Gitiles
Code Review Sign In
gerrit-public.fairphone.software / platform / external / python / cpython3 / 2bc930f0442e2bc7ea6d86b74352ad8ac90a40c8 / . / Lib / http / server.py
blob: 7b577b49ac103f3ad2642fea4df0f065d0d71d9b [file] [log] [blame]
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]1"""HTTP server classes.
2
3Note: BaseHTTPRequestHandler doesn't implement any HTTP request; see
4SimpleHTTPRequestHandler for simple implementations of GET, HEAD and POST,
5and CGIHTTPRequestHandler for CGI scripts.
6
7It does, however, optionally implement HTTP/1.1 persistent connections,
8as of version 0.3.
9
10Notes on CGIHTTPRequestHandler
11------------------------------
12
13This class implements GET and POST requests to cgi-bin scripts.
14
15If the os.fork() function is not present (e.g. on Windows),
Amaury Forgeot d'Arccb0d2d72008-06-18 22:19:22 +0000[diff] [blame]16subprocess.Popen() is used as a fallback, with slightly altered semantics.
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]17
18In all cases, the implementation is intentionally naive -- all
19requests are executed synchronously.
20
21SECURITY WARNING: DON'T USE THIS CODE UNLESS YOU ARE INSIDE A FIREWALL
22-- it may execute arbitrary Python code or external programs.
23
24Note that status code 200 is sent prior to execution of a CGI script, so
25scripts cannot send other status codes such as 302 (redirect).
26
27XXX To do:
28
29- log requests even later (to capture byte count)
30- log user-agent header and other interesting goodies
31- send error log to separate file
32"""
33
34
35# See also:
36#
37# HTTP Working Group T. Berners-Lee
38# INTERNET-DRAFT R. T. Fielding
39# <draft-ietf-http-v10-spec-00.txt> H. Frystyk Nielsen
40# Expires September 8, 1995 March 8, 1995
41#
42# URL: http://www.ics.uci.edu/pub/ietf/http/draft-ietf-http-v10-spec-00.txt
43#
44# and
45#
46# Network Working Group R. Fielding
47# Request for Comments: 2616 et al
48# Obsoletes: 2068 June 1999
49# Category: Standards Track
50#
51# URL: http://www.faqs.org/rfcs/rfc2616.html
52
53# Log files
54# ---------
55#
56# Here's a quote from the NCSA httpd docs about log file format.
57#
58# | The logfile format is as follows. Each line consists of:
59# |
60# | host rfc931 authuser [DD/Mon/YYYY:hh:mm:ss] "request" ddd bbbb
61# |
62# | host: Either the DNS name or the IP number of the remote client
63# | rfc931: Any information returned by identd for this person,
64# | - otherwise.
65# | authuser: If user sent a userid for authentication, the user name,
66# | - otherwise.
67# | DD: Day
68# | Mon: Month (calendar name)
69# | YYYY: Year
70# | hh: hour (24-hour format, the machine's timezone)
71# | mm: minutes
72# | ss: seconds
73# | request: The first line of the HTTP request as sent by the client.
74# | ddd: the status code returned by the server, - if not available.
75# | bbbb: the total number of bytes sent,
76# | *not including the HTTP/1.0 header*, - if not available
77# |
78# | You can determine the name of the file accessed through request.
79#
80# (Actually, the latter is only true if you know the server configuration
81# at the time the request was made!)
82
83__version__ = "0.6"
84
85__all__ = ["HTTPServer", "BaseHTTPRequestHandler"]
86
Georg Brandl1f7fffb2010-10-15 15:57:45 +0000[diff] [blame]87import html
Barry Warsaw820c1202008-06-12 04:06:45 +0000[diff] [blame]88import email.message
89import email.parser
Jeremy Hylton914ab452009-03-27 17:16:06 +0000[diff] [blame]90import http.client
91import io
92import mimetypes
93import os
94import posixpath
95import select
96import shutil
97import socket # For gethostbyaddr()
98import socketserver
99import sys
100import time
101import urllib.parse
Senthil Kumaran42713722010-10-03 17:55:45 +0000[diff] [blame]102import copy
Senthil Kumaran1251faf2012-06-03 16:15:54 +0800[diff] [blame]103import argparse
104
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]105
106# Default error message template
107DEFAULT_ERROR_MESSAGE = """\
Senthil Kumaran1b407fe2011-03-20 10:44:30 +0800[diff] [blame]108<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01//EN"
109 "http://www.w3.org/TR/html4/strict.dtd">
Ezio Melottica897e92011-11-02 19:33:29 +0200[diff] [blame]110<html>
Senthil Kumaranb253c9f2011-03-17 16:43:22 +0800[diff] [blame]111 <head>
Senthil Kumaran1b407fe2011-03-20 10:44:30 +0800[diff] [blame]112 <meta http-equiv="Content-Type" content="text/html;charset=utf-8">
Senthil Kumaranb253c9f2011-03-17 16:43:22 +0800[diff] [blame]113 <title>Error response</title>
114 </head>
115 <body>
116 <h1>Error response</h1>
117 <p>Error code: %(code)d</p>
118 <p>Message: %(message)s.</p>
119 <p>Error code explanation: %(code)s - %(explain)s.</p>
120 </body>
121</html>
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]122"""
123
124DEFAULT_ERROR_CONTENT_TYPE = "text/html;charset=utf-8"
125
126def _quote_html(html):
127 return html.replace("&", "&amp;").replace("<", "&lt;").replace(">", "&gt;")
128
129class HTTPServer(socketserver.TCPServer):
130
131 allow_reuse_address = 1 # Seems to make sense in testing environment
132
133 def server_bind(self):
134 """Override server_bind to store the server name."""
135 socketserver.TCPServer.server_bind(self)
136 host, port = self.socket.getsockname()[:2]
137 self.server_name = socket.getfqdn(host)
138 self.server_port = port
139
140
141class BaseHTTPRequestHandler(socketserver.StreamRequestHandler):
142
143 """HTTP request handler base class.
144
145 The following explanation of HTTP serves to guide you through the
146 code as well as to expose any misunderstandings I may have about
147 HTTP (so you don't need to read the code to figure out I'm wrong
148 :-).
149
150 HTTP (HyperText Transfer Protocol) is an extensible protocol on
151 top of a reliable stream transport (e.g. TCP/IP). The protocol
152 recognizes three parts to a request:
153
154 1. One line identifying the request type and path
155 2. An optional set of RFC-822-style headers
156 3. An optional data part
157
158 The headers and data are separated by a blank line.
159
160 The first line of the request has the form
161
162 <command> <path> <version>
163
164 where <command> is a (case-sensitive) keyword such as GET or POST,
165 <path> is a string containing path information for the request,
166 and <version> should be the string "HTTP/1.0" or "HTTP/1.1".
167 <path> is encoded using the URL encoding scheme (using %xx to signify
168 the ASCII character with hex code xx).
169
170 The specification specifies that lines are separated by CRLF but
171 for compatibility with the widest range of clients recommends
172 servers also handle LF. Similarly, whitespace in the request line
173 is treated sensibly (allowing multiple spaces between components
174 and allowing trailing whitespace).
175
176 Similarly, for output, lines ought to be separated by CRLF pairs
177 but most clients grok LF characters just fine.
178
179 If the first line of the request has the form
180
181 <command> <path>
182
183 (i.e. <version> is left out) then this is assumed to be an HTTP
184 0.9 request; this form has no optional headers and data part and
185 the reply consists of just the data.
186
187 The reply form of the HTTP 1.x protocol again has three parts:
188
189 1. One line giving the response code
190 2. An optional set of RFC-822-style headers
191 3. The data
192
193 Again, the headers and data are separated by a blank line.
194
195 The response code line has the form
196
197 <version> <responsecode> <responsestring>
198
199 where <version> is the protocol version ("HTTP/1.0" or "HTTP/1.1"),
200 <responsecode> is a 3-digit response code indicating success or
201 failure of the request, and <responsestring> is an optional
202 human-readable string explaining what the response code means.
203
204 This server parses the request and the headers, and then calls a
205 function specific to the request type (<command>). Specifically,
206 a request SPAM will be handled by a method do_SPAM(). If no
207 such method exists the server sends an error response to the
208 client. If it exists, it is called with no arguments:
209
210 do_SPAM()
211
212 Note that the request name is case sensitive (i.e. SPAM and spam
213 are different requests).
214
215 The various request details are stored in instance variables:
216
217 - client_address is the client IP address in the form (host,
218 port);
219
220 - command, path and version are the broken-down request line;
221
Barry Warsaw820c1202008-06-12 04:06:45 +0000[diff] [blame]222 - headers is an instance of email.message.Message (or a derived
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]223 class) containing the header information;
224
225 - rfile is a file object open for reading positioned at the
226 start of the optional input data part;
227
228 - wfile is a file object open for writing.
229
230 IT IS IMPORTANT TO ADHERE TO THE PROTOCOL FOR WRITING!
231
232 The first thing to be written must be the response line. Then
233 follow 0 or more header lines, then a blank line, and then the
234 actual data (if any). The meaning of the header lines depends on
235 the command executed by the server; in most cases, when data is
236 returned, there should be at least one header line of the form
237
238 Content-type: <type>/<subtype>
239
240 where <type> and <subtype> should be registered MIME types,
241 e.g. "text/html" or "text/plain".
242
243 """
244
245 # The Python system version, truncated to its first component.
246 sys_version = "Python/" + sys.version.split()[0]
247
248 # The server software version. You may want to override this.
249 # The format is multiple whitespace-separated strings,
250 # where each string is of the form name[/version].
251 server_version = "BaseHTTP/" + __version__
252
253 error_message_format = DEFAULT_ERROR_MESSAGE
254 error_content_type = DEFAULT_ERROR_CONTENT_TYPE
255
256 # The default request version. This only affects responses up until
257 # the point where the request line is parsed, so it mainly decides what
258 # the client gets back when sending a malformed request line.
259 # Most web servers default to HTTP 0.9, i.e. don't send a status line.
260 default_request_version = "HTTP/0.9"
261
262 def parse_request(self):
263 """Parse a request (internal).
264
265 The request should be stored in self.raw_requestline; the results
266 are in self.command, self.path, self.request_version and
267 self.headers.
268
269 Return True for success, False for failure; on failure, an
270 error is sent back.
271
272 """
273 self.command = None # set in case of error on the first line
274 self.request_version = version = self.default_request_version
275 self.close_connection = 1
276 requestline = str(self.raw_requestline, 'iso-8859-1')
Senthil Kumaran30755492011-12-23 17:03:41 +0800[diff] [blame]277 requestline = requestline.rstrip('\r\n')
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]278 self.requestline = requestline
279 words = requestline.split()
280 if len(words) == 3:
Senthil Kumaran30755492011-12-23 17:03:41 +0800[diff] [blame]281 command, path, version = words
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]282 if version[:5] != 'HTTP/':
283 self.send_error(400, "Bad request version (%r)" % version)
284 return False
285 try:
286 base_version_number = version.split('/', 1)[1]
287 version_number = base_version_number.split(".")
288 # RFC 2145 section 3.1 says there can be only one "." and
289 # - major and minor numbers MUST be treated as
290 # separate integers;
291 # - HTTP/2.4 is a lower version than HTTP/2.13, which in
292 # turn is lower than HTTP/12.3;
293 # - Leading zeros MUST be ignored by recipients.
294 if len(version_number) != 2:
295 raise ValueError
296 version_number = int(version_number[0]), int(version_number[1])
297 except (ValueError, IndexError):
298 self.send_error(400, "Bad request version (%r)" % version)
299 return False
300 if version_number >= (1, 1) and self.protocol_version >= "HTTP/1.1":
301 self.close_connection = 0
302 if version_number >= (2, 0):
303 self.send_error(505,
304 "Invalid HTTP Version (%s)" % base_version_number)
305 return False
306 elif len(words) == 2:
Senthil Kumaran30755492011-12-23 17:03:41 +0800[diff] [blame]307 command, path = words
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]308 self.close_connection = 1
309 if command != 'GET':
310 self.send_error(400,
311 "Bad HTTP/0.9 request type (%r)" % command)
312 return False
313 elif not words:
314 return False
315 else:
316 self.send_error(400, "Bad request syntax (%r)" % requestline)
317 return False
318 self.command, self.path, self.request_version = command, path, version
319
320 # Examine the headers and look for a Connection directive.
Senthil Kumaran5466bf12010-12-18 16:55:23 +0000[diff] [blame]321 try:
322 self.headers = http.client.parse_headers(self.rfile,
323 _class=self.MessageClass)
324 except http.client.LineTooLong:
325 self.send_error(400, "Line too long")
326 return False
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]327
328 conntype = self.headers.get('Connection', "")
329 if conntype.lower() == 'close':
330 self.close_connection = 1
331 elif (conntype.lower() == 'keep-alive' and
332 self.protocol_version >= "HTTP/1.1"):
333 self.close_connection = 0
Senthil Kumaran0f476d42010-09-30 06:09:18 +0000[diff] [blame]334 # Examine the headers and look for an Expect directive
335 expect = self.headers.get('Expect', "")
336 if (expect.lower() == "100-continue" and
337 self.protocol_version >= "HTTP/1.1" and
338 self.request_version >= "HTTP/1.1"):
339 if not self.handle_expect_100():
340 return False
341 return True
342
343 def handle_expect_100(self):
344 """Decide what to do with an "Expect: 100-continue" header.
345
346 If the client is expecting a 100 Continue response, we must
347 respond with either a 100 Continue or a final response before
348 waiting for the request body. The default is to always respond
349 with a 100 Continue. You can behave differently (for example,
350 reject unauthorized requests) by overriding this method.
351
352 This method should either return True (possibly after sending
353 a 100 Continue response) or send an error response and return
354 False.
355
356 """
357 self.send_response_only(100)
Senthil Kumaranc7ae19b2011-05-09 23:25:02 +0800[diff] [blame]358 self.flush_headers()
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]359 return True
360
361 def handle_one_request(self):
362 """Handle a single HTTP request.
363
364 You normally don't need to override this method; see the class
365 __doc__ string for information on how to handle specific HTTP
366 commands such as GET and POST.
367
368 """
Kristján Valur Jónsson985fc6a2009-07-01 10:01:31 +0000[diff] [blame]369 try:
Antoine Pitrouc4924372010-12-16 16:48:36 +0000[diff] [blame]370 self.raw_requestline = self.rfile.readline(65537)
371 if len(self.raw_requestline) > 65536:
372 self.requestline = ''
373 self.request_version = ''
374 self.command = ''
375 self.send_error(414)
376 return
Kristján Valur Jónsson985fc6a2009-07-01 10:01:31 +0000[diff] [blame]377 if not self.raw_requestline:
378 self.close_connection = 1
379 return
380 if not self.parse_request():
381 # An error code has been sent, just exit
382 return
383 mname = 'do_' + self.command
384 if not hasattr(self, mname):
385 self.send_error(501, "Unsupported method (%r)" % self.command)
386 return
387 method = getattr(self, mname)
388 method()
389 self.wfile.flush() #actually send the response if not already done.
390 except socket.timeout as e:
391 #a read or a write timed out. Discard this connection
392 self.log_error("Request timed out: %r", e)
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]393 self.close_connection = 1
394 return
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]395
396 def handle(self):
397 """Handle multiple requests if necessary."""
398 self.close_connection = 1
399
400 self.handle_one_request()
401 while not self.close_connection:
402 self.handle_one_request()
403
Senthil Kumaran26886442013-03-15 07:53:21 -0700[diff] [blame]404 def send_error(self, code, message=None, explain=None):
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]405 """Send and log an error reply.
406
Senthil Kumaran26886442013-03-15 07:53:21 -0700[diff] [blame]407 Arguments are
408 * code: an HTTP error code
409 3 digits
410 * message: a simple optional 1 line reason phrase.
411 *( HTAB / SP / VCHAR / %x80-FF )
412 defaults to short entry matching the response code
413 * explain: a detailed message defaults to the long entry
414 matching the response code.
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]415
416 This sends an error response (so it must be called before any
417 output has been generated), logs the error, and finally sends
418 a piece of HTML explaining the error to the user.
419
420 """
421
422 try:
423 shortmsg, longmsg = self.responses[code]
424 except KeyError:
425 shortmsg, longmsg = '???', '???'
426 if message is None:
427 message = shortmsg
Senthil Kumaran26886442013-03-15 07:53:21 -0700[diff] [blame]428 if explain is None:
429 explain = longmsg
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]430 self.log_error("code %d, message %s", code, message)
431 # using _quote_html to prevent Cross Site Scripting attacks (see bug #1100201)
432 content = (self.error_message_format %
Senthil Kumaran26886442013-03-15 07:53:21 -0700[diff] [blame]433 {'code': code, 'message': _quote_html(message), 'explain': _quote_html(explain)})
Senthil Kumaran52d27202012-10-10 23:16:21 -0700[diff] [blame]434 body = content.encode('UTF-8', 'replace')
Senthil Kumaran1e7551d2013-03-05 02:25:58 -0800[diff] [blame]435 self.send_response(code, message)
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]436 self.send_header("Content-Type", self.error_content_type)
437 self.send_header('Connection', 'close')
Senthil Kumaran52d27202012-10-10 23:16:21 -0700[diff] [blame]438 self.send_header('Content-Length', int(len(body)))
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]439 self.end_headers()
440 if self.command != 'HEAD' and code >= 200 and code not in (204, 304):
Senthil Kumaran52d27202012-10-10 23:16:21 -0700[diff] [blame]441 self.wfile.write(body)
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]442
443 def send_response(self, code, message=None):
Senthil Kumaranc7ae19b2011-05-09 23:25:02 +0800[diff] [blame]444 """Add the response header to the headers buffer and log the
445 response code.
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]446
447 Also send two standard headers with the server software
448 version and the current date.
449
450 """
451 self.log_request(code)
Senthil Kumaran0f476d42010-09-30 06:09:18 +0000[diff] [blame]452 self.send_response_only(code, message)
453 self.send_header('Server', self.version_string())
454 self.send_header('Date', self.date_time_string())
455
456 def send_response_only(self, code, message=None):
457 """Send the response header only."""
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]458 if message is None:
459 if code in self.responses:
460 message = self.responses[code][0]
461 else:
462 message = ''
463 if self.request_version != 'HTTP/0.9':
Senthil Kumaranc7ae19b2011-05-09 23:25:02 +0800[diff] [blame]464 if not hasattr(self, '_headers_buffer'):
465 self._headers_buffer = []
466 self._headers_buffer.append(("%s %d %s\r\n" %
467 (self.protocol_version, code, message)).encode(
468 'latin-1', 'strict'))
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]469
470 def send_header(self, keyword, value):
Senthil Kumaranc7ae19b2011-05-09 23:25:02 +0800[diff] [blame]471 """Send a MIME header to the headers buffer."""
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]472 if self.request_version != 'HTTP/0.9':
Senthil Kumarane4dad4f2010-11-21 14:36:14 +0000[diff] [blame]473 if not hasattr(self, '_headers_buffer'):
474 self._headers_buffer = []
475 self._headers_buffer.append(
Marc-André Lemburg8f36af72011-02-25 15:42:01 +0000[diff] [blame]476 ("%s: %s\r\n" % (keyword, value)).encode('latin-1', 'strict'))
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]477
478 if keyword.lower() == 'connection':
479 if value.lower() == 'close':
480 self.close_connection = 1
481 elif value.lower() == 'keep-alive':
482 self.close_connection = 0
483
484 def end_headers(self):
485 """Send the blank line ending the MIME headers."""
486 if self.request_version != 'HTTP/0.9':
Senthil Kumarane4dad4f2010-11-21 14:36:14 +0000[diff] [blame]487 self._headers_buffer.append(b"\r\n")
Senthil Kumaranc7ae19b2011-05-09 23:25:02 +0800[diff] [blame]488 self.flush_headers()
489
490 def flush_headers(self):
491 if hasattr(self, '_headers_buffer'):
Senthil Kumarane4dad4f2010-11-21 14:36:14 +0000[diff] [blame]492 self.wfile.write(b"".join(self._headers_buffer))
493 self._headers_buffer = []
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]494
495 def log_request(self, code='-', size='-'):
496 """Log an accepted request.
497
498 This is called by send_response().
499
500 """
501
502 self.log_message('"%s" %s %s',
503 self.requestline, str(code), str(size))
504
505 def log_error(self, format, *args):
506 """Log an error.
507
508 This is called when a request cannot be fulfilled. By
509 default it passes the message on to log_message().
510
511 Arguments are the same as for log_message().
512
513 XXX This should go to the separate error log.
514
515 """
516
517 self.log_message(format, *args)
518
519 def log_message(self, format, *args):
520 """Log an arbitrary message.
521
522 This is used by all other logging functions. Override
523 it if you have specific logging wishes.
524
525 The first argument, FORMAT, is a format string for the
526 message to be logged. If the format string contains
527 any % escapes requiring parameters, they should be
528 specified as subsequent arguments (it's just like
529 printf!).
530
Senthil Kumarandb727b42012-04-29 13:41:03 +0800[diff] [blame]531 The client ip and current date/time are prefixed to
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]532 every message.
533
534 """
535
536 sys.stderr.write("%s - - [%s] %s\n" %
537 (self.address_string(),
538 self.log_date_time_string(),
539 format%args))
540
541 def version_string(self):
542 """Return the server software version string."""
543 return self.server_version + ' ' + self.sys_version
544
545 def date_time_string(self, timestamp=None):
546 """Return the current date and time formatted for a message header."""
547 if timestamp is None:
548 timestamp = time.time()
549 year, month, day, hh, mm, ss, wd, y, z = time.gmtime(timestamp)
550 s = "%s, %02d %3s %4d %02d:%02d:%02d GMT" % (
551 self.weekdayname[wd],
552 day, self.monthname[month], year,
553 hh, mm, ss)
554 return s
555
556 def log_date_time_string(self):
557 """Return the current time formatted for logging."""
558 now = time.time()
559 year, month, day, hh, mm, ss, x, y, z = time.localtime(now)
560 s = "%02d/%3s/%04d %02d:%02d:%02d" % (
561 day, self.monthname[month], year, hh, mm, ss)
562 return s
563
564 weekdayname = ['Mon', 'Tue', 'Wed', 'Thu', 'Fri', 'Sat', 'Sun']
565
566 monthname = [None,
567 'Jan', 'Feb', 'Mar', 'Apr', 'May', 'Jun',
568 'Jul', 'Aug', 'Sep', 'Oct', 'Nov', 'Dec']
569
570 def address_string(self):
Senthil Kumaran1aacba42012-04-29 12:51:54 +0800[diff] [blame]571 """Return the client address."""
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]572
Senthil Kumaran1aacba42012-04-29 12:51:54 +0800[diff] [blame]573 return self.client_address[0]
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]574
575 # Essentially static class variables
576
577 # The version of the HTTP protocol we support.
578 # Set this to HTTP/1.1 to enable automatic keepalive
579 protocol_version = "HTTP/1.0"
580
Barry Warsaw820c1202008-06-12 04:06:45 +0000[diff] [blame]581 # MessageClass used to parse headers
Barry Warsaw820c1202008-06-12 04:06:45 +0000[diff] [blame]582 MessageClass = http.client.HTTPMessage
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]583
584 # Table mapping response codes to messages; entries have the
585 # form {code: (shortmessage, longmessage)}.
Hynek Schlawack51b2ed52012-05-16 09:51:07 +0200[diff] [blame]586 # See RFC 2616 and 6585.
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]587 responses = {
588 100: ('Continue', 'Request received, please continue'),
589 101: ('Switching Protocols',
590 'Switching to new protocol; obey Upgrade header'),
591
592 200: ('OK', 'Request fulfilled, document follows'),
593 201: ('Created', 'Document created, URL follows'),
594 202: ('Accepted',
595 'Request accepted, processing continues off-line'),
596 203: ('Non-Authoritative Information', 'Request fulfilled from cache'),
597 204: ('No Content', 'Request fulfilled, nothing follows'),
598 205: ('Reset Content', 'Clear input form for further input.'),
599 206: ('Partial Content', 'Partial content follows.'),
600
601 300: ('Multiple Choices',
602 'Object has several resources -- see URI list'),
603 301: ('Moved Permanently', 'Object moved permanently -- see URI list'),
604 302: ('Found', 'Object moved temporarily -- see URI list'),
605 303: ('See Other', 'Object moved -- see Method and URL list'),
606 304: ('Not Modified',
607 'Document has not changed since given time'),
608 305: ('Use Proxy',
609 'You must use proxy specified in Location to access this '
610 'resource.'),
611 307: ('Temporary Redirect',
612 'Object moved temporarily -- see URI list'),
613
614 400: ('Bad Request',
615 'Bad request syntax or unsupported method'),
616 401: ('Unauthorized',
617 'No permission -- see authorization schemes'),
618 402: ('Payment Required',
619 'No payment -- see charging schemes'),
620 403: ('Forbidden',
621 'Request forbidden -- authorization will not help'),
622 404: ('Not Found', 'Nothing matches the given URI'),
623 405: ('Method Not Allowed',
Senthil Kumaran7aa26212010-02-22 11:00:50 +0000[diff] [blame]624 'Specified method is invalid for this resource.'),
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]625 406: ('Not Acceptable', 'URI not available in preferred format.'),
626 407: ('Proxy Authentication Required', 'You must authenticate with '
627 'this proxy before proceeding.'),
628 408: ('Request Timeout', 'Request timed out; try again later.'),
629 409: ('Conflict', 'Request conflict.'),
630 410: ('Gone',
631 'URI no longer exists and has been permanently removed.'),
632 411: ('Length Required', 'Client must specify Content-Length.'),
633 412: ('Precondition Failed', 'Precondition in headers is false.'),
634 413: ('Request Entity Too Large', 'Entity is too large.'),
635 414: ('Request-URI Too Long', 'URI is too long.'),
636 415: ('Unsupported Media Type', 'Entity body in unsupported format.'),
637 416: ('Requested Range Not Satisfiable',
638 'Cannot satisfy request range.'),
639 417: ('Expectation Failed',
640 'Expect condition could not be satisfied.'),
Hynek Schlawack51b2ed52012-05-16 09:51:07 +0200[diff] [blame]641 428: ('Precondition Required',
642 'The origin server requires the request to be conditional.'),
643 429: ('Too Many Requests', 'The user has sent too many requests '
644 'in a given amount of time ("rate limiting").'),
645 431: ('Request Header Fields Too Large', 'The server is unwilling to '
646 'process the request because its header fields are too large.'),
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]647
648 500: ('Internal Server Error', 'Server got itself in trouble'),
649 501: ('Not Implemented',
650 'Server does not support this operation'),
651 502: ('Bad Gateway', 'Invalid responses from another server/proxy.'),
652 503: ('Service Unavailable',
653 'The server cannot process the request due to a high load'),
654 504: ('Gateway Timeout',
655 'The gateway server did not receive a timely response'),
656 505: ('HTTP Version Not Supported', 'Cannot fulfill request.'),
Hynek Schlawack51b2ed52012-05-16 09:51:07 +0200[diff] [blame]657 511: ('Network Authentication Required',
658 'The client needs to authenticate to gain network access.'),
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]659 }
660
661
662class SimpleHTTPRequestHandler(BaseHTTPRequestHandler):
663
664 """Simple HTTP request handler with GET and HEAD commands.
665
666 This serves files from the current directory and any of its
667 subdirectories. The MIME type for files is determined by
668 calling the .guess_type() method.
669
670 The GET and HEAD requests are identical except that the HEAD
671 request omits the actual contents of the file.
672
673 """
674
675 server_version = "SimpleHTTP/" + __version__
676
677 def do_GET(self):
678 """Serve a GET request."""
679 f = self.send_head()
680 if f:
681 self.copyfile(f, self.wfile)
682 f.close()
683
684 def do_HEAD(self):
685 """Serve a HEAD request."""
686 f = self.send_head()
687 if f:
688 f.close()
689
690 def send_head(self):
691 """Common code for GET and HEAD commands.
692
693 This sends the response code and MIME headers.
694
695 Return value is either a file object (which has to be copied
696 to the outputfile by the caller unless the command was HEAD,
697 and must be closed by the caller under all circumstances), or
698 None, in which case the caller has nothing further to do.
699
700 """
701 path = self.translate_path(self.path)
702 f = None
703 if os.path.isdir(path):
704 if not self.path.endswith('/'):
705 # redirect browser - doing basically what apache does
706 self.send_response(301)
707 self.send_header("Location", self.path + "/")
708 self.end_headers()
709 return None
710 for index in "index.html", "index.htm":
711 index = os.path.join(path, index)
712 if os.path.exists(index):
713 path = index
714 break
715 else:
716 return self.list_directory(path)
717 ctype = self.guess_type(path)
718 try:
719 f = open(path, 'rb')
Andrew Svetlovf7a17b42012-12-25 16:47:37 +0200[diff] [blame]720 except OSError:
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]721 self.send_error(404, "File not found")
722 return None
723 self.send_response(200)
724 self.send_header("Content-type", ctype)
725 fs = os.fstat(f.fileno())
726 self.send_header("Content-Length", str(fs[6]))
727 self.send_header("Last-Modified", self.date_time_string(fs.st_mtime))
728 self.end_headers()
729 return f
730
731 def list_directory(self, path):
732 """Helper to produce a directory listing (absent index.html).
733
734 Return value is either a file object, or None (indicating an
735 error). In either case, the headers are sent, making the
736 interface the same as for send_head().
737
738 """
739 try:
740 list = os.listdir(path)
Andrew Svetlovad28c7f2012-12-18 22:02:39 +0200[diff] [blame]741 except OSError:
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]742 self.send_error(404, "No permission to list directory")
743 return None
744 list.sort(key=lambda a: a.lower())
745 r = []
Georg Brandl1f7fffb2010-10-15 15:57:45 +0000[diff] [blame]746 displaypath = html.escape(urllib.parse.unquote(self.path))
Ezio Melottica897e92011-11-02 19:33:29 +0200[diff] [blame]747 enc = sys.getfilesystemencoding()
748 title = 'Directory listing for %s' % displaypath
749 r.append('<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01//EN" '
750 '"http://www.w3.org/TR/html4/strict.dtd">')
751 r.append('<html>\n<head>')
752 r.append('<meta http-equiv="Content-Type" '
753 'content="text/html; charset=%s">' % enc)
754 r.append('<title>%s</title>\n</head>' % title)
755 r.append('<body>\n<h1>%s</h1>' % title)
756 r.append('<hr>\n<ul>')
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]757 for name in list:
758 fullname = os.path.join(path, name)
759 displayname = linkname = name
760 # Append / for directories or @ for symbolic links
761 if os.path.isdir(fullname):
762 displayname = name + "/"
763 linkname = name + "/"
764 if os.path.islink(fullname):
765 displayname = name + "@"
766 # Note: a link to a directory displays with @ and links with /
Ezio Melottica897e92011-11-02 19:33:29 +0200[diff] [blame]767 r.append('<li><a href="%s">%s</a></li>'
Georg Brandl1f7fffb2010-10-15 15:57:45 +0000[diff] [blame]768 % (urllib.parse.quote(linkname), html.escape(displayname)))
Ezio Melottica897e92011-11-02 19:33:29 +0200[diff] [blame]769 r.append('</ul>\n<hr>\n</body>\n</html>\n')
770 encoded = '\n'.join(r).encode(enc)
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]771 f = io.BytesIO()
772 f.write(encoded)
773 f.seek(0)
774 self.send_response(200)
775 self.send_header("Content-type", "text/html; charset=%s" % enc)
776 self.send_header("Content-Length", str(len(encoded)))
777 self.end_headers()
778 return f
779
780 def translate_path(self, path):
781 """Translate a /-separated PATH to the local filename syntax.
782
783 Components that mean special things to the local file system
784 (e.g. drive or directory names) are ignored. (XXX They should
785 probably be diagnosed.)
786
787 """
788 # abandon query parameters
789 path = path.split('?',1)[0]
790 path = path.split('#',1)[0]
Senthil Kumaran72c238e2013-09-13 00:21:18 -0700[diff] [blame]791 # Don't forget explicit trailing slash when normalizing. Issue17324
Senthil Kumaran600b7352013-09-29 18:59:04 -0700[diff] [blame]792 trailing_slash = path.rstrip().endswith('/')
Jeremy Hylton1afc1692008-06-18 20:49:58 +0000[diff] [blame]793 path = posixpath.normpath(urllib.parse.unquote(path))
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]794 words = path.split('/')
795 words = filter(None, words)
796 path = os.getcwd()
797 for word in words:
798 drive, word = os.path.splitdrive(word)
799 head, word = os.path.split(word)
800 if word in (os.curdir, os.pardir): continue
801 path = os.path.join(path, word)
Senthil Kumaran72c238e2013-09-13 00:21:18 -0700[diff] [blame]802 if trailing_slash:
803 path += '/'
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]804 return path
805
806 def copyfile(self, source, outputfile):
807 """Copy all data between two file objects.
808
809 The SOURCE argument is a file object open for reading
810 (or anything with a read() method) and the DESTINATION
811 argument is a file object open for writing (or
812 anything with a write() method).
813
814 The only reason for overriding this would be to change
815 the block size or perhaps to replace newlines by CRLF
816 -- note however that this the default server uses this
817 to copy binary data as well.
818
819 """
820 shutil.copyfileobj(source, outputfile)
821
822 def guess_type(self, path):
823 """Guess the type of a file.
824
825 Argument is a PATH (a filename).
826
827 Return value is a string of the form type/subtype,
828 usable for a MIME Content-type header.
829
830 The default implementation looks the file's extension
831 up in the table self.extensions_map, using application/octet-stream
832 as a default; however it would be permissible (if
833 slow) to look inside the data to make a better guess.
834
835 """
836
837 base, ext = posixpath.splitext(path)
838 if ext in self.extensions_map:
839 return self.extensions_map[ext]
840 ext = ext.lower()
841 if ext in self.extensions_map:
842 return self.extensions_map[ext]
843 else:
844 return self.extensions_map['']
845
846 if not mimetypes.inited:
847 mimetypes.init() # try to read system mime.types
848 extensions_map = mimetypes.types_map.copy()
849 extensions_map.update({
850 '': 'application/octet-stream', # Default
851 '.py': 'text/plain',
852 '.c': 'text/plain',
853 '.h': 'text/plain',
854 })
855
856
857# Utilities for CGIHTTPRequestHandler
858
Senthil Kumarand70846b2012-04-12 02:34:32 +0800[diff] [blame]859def _url_collapse_path(path):
Benjamin Petersonad71f0f2009-04-11 20:12:10 +0000[diff] [blame]860 """
861 Given a URL path, remove extra '/'s and '.' path elements and collapse
Senthil Kumarand70846b2012-04-12 02:34:32 +0800[diff] [blame]862 any '..' references and returns a colllapsed path.
Benjamin Petersonad71f0f2009-04-11 20:12:10 +0000[diff] [blame]863
864 Implements something akin to RFC-2396 5.2 step 6 to parse relative paths.
Senthil Kumarand70846b2012-04-12 02:34:32 +0800[diff] [blame]865 The utility of this function is limited to is_cgi method and helps
866 preventing some security attacks.
Benjamin Petersonad71f0f2009-04-11 20:12:10 +0000[diff] [blame]867
868 Returns: A tuple of (head, tail) where tail is everything after the final /
869 and head is everything before it. Head will always start with a '/' and,
870 if it contains anything else, never have a trailing '/'.
871
872 Raises: IndexError if too many '..' occur within the path.
Senthil Kumarand70846b2012-04-12 02:34:32 +0800[diff] [blame]873
Benjamin Petersonad71f0f2009-04-11 20:12:10 +0000[diff] [blame]874 """
875 # Similar to os.path.split(os.path.normpath(path)) but specific to URL
876 # path semantics rather than local operating system semantics.
Senthil Kumarand70846b2012-04-12 02:34:32 +0800[diff] [blame]877 path_parts = path.split('/')
878 head_parts = []
879 for part in path_parts[:-1]:
880 if part == '..':
881 head_parts.pop() # IndexError if more '..' than prior parts
882 elif part and part != '.':
883 head_parts.append( part )
Benjamin Petersonad71f0f2009-04-11 20:12:10 +0000[diff] [blame]884 if path_parts:
Senthil Kumarandbb369d2012-04-11 03:15:28 +0800[diff] [blame]885 tail_part = path_parts.pop()
Senthil Kumarand70846b2012-04-12 02:34:32 +0800[diff] [blame]886 if tail_part:
887 if tail_part == '..':
888 head_parts.pop()
889 tail_part = ''
890 elif tail_part == '.':
891 tail_part = ''
Benjamin Petersonad71f0f2009-04-11 20:12:10 +0000[diff] [blame]892 else:
893 tail_part = ''
Senthil Kumarand70846b2012-04-12 02:34:32 +0800[diff] [blame]894
895 splitpath = ('/' + '/'.join(head_parts), tail_part)
896 collapsed_path = "/".join(splitpath)
897
898 return collapsed_path
899
Benjamin Petersonad71f0f2009-04-11 20:12:10 +0000[diff] [blame]900
901
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]902nobody = None
903
904def nobody_uid():
905 """Internal routine to get nobody's uid"""
906 global nobody
907 if nobody:
908 return nobody
909 try:
910 import pwd
Brett Cannoncd171c82013-07-04 17:43:24 -0400[diff] [blame]911 except ImportError:
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]912 return -1
913 try:
914 nobody = pwd.getpwnam('nobody')[2]
915 except KeyError:
Georg Brandlcbd2ab12010-12-04 10:39:14 +0000[diff] [blame]916 nobody = 1 + max(x[2] for x in pwd.getpwall())
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]917 return nobody
918
919
920def executable(path):
921 """Test for executable file."""
Victor Stinnerfb25ba92011-06-20 17:45:54 +0200[diff] [blame]922 return os.access(path, os.X_OK)
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]923
924
925class CGIHTTPRequestHandler(SimpleHTTPRequestHandler):
926
927 """Complete HTTP server with GET, HEAD and POST commands.
928
929 GET and HEAD also support running CGI scripts.
930
931 The POST command is *only* implemented for CGI scripts.
932
933 """
934
935 # Determine platform specifics
936 have_fork = hasattr(os, 'fork')
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]937
938 # Make rfile unbuffered -- we need to read one line and then pass
939 # the rest to a subprocess, so we can't use buffered input.
940 rbufsize = 0
941
942 def do_POST(self):
943 """Serve a POST request.
944
945 This is only implemented for CGI scripts.
946
947 """
948
949 if self.is_cgi():
950 self.run_cgi()
951 else:
952 self.send_error(501, "Can only POST to CGI scripts")
953
954 def send_head(self):
955 """Version of send_head that support CGI scripts"""
956 if self.is_cgi():
957 return self.run_cgi()
958 else:
959 return SimpleHTTPRequestHandler.send_head(self)
960
961 def is_cgi(self):
962 """Test whether self.path corresponds to a CGI script.
963
Benjamin Petersonad71f0f2009-04-11 20:12:10 +0000[diff] [blame]964 Returns True and updates the cgi_info attribute to the tuple
965 (dir, rest) if self.path requires running a CGI script.
966 Returns False otherwise.
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]967
Benjamin Petersona7deeee2009-05-08 20:54:42 +0000[diff] [blame]968 If any exception is raised, the caller should assume that
969 self.path was rejected as invalid and act accordingly.
970
Benjamin Petersonad71f0f2009-04-11 20:12:10 +0000[diff] [blame]971 The default implementation tests whether the normalized url
972 path begins with one of the strings in self.cgi_directories
973 (and the next character is a '/' or the end of the string).
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]974
975 """
Senthil Kumarand70846b2012-04-12 02:34:32 +0800[diff] [blame]976 collapsed_path = _url_collapse_path(self.path)
977 dir_sep = collapsed_path.find('/', 1)
978 head, tail = collapsed_path[:dir_sep], collapsed_path[dir_sep+1:]
Senthil Kumarandbb369d2012-04-11 03:15:28 +0800[diff] [blame]979 if head in self.cgi_directories:
980 self.cgi_info = head, tail
Benjamin Petersonad71f0f2009-04-11 20:12:10 +0000[diff] [blame]981 return True
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]982 return False
983
Senthil Kumarand70846b2012-04-12 02:34:32 +0800[diff] [blame]984
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]985 cgi_directories = ['/cgi-bin', '/htbin']
986
987 def is_executable(self, path):
988 """Test whether argument path is an executable file."""
989 return executable(path)
990
991 def is_python(self, path):
992 """Test whether argument path is a Python script."""
993 head, tail = os.path.splitext(path)
994 return tail.lower() in (".py", ".pyw")
995
996 def run_cgi(self):
997 """Execute a CGI script."""
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]998 dir, rest = self.cgi_info
999
Benjamin Peterson04e9de42013-10-30 12:43:09 -0400[diff] [blame]1000 i = rest.find('/')
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]1001 while i >= 0:
Benjamin Peterson04e9de42013-10-30 12:43:09 -0400[diff] [blame]1002 nextdir = rest[:i]
1003 nextrest = rest[i+1:]
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]1004
1005 scriptdir = self.translate_path(nextdir)
1006 if os.path.isdir(scriptdir):
1007 dir, rest = nextdir, nextrest
Benjamin Peterson04e9de42013-10-30 12:43:09 -0400[diff] [blame]1008 i = rest.find('/')
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]1009 else:
1010 break
1011
1012 # find an explicit query string, if present.
1013 i = rest.rfind('?')
1014 if i >= 0:
1015 rest, query = rest[:i], rest[i+1:]
1016 else:
1017 query = ''
1018
1019 # dissect the part after the directory name into a script name &
1020 # a possible additional path, to be stored in PATH_INFO.
1021 i = rest.find('/')
1022 if i >= 0:
1023 script, rest = rest[:i], rest[i:]
1024 else:
1025 script, rest = rest, ''
1026
1027 scriptname = dir + '/' + script
1028 scriptfile = self.translate_path(scriptname)
1029 if not os.path.exists(scriptfile):
1030 self.send_error(404, "No such CGI script (%r)" % scriptname)
1031 return
1032 if not os.path.isfile(scriptfile):
1033 self.send_error(403, "CGI script is not a plain file (%r)" %
1034 scriptname)
1035 return
1036 ispy = self.is_python(scriptname)
Victor Stinnerfb25ba92011-06-20 17:45:54 +0200[diff] [blame]1037 if self.have_fork or not ispy:
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]1038 if not self.is_executable(scriptfile):
1039 self.send_error(403, "CGI script is not executable (%r)" %
1040 scriptname)
1041 return
1042
1043 # Reference: http://hoohoo.ncsa.uiuc.edu/cgi/env.html
1044 # XXX Much of the following could be prepared ahead of time!
Senthil Kumaran42713722010-10-03 17:55:45 +0000[diff] [blame]1045 env = copy.deepcopy(os.environ)
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]1046 env['SERVER_SOFTWARE'] = self.version_string()
1047 env['SERVER_NAME'] = self.server.server_name
1048 env['GATEWAY_INTERFACE'] = 'CGI/1.1'
1049 env['SERVER_PROTOCOL'] = self.protocol_version
1050 env['SERVER_PORT'] = str(self.server.server_port)
1051 env['REQUEST_METHOD'] = self.command
Jeremy Hylton1afc1692008-06-18 20:49:58 +0000[diff] [blame]1052 uqrest = urllib.parse.unquote(rest)
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]1053 env['PATH_INFO'] = uqrest
1054 env['PATH_TRANSLATED'] = self.translate_path(uqrest)
1055 env['SCRIPT_NAME'] = scriptname
1056 if query:
1057 env['QUERY_STRING'] = query
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]1058 env['REMOTE_ADDR'] = self.client_address[0]
Barry Warsaw820c1202008-06-12 04:06:45 +0000[diff] [blame]1059 authorization = self.headers.get("authorization")
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]1060 if authorization:
1061 authorization = authorization.split()
1062 if len(authorization) == 2:
1063 import base64, binascii
1064 env['AUTH_TYPE'] = authorization[0]
1065 if authorization[0].lower() == "basic":
1066 try:
1067 authorization = authorization[1].encode('ascii')
Georg Brandl706824f2009-06-04 09:42:55 +0000[diff] [blame]1068 authorization = base64.decodebytes(authorization).\
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]1069 decode('ascii')
1070 except (binascii.Error, UnicodeError):
1071 pass
1072 else:
1073 authorization = authorization.split(':')
1074 if len(authorization) == 2:
1075 env['REMOTE_USER'] = authorization[0]
1076 # XXX REMOTE_IDENT
Barry Warsaw820c1202008-06-12 04:06:45 +0000[diff] [blame]1077 if self.headers.get('content-type') is None:
1078 env['CONTENT_TYPE'] = self.headers.get_content_type()
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]1079 else:
Barry Warsaw820c1202008-06-12 04:06:45 +0000[diff] [blame]1080 env['CONTENT_TYPE'] = self.headers['content-type']
1081 length = self.headers.get('content-length')
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]1082 if length:
1083 env['CONTENT_LENGTH'] = length
Barry Warsaw820c1202008-06-12 04:06:45 +0000[diff] [blame]1084 referer = self.headers.get('referer')
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]1085 if referer:
1086 env['HTTP_REFERER'] = referer
1087 accept = []
1088 for line in self.headers.getallmatchingheaders('accept'):
1089 if line[:1] in "\t\n\r ":
1090 accept.append(line.strip())
1091 else:
1092 accept = accept + line[7:].split(',')
1093 env['HTTP_ACCEPT'] = ','.join(accept)
Barry Warsaw820c1202008-06-12 04:06:45 +0000[diff] [blame]1094 ua = self.headers.get('user-agent')
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]1095 if ua:
1096 env['HTTP_USER_AGENT'] = ua
Barry Warsaw820c1202008-06-12 04:06:45 +0000[diff] [blame]1097 co = filter(None, self.headers.get_all('cookie', []))
Georg Brandl62e2ca22010-07-31 21:54:24 +0000[diff] [blame]1098 cookie_str = ', '.join(co)
1099 if cookie_str:
1100 env['HTTP_COOKIE'] = cookie_str
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]1101 # XXX Other HTTP_* headers
1102 # Since we're setting the env in the parent, provide empty
1103 # values to override previously set values
1104 for k in ('QUERY_STRING', 'REMOTE_HOST', 'CONTENT_LENGTH',
1105 'HTTP_USER_AGENT', 'HTTP_COOKIE', 'HTTP_REFERER'):
1106 env.setdefault(k, "")
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]1107
1108 self.send_response(200, "Script output follows")
Senthil Kumaranc7ae19b2011-05-09 23:25:02 +0800[diff] [blame]1109 self.flush_headers()
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]1110
1111 decoded_query = query.replace('+', ' ')
1112
1113 if self.have_fork:
1114 # Unix -- fork as we should
1115 args = [script]
1116 if '=' not in decoded_query:
1117 args.append(decoded_query)
1118 nobody = nobody_uid()
1119 self.wfile.flush() # Always flush before forking
1120 pid = os.fork()
1121 if pid != 0:
1122 # Parent
1123 pid, sts = os.waitpid(pid, 0)
1124 # throw away additional data [see bug #427345]
1125 while select.select([self.rfile], [], [], 0)[0]:
1126 if not self.rfile.read(1):
1127 break
1128 if sts:
1129 self.log_error("CGI script exit status %#x", sts)
1130 return
1131 # Child
1132 try:
1133 try:
1134 os.setuid(nobody)
Andrew Svetlovad28c7f2012-12-18 22:02:39 +0200[diff] [blame]1135 except OSError:
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]1136 pass
1137 os.dup2(self.rfile.fileno(), 0)
1138 os.dup2(self.wfile.fileno(), 1)
Senthil Kumaran42713722010-10-03 17:55:45 +0000[diff] [blame]1139 os.execve(scriptfile, args, env)
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]1140 except:
1141 self.server.handle_error(self.request, self.client_address)
1142 os._exit(127)
1143
Amaury Forgeot d'Arccb0d2d72008-06-18 22:19:22 +0000[diff] [blame]1144 else:
1145 # Non-Unix -- use subprocess
1146 import subprocess
Senthil Kumarane29cd162009-11-11 04:17:53 +0000[diff] [blame]1147 cmdline = [scriptfile]
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]1148 if self.is_python(scriptfile):
1149 interp = sys.executable
1150 if interp.lower().endswith("w.exe"):
1151 # On Windows, use python.exe, not pythonw.exe
1152 interp = interp[:-5] + interp[-4:]
Senthil Kumarane29cd162009-11-11 04:17:53 +0000[diff] [blame]1153 cmdline = [interp, '-u'] + cmdline
1154 if '=' not in query:
1155 cmdline.append(query)
1156 self.log_message("command: %s", subprocess.list2cmdline(cmdline))
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]1157 try:
1158 nbytes = int(length)
1159 except (TypeError, ValueError):
1160 nbytes = 0
Amaury Forgeot d'Arccb0d2d72008-06-18 22:19:22 +0000[diff] [blame]1161 p = subprocess.Popen(cmdline,
1162 stdin=subprocess.PIPE,
1163 stdout=subprocess.PIPE,
Senthil Kumaran42713722010-10-03 17:55:45 +0000[diff] [blame]1164 stderr=subprocess.PIPE,
1165 env = env
Amaury Forgeot d'Arccb0d2d72008-06-18 22:19:22 +0000[diff] [blame]1166 )
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]1167 if self.command.lower() == "post" and nbytes > 0:
1168 data = self.rfile.read(nbytes)
Amaury Forgeot d'Arccb0d2d72008-06-18 22:19:22 +0000[diff] [blame]1169 else:
1170 data = None
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]1171 # throw away additional data [see bug #427345]
1172 while select.select([self.rfile._sock], [], [], 0)[0]:
1173 if not self.rfile._sock.recv(1):
1174 break
Amaury Forgeot d'Arccb0d2d72008-06-18 22:19:22 +0000[diff] [blame]1175 stdout, stderr = p.communicate(data)
1176 self.wfile.write(stdout)
1177 if stderr:
1178 self.log_error('%s', stderr)
Brian Curtincbad4df2010-11-05 15:04:48 +0000[diff] [blame]1179 p.stderr.close()
1180 p.stdout.close()
Amaury Forgeot d'Arccb0d2d72008-06-18 22:19:22 +0000[diff] [blame]1181 status = p.returncode
1182 if status:
1183 self.log_error("CGI script exit status %#x", status)
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]1184 else:
1185 self.log_message("CGI script exited OK")
1186
1187
Senthil Kumarandefe7f42013-09-15 09:37:27 -0700[diff] [blame]1188def test(HandlerClass=BaseHTTPRequestHandler,
1189 ServerClass=HTTPServer, protocol="HTTP/1.0", port=8000, bind=""):
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]1190 """Test the HTTP request handler class.
1191
1192 This runs an HTTP server on port 8000 (or the first command line
1193 argument).
1194
1195 """
Senthil Kumarandefe7f42013-09-15 09:37:27 -0700[diff] [blame]1196 server_address = (bind, port)
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]1197
1198 HandlerClass.protocol_version = protocol
1199 httpd = ServerClass(server_address, HandlerClass)
1200
1201 sa = httpd.socket.getsockname()
1202 print("Serving HTTP on", sa[0], "port", sa[1], "...")
Alexandre Vassalottib5292a22009-04-03 07:16:55 +0000[diff] [blame]1203 try:
1204 httpd.serve_forever()
1205 except KeyboardInterrupt:
1206 print("\nKeyboard interrupt received, exiting.")
1207 httpd.server_close()
1208 sys.exit(0)
Georg Brandl24420152008-05-26 16:32:26 +0000[diff] [blame]1209
1210if __name__ == '__main__':
Senthil Kumaran1251faf2012-06-03 16:15:54 +0800[diff] [blame]1211 parser = argparse.ArgumentParser()
1212 parser.add_argument('--cgi', action='store_true',
1213 help='Run as CGI Server')
Senthil Kumarandefe7f42013-09-15 09:37:27 -0700[diff] [blame]1214 parser.add_argument('--bind', '-b', default='', metavar='ADDRESS',
1215 help='Specify alternate bind address '
1216 '[default: all interfaces]')
Senthil Kumaran1251faf2012-06-03 16:15:54 +0800[diff] [blame]1217 parser.add_argument('port', action='store',
1218 default=8000, type=int,
1219 nargs='?',
1220 help='Specify alternate port [default: 8000]')
1221 args = parser.parse_args()
1222 if args.cgi:
Senthil Kumarandefe7f42013-09-15 09:37:27 -0700[diff] [blame]1223 handler_class = CGIHTTPRequestHandler
Senthil Kumaran1251faf2012-06-03 16:15:54 +0800[diff] [blame]1224 else:
Senthil Kumarandefe7f42013-09-15 09:37:27 -0700[diff] [blame]1225 handler_class = SimpleHTTPRequestHandler
1226 test(HandlerClass=handler_class, port=args.port, bind=args.bind)