# -*- coding: UTF-8 -*-
from xml.dom import minidom
import sys,os
import http.client
from urllib.parse import urlparse
import ssl
ssl._create_default_https_context = ssl._create_unverified_context
countRows=0
errors=0
for fileName in os.listdir(sys.path[0]):
if fileName[-4:]==".xml":
print("-----",fileName,"-----")
xmldoc=minidom.parse(fileName)
itemlist = xmldoc.getElementsByTagName("loc")
for i in itemlist:
url=i.firstChild.nodeValue
parse=urlparse(url)
if parse.scheme=="http":
conn = http.client.HTTPConnection(parse.netloc)
elif parse.scheme=="https":
conn = http.client.HTTPSConnection(parse.netloc)
else:
print("Error",url)
continue
conn.request("GET", parse.path)
r1 = conn.getresponse()
if r1.status!=200:
errors+=1
print(r1.status, r1.reason, url)
countRows+=1
sys.stdout.write("countRows/errors: %s/%s \r" % (countRows,errors))
sys.stdout.flush()
Comentarios sobre la versión: Versión 1.0 (0)
No hay comentarios