trunk/pycurl/python/curl/__init__.py

   1 # A high-level interface to the pycurl extension
   2 #
   3 # ** mfx NOTE: the CGI class uses "black magic" using COOKIEFILE in
   4 #    combination with a non-existant file name. See the libcurl docs
   5 #    for more info.
   6 #
   7 # If you want thread-safe operation, you'll have to set the NOSIGNAL option
   8 # yourself.
   9 #
  10 # By Eric S. Raymond, April 2003.
  11
  12 import os, sys, urllib, exceptions, mimetools, pycurl
  13 try:
  14     from cStringIO import StringIO
  15 except ImportError:
  16     from StringIO import StringIO
  17
  18
  19 class Curl:
  20     "High-level interface to cURL functions."
  21     def __init__(self, base_url="", fakeheaders=[]):
  22         self.handle = pycurl.Curl()
  23         # These members might be set.
  24         self.set_url(base_url)
  25         self.verbosity = 0
  26         self.fakeheaders = fakeheaders
  27         # Nothing past here should be modified by the caller.
  28         self.payload = ""
  29         self.header = StringIO()
  30         # Verify that we've got the right site; harmless on a non-SSL connect.
  31         self.set_option(pycurl.SSL_VERIFYHOST, 2)
  32         # Follow redirects in case it wants to take us to a CGI...
  33         self.set_option(pycurl.FOLLOWLOCATION, 1)
  34         self.set_option(pycurl.MAXREDIRS, 5)
  35         # Setting this option with even a nonexistent file makes libcurl
  36         # handle cookie capture and playback automatically.
  37         self.set_option(pycurl.COOKIEFILE, "/dev/null")
  38         # Set timeouts to avoid hanging too long
  39         self.set_timeout(30)
  40         # Use password identification from .netrc automatically
  41         self.set_option(pycurl.NETRC, 1)
  42         # Set up a callback to capture the payload
  43         def payload_callback(x):
  44             self.payload += x
  45         self.set_option(pycurl.WRITEFUNCTION, payload_callback)
  46         def header_callback(x):
  47             self.header.write(x)
  48         self.set_option(pycurl.HEADERFUNCTION, header_callback)
  49
  50     def set_timeout(self, timeout):
  51         "Set timeout for connect and object retrieval (applies for both)"
  52         self.set_option(pycurl.CONNECTTIMEOUT, timeout)
  53         self.set_option(pycurl.TIMEOUT, timeout)
  54
  55     def set_url(self, url):
  56         "Set the base URL to be retrieved."
  57         self.base_url = url
  58         self.set_option(pycurl.URL, self.base_url)
  59
  60     def set_option(self, *args):
  61         "Set an option on the retrieval,"
  62         apply(self.handle.setopt, args)
  63
  64     def set_verbosity(self, level):
  65         "Set verbosity to 1 to see transactions."
  66         self.set_option(pycurl.VERBOSE, level)
  67
  68     def __request(self, relative_url=None):
  69         "Perform the pending request."
  70         if self.fakeheaders:
  71             self.set_option(pycurl.HTTPHEADER, self.fakeheaders)
  72         if relative_url:
  73             self.set_option(pycurl.URL,os.path.join(self.base_url,relative_url))
  74         self.header.seek(0,0)
  75         self.payload = ""
  76         self.handle.perform()
  77         return self.payload
  78
  79     def get(self, url="", params=None):
  80         "Ship a GET request for a specified URL, capture the response."
  81         if params:
  82             url += "?" + urllib.urlencode(params)
  83         self.set_option(pycurl.HTTPGET, 1)
  84         return self.__request(url)
  85
  86     def post(self, cgi, params):
  87         "Ship a POST request to a specified CGI, capture the response."
  88         self.set_option(pycurl.POST, 1)
  89         self.set_option(pycurl.POSTFIELDS, urllib.urlencode(params))
  90         return self.__request(cgi)
  91
  92     def body(self):
  93         "Return the body from the last response."
  94         return self.payload
  95
  96     def info(self):
  97         "Return an RFC822 object with info on the page."
  98         self.header.seek(0,0)
  99         url = self.handle.getinfo(pycurl.EFFECTIVE_URL)
 100         if url[:5] == 'http:':
 101             self.header.readline()
 102             m = mimetools.Message(self.header)
 103         else:
 104             m = mimetools.Message(StringIO())
 105         m['effective-url'] = url
 106         m['http-code'] = str(self.handle.getinfo(pycurl.HTTP_CODE))
 107         m['total-time'] = str(self.handle.getinfo(pycurl.TOTAL_TIME))
 108         m['namelookup-time'] = str(self.handle.getinfo(pycurl.NAMELOOKUP_TIME))
 109         m['connect-time'] = str(self.handle.getinfo(pycurl.CONNECT_TIME))
 110         m['pretransfer-time'] = str(self.handle.getinfo(pycurl.PRETRANSFER_TIME))
 111         m['redirect-time'] = str(self.handle.getinfo(pycurl.REDIRECT_TIME))
 112         m['redirect-count'] = str(self.handle.getinfo(pycurl.REDIRECT_COUNT))
 113         m['size-upload'] = str(self.handle.getinfo(pycurl.SIZE_UPLOAD))
 114         m['size-download'] = str(self.handle.getinfo(pycurl.SIZE_DOWNLOAD))
 115         m['speed-upload'] = str(self.handle.getinfo(pycurl.SPEED_UPLOAD))
 116         m['header-size'] = str(self.handle.getinfo(pycurl.HEADER_SIZE))
 117         m['request-size'] = str(self.handle.getinfo(pycurl.REQUEST_SIZE))
 118         m['content-length-download'] = str(self.handle.getinfo(pycurl.CONTENT_LENGTH_DOWNLOAD))
 119         m['content-length-upload'] = str(self.handle.getinfo(pycurl.CONTENT_LENGTH_UPLOAD))
 120         m['content-type'] = (self.handle.getinfo(pycurl.CONTENT_TYPE) or '').strip(';')
 121         return m
 122
 123     def answered(self, check):
 124         "Did a given check string occur in the last payload?"
 125         return self.payload.find(check) >= 0
 126
 127     def close(self):
 128         "Close a session, freeing resources."
 129         self.handle.close()
 130         self.header.close()
 131
 132     def __del__(self):
 133         self.close()
 134
 135
 136 if __name__ == "__main__":
 137     if len(sys.argv) < 2:
 138         url = 'http://curl.haxx.se'
 139     else:
 140         url = sys.argv[1]
 141     c = Curl()
 142     c.get(url)
 143     print c.body()
 144     print '='*74 + '\n'
 145     print c.info()
 146     c.close()