3 @copyright: (c) 2005 by Szoftver Messias Bt.
6 Objects of the MozillaEmulator class can emulate a browser that is capable of:
9 - configurable user agent string
11 - multipart POST (send files)
12 - receive content into file
14 I have seen many requests on the python mailing list about how to emulate a browser. I'm using this class for years now, without any problems. This is how you can use it:
17 2. Install and open the livehttpheaders plugin
18 3. Use the website manually with firefox
19 4. Check the GET and POST requests in the livehttpheaders capture window
20 5. Create an instance of the above class and send the same GET and POST requests to the server.
24 - You can change user agent string in the build_opened method
25 - The "encode_multipart_formdata" function can be used alone to create POST data from a list of field values and files
35 _moduleLogger = logging.getLogger(__name__)
36 socket.setdefaulttimeout(45)
39 def add_proxy(protocol, url, port):
40 proxyInfo = "%s:%s" % (url, port)
41 proxy = urllib2.ProxyHandler(
44 opener = urllib2.build_opener(proxy)
45 urllib2.install_opener(opener)
48 class MozillaEmulator(object):
50 USER_AGENT = 'Mozilla/5.0 (Windows; U; Windows NT 5.1; de; rv:1.9.1.4) Gecko/20091016 Firefox/3.5.4 (.NET CLR 3.5.30729)'
52 def __init__(self, trycount = 1):
53 """Create a new MozillaEmulator object.
55 @param trycount: The download() method will retry the operation if it
56 fails. You can specify -1 for infinite retrying. A value of 0 means no
57 retrying. A value of 1 means one retry. etc."""
59 self.trycount = trycount
60 self._cookies = cookielib.LWPCookieJar()
61 self._loadedFromCookies = False
62 self._storeCookies = False
64 def load_cookies(self, path):
65 assert not self._loadedFromCookies, "Load cookies only once"
69 self._cookies.filename = path
72 except cookielib.LoadError:
73 _moduleLogger.exception("Bad cookie file")
75 _moduleLogger.exception("No cookie file")
77 _moduleLogger.exception("Unknown error with cookies")
79 self._loadedFromCookies = True
80 self._storeCookies = True
82 return self._loadedFromCookies
84 def save_cookies(self):
85 if self._storeCookies:
88 def clear_cookies(self):
89 if self._storeCookies:
92 def download(self, url,
93 postdata = None, extraheaders = None, forbidRedirect = False,
94 trycount = None, only_head = False,
96 """Download an URL with GET or POST methods.
98 @param postdata: It can be a string that will be POST-ed to the URL.
99 When None is given, the method will be GET instead.
100 @param extraheaders: You can add/modify HTTP headers with a dict here.
101 @param forbidRedirect: Set this flag if you do not want to handle
102 HTTP 301 and 302 redirects.
103 @param trycount: Specify the maximum number of retries here.
104 0 means no retry on error. Using -1 means infinite retring.
105 None means the default value (that is self.trycount).
106 @param only_head: Create the openerdirector and return it. In other
107 words, this will not retrieve any content except HTTP headers.
109 @return: The raw HTML page data
111 _moduleLogger.debug("Performing download of %s" % url)
113 if extraheaders is None:
116 trycount = self.trycount
121 req, u = self._build_opener(url, postdata, extraheaders, forbidRedirect)
122 openerdirector = u.open(req)
124 _moduleLogger.info("%r - %r" % (req.get_method(), url))
125 _moduleLogger.info("%r - %r" % (openerdirector.code, openerdirector.msg))
126 _moduleLogger.info("%r" % (openerdirector.headers))
127 self._cookies.extract_cookies(openerdirector, req)
129 return openerdirector
131 return self._read(openerdirector, trycount)
132 except urllib2.URLError, e:
133 _moduleLogger.debug("%s: %s" % (e, url))
135 if (-1 < trycount) and (trycount < cnt):
139 _moduleLogger.debug("MozillaEmulator: urllib2.URLError, retrying %d" % cnt)
141 def _build_opener(self, url, postdata = None, extraheaders = None, forbidRedirect = False):
142 if extraheaders is None:
146 'Accept': 'text/xml,application/xml,application/xhtml+xml,text/html;q=0.9,text/plain;q=0.8,image/png',
147 'Accept-Language': 'en,en-us;q=0.5',
148 'Accept-Charset': 'ISO-8859-1,utf-8;q=0.7,*;q=0.7',
149 'User-Agent': self.USER_AGENT,
151 for key, value in extraheaders.iteritems():
152 txheaders[key] = value
153 req = urllib2.Request(url, postdata, txheaders)
154 self._cookies.add_cookie_header(req)
156 redirector = HTTPNoRedirector()
157 #_moduleLogger.info("Redirection disabled")
159 redirector = urllib2.HTTPRedirectHandler()
160 #_moduleLogger.info("Redirection enabled")
162 http_handler = urllib2.HTTPHandler(debuglevel=self.debug)
163 https_handler = urllib2.HTTPSHandler(debuglevel=self.debug)
165 u = urllib2.build_opener(
168 urllib2.HTTPCookieProcessor(self._cookies),
171 if not postdata is None:
172 req.add_data(postdata)
175 def _read(self, openerdirector, trycount):
178 chunk = openerdirector.read()
180 #while chunk and cnt < trycount:
183 # chunk = openerdirector.read()
184 # chunks.append(chunk)
186 data = "".join(chunks)
188 if "Content-Length" in openerdirector.info():
189 assert len(data) == int(openerdirector.info()["Content-Length"]), "The packet header promised %s of data but only was able to read %s of data" % (
190 openerdirector.info()["Content-Length"],
197 class HTTPNoRedirector(urllib2.HTTPRedirectHandler):
198 """This is a custom http redirect handler that FORBIDS redirection."""
200 def http_error_302(self, req, fp, code, msg, headers):
201 e = urllib2.HTTPError(req.get_full_url(), code, msg, headers, fp)
202 if e.code in (301, 302):
203 if 'location' in headers:
204 newurl = headers.getheaders('location')[0]
205 elif 'uri' in headers:
206 newurl = headers.getheaders('uri')[0]
208 _moduleLogger.info("New url: %s" % e.newurl)