#!/usr/bin/python
# -*- coding: utf-8 -*-
'''
Created on Nov 9, 2011
 
@author: li3huo
'
 
''
 
import urllib, urllib2
import sys
import time
class DefaultErrorHandler(urllib2.HTTPDefaultErrorHandler):
	"""用來保證請求中記錄Http狀態
"""
	def http_error_default(self, req, fp, code, msg, headers):
		result = urllib2.HTTPError(
			req.get_full_url(), code, msg, headers, fp)
		result.status = code
		return
 
 result
 
class Sample():
	"""a sample is the url i want to download
"""
	url = None
	contentLength = 0
	etag = None
	lastModified = None
	data = None
	path = None
 
	def __init__(self, url, contentLength=0, etag=None, lastModified=None):
		self.url = url
		self.contentLength = 0
		self.etag = etag
		self.lastModified = lastModified
		self.status = 200
		self.file = file
	def __repr__(self):
		return repr("Http Status=%d; Length=%d; Last Modified Time=%s; eTag=%s" % (self.status, self.contentLength, self.lastModified, self.etag))
 
	def downloadSample(self):
		request = urllib2.Request(self.url)
		request.add_header('User-Agent', "Mozilla/5.0 (Macintosh; U; PPC Mac OS X; en)")
		if self.lastModified:
			request.add_header('If-Modified-Since', self.lastModified)
		if self.etag:
			request.add_header('If-None-Match', self.etag)
		conn = urllib2.build_opener(DefaultErrorHandler()).open(request)
 
 
		if hasattr(conn, 'headers'):
			# save ETag, if the server sent one
			self.etag = conn.headers.get('ETag')
			# save Last-Modified header, if the server sent one
			self.lastModified = conn.headers.get('Last-Modified')
 
			self.contentLength = conn.headers.get("content-length")
 
		if hasattr(conn, 'status'):
			self.status = conn.status
			print "status=%d" % self.status
 
		self.data = conn.read()
 
		if self.status == 304:
			print "the content is same, so return nothing!"
 
		if not self.contentLength:
			self.contentLength = len(self.data)
 
		conn.close()
 
	if __name__ == '__main__':
	url = 'http://www.sina.com.cn'
	sample = Sample(url)
	sample.downloadSample()
	print sample
	sample.downloadSample()
	print sample