exit(message, status = -1):
    sys.stderr.write(message + '\n')
    sys.exit(status)
 
def get_text_from_html_tag(html):
    pattern_text =  re.compile(r">.*?    return pattern_text.findall(html)[0][1:-2].strip()
 
def parse_alexa(url):
    url_alexa = "http://icp.alexa.cn/index.php?q=%s" % url
    print url_alexa
    #handle exception
    times = 0
    while times < 5000: #等待有一定次數限制
        try:
            alexa = urllib2.urlopen(url_alexa).read()
 
            pattern_table = re.compile(r".*?", re.DOTALL | re.MULTILINE)
            match_table = pattern_table.search(alexa)
            if not match_table:
                raise BaseException("No table in HTML")
            break
        except:
            print "try %s times:sleep %s seconds" % (times, 2**times)
            times += 1
            time.sleep(2**times)
            continue
 
    table = match_table.group()
    pattern_tr = re.compile(r".*?", re.DOTALL | re.MULTILINE)
    match_tr = pattern_tr.findall(table)
    if len(match_tr) != 2:
        perror_and_exit("table format is incorrect")
    
    icp_tr = match_tr[1]
    pattern_td = re.compile(r".*?", re.DOTALL | re.MULTILINE)
    match_td = pattern_td.findall(icp_tr)
    
    #print match_td
    company_name = get_text_from_html_tag(match_td[1])
    company_properties = get_text_from_html_tag(match_td[2])
    company_icp = get_text_from_html_tag(match_td[3])
    company_icp = company_icp[company_icp.find(">") + 1:]
    company_website_name = get_text_from_html_tag(match_td[4])
    company_website_home_page = get_text_from_html_tag(match_td[5])
    company_website_home_page = company_website_home_page[company_website_home_page.rfind(">") + 1:]
    company_detail_url = get_text_from_html_tag(match_td[7])
    pattern_href = re.compile(r"href=\".*?\"", re.DOTALL | re.MULTILINE)
    match_href = pattern_href.findall(company_detail_url)
    if len(match_href) == 0:
        company_detail_url = ""
    else:
        company_detail_url = match_href[0][len("href=\""):-1]
    return [url, company_name, company_properties, company_icp, company_website_name, company_website_home_page, company_detail_url]
    pass
 
if __name__ == "__main__":
    fw = file("out.txt", "w")
    for url in sys.stdin:
        fw.write("\t".join(parse_alexa(url)) + "\n")[python] view plaincopyprint? time.sleep(2) 
    pass
 
 time.sleep(2)
    pass


每次抓取都會sleep 2s，防止ip被封，實際上即使sleep了IP過一段時間還是會被封

由於是結構化抓取，當網站格式變化此程式將無法使用


相關文章