Guido van Rossum | 41999c1 | 1997-12-09 00:12:23 +0000 | [diff] [blame] | 1 | """HTTP client class |
Guido van Rossum | 23acc95 | 1994-02-21 16:36:04 +0000 | [diff] [blame] | 2 | |
Guido van Rossum | 41999c1 | 1997-12-09 00:12:23 +0000 | [diff] [blame] | 3 | See the following URL for a description of the HTTP/1.0 protocol: |
| 4 | http://www.w3.org/hypertext/WWW/Protocols/ |
| 5 | (I actually implemented it from a much earlier draft.) |
| 6 | |
| 7 | Example: |
| 8 | |
| 9 | >>> from httplib import HTTP |
| 10 | >>> h = HTTP('www.python.org') |
| 11 | >>> h.putrequest('GET', '/index.html') |
| 12 | >>> h.putheader('Accept', 'text/html') |
| 13 | >>> h.putheader('Accept', 'text/plain') |
| 14 | >>> h.endheaders() |
| 15 | >>> errcode, errmsg, headers = h.getreply() |
| 16 | >>> if errcode == 200: |
| 17 | ... f = h.getfile() |
| 18 | ... print f.read() # Print the raw HTML |
| 19 | ... |
| 20 | <HEAD> |
| 21 | <TITLE>Python Language Home Page</TITLE> |
| 22 | [...many more lines...] |
| 23 | >>> |
| 24 | |
| 25 | Note that an HTTP object is used for a single request -- to issue a |
| 26 | second request to the same server, you create a new HTTP object. |
| 27 | (This is in accordance with the protocol, which uses a new TCP |
| 28 | connection for each request.) |
| 29 | """ |
Guido van Rossum | 23acc95 | 1994-02-21 16:36:04 +0000 | [diff] [blame] | 30 | |
Guido van Rossum | 09c8b6c | 1999-12-07 21:37:17 +0000 | [diff] [blame] | 31 | import os |
Guido van Rossum | 23acc95 | 1994-02-21 16:36:04 +0000 | [diff] [blame] | 32 | import socket |
| 33 | import string |
Guido van Rossum | 65ab98c | 1995-08-07 20:13:02 +0000 | [diff] [blame] | 34 | import mimetools |
Guido van Rossum | 23acc95 | 1994-02-21 16:36:04 +0000 | [diff] [blame] | 35 | |
Guido van Rossum | 09c8b6c | 1999-12-07 21:37:17 +0000 | [diff] [blame] | 36 | try: |
| 37 | from cStringIO import StringIO |
| 38 | except: |
| 39 | from StringIO import StringIO |
| 40 | |
Guido van Rossum | 23acc95 | 1994-02-21 16:36:04 +0000 | [diff] [blame] | 41 | HTTP_VERSION = 'HTTP/1.0' |
| 42 | HTTP_PORT = 80 |
Guido van Rossum | 09c8b6c | 1999-12-07 21:37:17 +0000 | [diff] [blame] | 43 | HTTPS_PORT = 443 |
| 44 | |
| 45 | class FakeSocket: |
| 46 | def __init__(self, sock, ssl): |
Fred Drake | 13a2c27 | 2000-02-10 17:17:14 +0000 | [diff] [blame] | 47 | self.__sock = sock |
| 48 | self.__ssl = ssl |
| 49 | return |
Guido van Rossum | 09c8b6c | 1999-12-07 21:37:17 +0000 | [diff] [blame] | 50 | |
Fred Drake | 13a2c27 | 2000-02-10 17:17:14 +0000 | [diff] [blame] | 51 | def makefile(self, mode): # hopefully, never have to write |
| 52 | msgbuf = "" |
| 53 | while 1: |
| 54 | try: |
| 55 | msgbuf = msgbuf + self.__ssl.read() |
| 56 | except socket.sslerror, msg: |
| 57 | break |
| 58 | return StringIO(msgbuf) |
Guido van Rossum | 09c8b6c | 1999-12-07 21:37:17 +0000 | [diff] [blame] | 59 | |
| 60 | def send(self, stuff, flags = 0): |
Fred Drake | 13a2c27 | 2000-02-10 17:17:14 +0000 | [diff] [blame] | 61 | return self.__ssl.write(stuff) |
Guido van Rossum | 09c8b6c | 1999-12-07 21:37:17 +0000 | [diff] [blame] | 62 | |
| 63 | def recv(self, len = 1024, flags = 0): |
Fred Drake | 13a2c27 | 2000-02-10 17:17:14 +0000 | [diff] [blame] | 64 | return self.__ssl.read(len) |
Guido van Rossum | 09c8b6c | 1999-12-07 21:37:17 +0000 | [diff] [blame] | 65 | |
| 66 | def __getattr__(self, attr): |
Fred Drake | 13a2c27 | 2000-02-10 17:17:14 +0000 | [diff] [blame] | 67 | return getattr(self.__sock, attr) |
Guido van Rossum | 23acc95 | 1994-02-21 16:36:04 +0000 | [diff] [blame] | 68 | |
Guido van Rossum | 23acc95 | 1994-02-21 16:36:04 +0000 | [diff] [blame] | 69 | class HTTP: |
Guido van Rossum | 41999c1 | 1997-12-09 00:12:23 +0000 | [diff] [blame] | 70 | """This class manages a connection to an HTTP server.""" |
Guido van Rossum | 09c8b6c | 1999-12-07 21:37:17 +0000 | [diff] [blame] | 71 | |
| 72 | def __init__(self, host = '', port = 0, **x509): |
Guido van Rossum | 41999c1 | 1997-12-09 00:12:23 +0000 | [diff] [blame] | 73 | """Initialize a new instance. |
Guido van Rossum | 23acc95 | 1994-02-21 16:36:04 +0000 | [diff] [blame] | 74 | |
Guido van Rossum | 41999c1 | 1997-12-09 00:12:23 +0000 | [diff] [blame] | 75 | If specified, `host' is the name of the remote host to which |
| 76 | to connect. If specified, `port' specifies the port to which |
| 77 | to connect. By default, httplib.HTTP_PORT is used. |
Guido van Rossum | 23acc95 | 1994-02-21 16:36:04 +0000 | [diff] [blame] | 78 | |
Guido van Rossum | 41999c1 | 1997-12-09 00:12:23 +0000 | [diff] [blame] | 79 | """ |
Guido van Rossum | 09c8b6c | 1999-12-07 21:37:17 +0000 | [diff] [blame] | 80 | self.key_file = x509.get('key_file') |
| 81 | self.cert_file = x509.get('cert_file') |
Guido van Rossum | 41999c1 | 1997-12-09 00:12:23 +0000 | [diff] [blame] | 82 | self.debuglevel = 0 |
| 83 | self.file = None |
| 84 | if host: self.connect(host, port) |
Guido van Rossum | 09c8b6c | 1999-12-07 21:37:17 +0000 | [diff] [blame] | 85 | |
Guido van Rossum | 41999c1 | 1997-12-09 00:12:23 +0000 | [diff] [blame] | 86 | def set_debuglevel(self, debuglevel): |
| 87 | """Set the debug output level. |
Guido van Rossum | 23acc95 | 1994-02-21 16:36:04 +0000 | [diff] [blame] | 88 | |
Guido van Rossum | 41999c1 | 1997-12-09 00:12:23 +0000 | [diff] [blame] | 89 | A non-false value results in debug messages for connection and |
| 90 | for all messages sent to and received from the server. |
Guido van Rossum | 23acc95 | 1994-02-21 16:36:04 +0000 | [diff] [blame] | 91 | |
Guido van Rossum | 41999c1 | 1997-12-09 00:12:23 +0000 | [diff] [blame] | 92 | """ |
| 93 | self.debuglevel = debuglevel |
Guido van Rossum | 09c8b6c | 1999-12-07 21:37:17 +0000 | [diff] [blame] | 94 | |
Guido van Rossum | 41999c1 | 1997-12-09 00:12:23 +0000 | [diff] [blame] | 95 | def connect(self, host, port = 0): |
| 96 | """Connect to a host on a given port. |
Guido van Rossum | 09c8b6c | 1999-12-07 21:37:17 +0000 | [diff] [blame] | 97 | |
Guido van Rossum | 41999c1 | 1997-12-09 00:12:23 +0000 | [diff] [blame] | 98 | Note: This method is automatically invoked by __init__, |
| 99 | if a host is specified during instantiation. |
Guido van Rossum | 23acc95 | 1994-02-21 16:36:04 +0000 | [diff] [blame] | 100 | |
Guido van Rossum | 41999c1 | 1997-12-09 00:12:23 +0000 | [diff] [blame] | 101 | """ |
| 102 | if not port: |
| 103 | i = string.find(host, ':') |
| 104 | if i >= 0: |
| 105 | host, port = host[:i], host[i+1:] |
| 106 | try: port = string.atoi(port) |
| 107 | except string.atoi_error: |
| 108 | raise socket.error, "nonnumeric port" |
| 109 | if not port: port = HTTP_PORT |
| 110 | self.sock = socket.socket(socket.AF_INET, socket.SOCK_STREAM) |
| 111 | if self.debuglevel > 0: print 'connect:', (host, port) |
| 112 | self.sock.connect(host, port) |
Guido van Rossum | 09c8b6c | 1999-12-07 21:37:17 +0000 | [diff] [blame] | 113 | |
Guido van Rossum | 41999c1 | 1997-12-09 00:12:23 +0000 | [diff] [blame] | 114 | def send(self, str): |
| 115 | """Send `str' to the server.""" |
| 116 | if self.debuglevel > 0: print 'send:', `str` |
| 117 | self.sock.send(str) |
Guido van Rossum | 09c8b6c | 1999-12-07 21:37:17 +0000 | [diff] [blame] | 118 | |
Guido van Rossum | 41999c1 | 1997-12-09 00:12:23 +0000 | [diff] [blame] | 119 | def putrequest(self, request, selector): |
| 120 | """Send a request to the server. |
Guido van Rossum | 23acc95 | 1994-02-21 16:36:04 +0000 | [diff] [blame] | 121 | |
Guido van Rossum | 41999c1 | 1997-12-09 00:12:23 +0000 | [diff] [blame] | 122 | `request' specifies an HTTP request method, e.g. 'GET'. |
| 123 | `selector' specifies the object being requested, e.g. |
| 124 | '/index.html'. |
Guido van Rossum | 23acc95 | 1994-02-21 16:36:04 +0000 | [diff] [blame] | 125 | |
Guido van Rossum | 41999c1 | 1997-12-09 00:12:23 +0000 | [diff] [blame] | 126 | """ |
| 127 | if not selector: selector = '/' |
| 128 | str = '%s %s %s\r\n' % (request, selector, HTTP_VERSION) |
| 129 | self.send(str) |
Guido van Rossum | 09c8b6c | 1999-12-07 21:37:17 +0000 | [diff] [blame] | 130 | |
Guido van Rossum | 41999c1 | 1997-12-09 00:12:23 +0000 | [diff] [blame] | 131 | def putheader(self, header, *args): |
| 132 | """Send a request header line to the server. |
Guido van Rossum | 23acc95 | 1994-02-21 16:36:04 +0000 | [diff] [blame] | 133 | |
Guido van Rossum | 41999c1 | 1997-12-09 00:12:23 +0000 | [diff] [blame] | 134 | For example: h.putheader('Accept', 'text/html') |
Guido van Rossum | 23acc95 | 1994-02-21 16:36:04 +0000 | [diff] [blame] | 135 | |
Guido van Rossum | 41999c1 | 1997-12-09 00:12:23 +0000 | [diff] [blame] | 136 | """ |
| 137 | str = '%s: %s\r\n' % (header, string.joinfields(args,'\r\n\t')) |
| 138 | self.send(str) |
Guido van Rossum | 09c8b6c | 1999-12-07 21:37:17 +0000 | [diff] [blame] | 139 | |
Guido van Rossum | 41999c1 | 1997-12-09 00:12:23 +0000 | [diff] [blame] | 140 | def endheaders(self): |
| 141 | """Indicate that the last header line has been sent to the server.""" |
| 142 | self.send('\r\n') |
Guido van Rossum | 09c8b6c | 1999-12-07 21:37:17 +0000 | [diff] [blame] | 143 | |
Guido van Rossum | 41999c1 | 1997-12-09 00:12:23 +0000 | [diff] [blame] | 144 | def getreply(self): |
| 145 | """Get a reply from the server. |
Guido van Rossum | 09c8b6c | 1999-12-07 21:37:17 +0000 | [diff] [blame] | 146 | |
Guido van Rossum | 41999c1 | 1997-12-09 00:12:23 +0000 | [diff] [blame] | 147 | Returns a tuple consisting of: |
| 148 | - server response code (e.g. '200' if all goes well) |
| 149 | - server response string corresponding to response code |
| 150 | - any RFC822 headers in the response from the server |
Guido van Rossum | 23acc95 | 1994-02-21 16:36:04 +0000 | [diff] [blame] | 151 | |
Guido van Rossum | 41999c1 | 1997-12-09 00:12:23 +0000 | [diff] [blame] | 152 | """ |
| 153 | self.file = self.sock.makefile('rb') |
| 154 | line = self.file.readline() |
| 155 | if self.debuglevel > 0: print 'reply:', `line` |
| 156 | try: |
| 157 | [ver, code, msg] = string.split(line, None, 2) |
| 158 | except ValueError: |
Guido van Rossum | 29c4688 | 1998-01-19 22:25:24 +0000 | [diff] [blame] | 159 | try: |
| 160 | [ver, code] = string.split(line, None, 1) |
| 161 | msg = "" |
| 162 | except ValueError: |
| 163 | self.headers = None |
| 164 | return -1, line, self.headers |
Guido van Rossum | 41999c1 | 1997-12-09 00:12:23 +0000 | [diff] [blame] | 165 | if ver[:5] != 'HTTP/': |
| 166 | self.headers = None |
| 167 | return -1, line, self.headers |
| 168 | errcode = string.atoi(code) |
| 169 | errmsg = string.strip(msg) |
| 170 | self.headers = mimetools.Message(self.file, 0) |
| 171 | return errcode, errmsg, self.headers |
Guido van Rossum | 09c8b6c | 1999-12-07 21:37:17 +0000 | [diff] [blame] | 172 | |
Guido van Rossum | 41999c1 | 1997-12-09 00:12:23 +0000 | [diff] [blame] | 173 | def getfile(self): |
| 174 | """Get a file object from which to receive data from the HTTP server. |
| 175 | |
| 176 | NOTE: This method must not be invoked until getreplies |
| 177 | has been invoked. |
| 178 | |
| 179 | """ |
| 180 | return self.file |
Guido van Rossum | 09c8b6c | 1999-12-07 21:37:17 +0000 | [diff] [blame] | 181 | |
Guido van Rossum | 41999c1 | 1997-12-09 00:12:23 +0000 | [diff] [blame] | 182 | def close(self): |
| 183 | """Close the connection to the HTTP server.""" |
| 184 | if self.file: |
| 185 | self.file.close() |
| 186 | self.file = None |
| 187 | if self.sock: |
| 188 | self.sock.close() |
| 189 | self.sock = None |
Guido van Rossum | 65ab98c | 1995-08-07 20:13:02 +0000 | [diff] [blame] | 190 | |
Guido van Rossum | 09c8b6c | 1999-12-07 21:37:17 +0000 | [diff] [blame] | 191 | if hasattr(socket, "ssl"): |
| 192 | class HTTPS(HTTP): |
| 193 | """This class allows communication via SSL.""" |
| 194 | |
| 195 | def connect(self, host, port = 0): |
| 196 | """Connect to a host on a given port. |
| 197 | |
| 198 | Note: This method is automatically invoked by __init__, |
| 199 | if a host is specified during instantiation. |
| 200 | |
| 201 | """ |
| 202 | if not port: |
| 203 | i = string.find(host, ':') |
| 204 | if i >= 0: |
| 205 | host, port = host[:i], host[i+1:] |
| 206 | try: port = string.atoi(port) |
| 207 | except string.atoi_error: |
| 208 | raise socket.error, "nonnumeric port" |
| 209 | if not port: port = HTTPS_PORT |
| 210 | sock = socket.socket(socket.AF_INET, socket.SOCK_STREAM) |
| 211 | if self.debuglevel > 0: print 'connect:', (host, port) |
| 212 | sock.connect(host, port) |
| 213 | ssl = socket.ssl(sock, self.key_file, self.cert_file) |
| 214 | self.sock = FakeSocket(sock, ssl) |
| 215 | |
Guido van Rossum | 23acc95 | 1994-02-21 16:36:04 +0000 | [diff] [blame] | 216 | |
| 217 | def test(): |
Guido van Rossum | 41999c1 | 1997-12-09 00:12:23 +0000 | [diff] [blame] | 218 | """Test this module. |
| 219 | |
| 220 | The test consists of retrieving and displaying the Python |
| 221 | home page, along with the error code and error string returned |
| 222 | by the www.python.org server. |
| 223 | |
| 224 | """ |
| 225 | import sys |
| 226 | import getopt |
| 227 | opts, args = getopt.getopt(sys.argv[1:], 'd') |
| 228 | dl = 0 |
| 229 | for o, a in opts: |
| 230 | if o == '-d': dl = dl + 1 |
Guido van Rossum | 09c8b6c | 1999-12-07 21:37:17 +0000 | [diff] [blame] | 231 | print "testing HTTP..." |
Guido van Rossum | 41999c1 | 1997-12-09 00:12:23 +0000 | [diff] [blame] | 232 | host = 'www.python.org' |
| 233 | selector = '/' |
| 234 | if args[0:]: host = args[0] |
| 235 | if args[1:]: selector = args[1] |
| 236 | h = HTTP() |
| 237 | h.set_debuglevel(dl) |
| 238 | h.connect(host) |
| 239 | h.putrequest('GET', selector) |
| 240 | h.endheaders() |
| 241 | errcode, errmsg, headers = h.getreply() |
| 242 | print 'errcode =', errcode |
| 243 | print 'errmsg =', errmsg |
| 244 | print |
| 245 | if headers: |
| 246 | for header in headers.headers: print string.strip(header) |
| 247 | print |
| 248 | print h.getfile().read() |
Guido van Rossum | 09c8b6c | 1999-12-07 21:37:17 +0000 | [diff] [blame] | 249 | if hasattr(socket, "ssl"): |
| 250 | print "-"*40 |
| 251 | print "testing HTTPS..." |
| 252 | host = 'synergy.as.cmu.edu' |
| 253 | selector = '/~geek/' |
| 254 | if args[0:]: host = args[0] |
| 255 | if args[1:]: selector = args[1] |
| 256 | h = HTTPS() |
| 257 | h.set_debuglevel(dl) |
| 258 | h.connect(host) |
| 259 | h.putrequest('GET', selector) |
| 260 | h.endheaders() |
| 261 | errcode, errmsg, headers = h.getreply() |
| 262 | print 'errcode =', errcode |
| 263 | print 'errmsg =', errmsg |
| 264 | print |
| 265 | if headers: |
| 266 | for header in headers.headers: print string.strip(header) |
| 267 | print |
| 268 | print h.getfile().read() |
Guido van Rossum | 23acc95 | 1994-02-21 16:36:04 +0000 | [diff] [blame] | 269 | |
Guido van Rossum | a0dfc7a | 1995-09-07 19:28:19 +0000 | [diff] [blame] | 270 | |
Guido van Rossum | 23acc95 | 1994-02-21 16:36:04 +0000 | [diff] [blame] | 271 | if __name__ == '__main__': |
Guido van Rossum | 41999c1 | 1997-12-09 00:12:23 +0000 | [diff] [blame] | 272 | test() |