PageRenderTime 7ms CodeModel.GetById 2ms app.highlight 33ms RepoModel.GetById 1ms app.codeStats 0ms

/Doc/library/socket.rst

https://bitbucket.org/rmtew/cpython-stackless2
ReStructuredText | 1276 lines | 894 code | 382 blank | 0 comment | 0 complexity | 0001ef5968351fb2b036d4c0c327eb9c MD5 | raw file

Large files files are truncated, but you can click here to view the full file

   1:mod:`socket` --- Low-level networking interface
   2================================================
   3
   4.. module:: socket
   5   :synopsis: Low-level networking interface.
   6
   7
   8This module provides access to the BSD *socket* interface. It is available on
   9all modern Unix systems, Windows, MacOS, OS/2, and probably additional
  10platforms.
  11
  12.. note::
  13
  14   Some behavior may be platform dependent, since calls are made to the operating
  15   system socket APIs.
  16
  17.. index:: object: socket
  18
  19The Python interface is a straightforward transliteration of the Unix system
  20call and library interface for sockets to Python's object-oriented style: the
  21:func:`socket` function returns a :dfn:`socket object` whose methods implement
  22the various socket system calls.  Parameter types are somewhat higher-level than
  23in the C interface: as with :meth:`read` and :meth:`write` operations on Python
  24files, buffer allocation on receive operations is automatic, and buffer length
  25is implicit on send operations.
  26
  27
  28.. seealso::
  29
  30   Module :mod:`socketserver`
  31      Classes that simplify writing network servers.
  32
  33   Module :mod:`ssl`
  34      A TLS/SSL wrapper for socket objects.
  35
  36
  37Socket families
  38---------------
  39
  40Depending on the system and the build options, various socket families
  41are supported by this module.
  42
  43Socket addresses are represented as follows:
  44
  45- A single string is used for the :const:`AF_UNIX` address family.
  46
  47- A pair ``(host, port)`` is used for the :const:`AF_INET` address family,
  48  where *host* is a string representing either a hostname in Internet domain
  49  notation like ``'daring.cwi.nl'`` or an IPv4 address like ``'100.50.200.5'``,
  50  and *port* is an integral port number.
  51
  52- For :const:`AF_INET6` address family, a four-tuple ``(host, port, flowinfo,
  53  scopeid)`` is used, where *flowinfo* and *scopeid* represent the ``sin6_flowinfo``
  54  and ``sin6_scope_id`` members in :const:`struct sockaddr_in6` in C.  For
  55  :mod:`socket` module methods, *flowinfo* and *scopeid* can be omitted just for
  56  backward compatibility.  Note, however, omission of *scopeid* can cause problems
  57  in manipulating scoped IPv6 addresses.
  58
  59- :const:`AF_NETLINK` sockets are represented as pairs ``(pid, groups)``.
  60
  61- Linux-only support for TIPC is available using the :const:`AF_TIPC`
  62  address family.  TIPC is an open, non-IP based networked protocol designed
  63  for use in clustered computer environments.  Addresses are represented by a
  64  tuple, and the fields depend on the address type. The general tuple form is
  65  ``(addr_type, v1, v2, v3 [, scope])``, where:
  66
  67  - *addr_type* is one of TIPC_ADDR_NAMESEQ, TIPC_ADDR_NAME, or
  68    TIPC_ADDR_ID.
  69  - *scope* is one of TIPC_ZONE_SCOPE, TIPC_CLUSTER_SCOPE, and
  70    TIPC_NODE_SCOPE.
  71  - If *addr_type* is TIPC_ADDR_NAME, then *v1* is the server type, *v2* is
  72    the port identifier, and *v3* should be 0.
  73
  74    If *addr_type* is TIPC_ADDR_NAMESEQ, then *v1* is the server type, *v2*
  75    is the lower port number, and *v3* is the upper port number.
  76
  77    If *addr_type* is TIPC_ADDR_ID, then *v1* is the node, *v2* is the
  78    reference, and *v3* should be set to 0.
  79
  80    If *addr_type* is TIPC_ADDR_ID, then *v1* is the node, *v2* is the
  81    reference, and *v3* should be set to 0.
  82
  83- Certain other address families (:const:`AF_BLUETOOTH`, :const:`AF_PACKET`)
  84  support specific representations.
  85
  86  .. XXX document them!
  87
  88For IPv4 addresses, two special forms are accepted instead of a host address:
  89the empty string represents :const:`INADDR_ANY`, and the string
  90``'<broadcast>'`` represents :const:`INADDR_BROADCAST`.  This behavior is not
  91compatible with IPv6, therefore, you may want to avoid these if you intend
  92to support IPv6 with your Python programs.
  93
  94If you use a hostname in the *host* portion of IPv4/v6 socket address, the
  95program may show a nondeterministic behavior, as Python uses the first address
  96returned from the DNS resolution.  The socket address will be resolved
  97differently into an actual IPv4/v6 address, depending on the results from DNS
  98resolution and/or the host configuration.  For deterministic behavior use a
  99numeric address in *host* portion.
 100
 101All errors raise exceptions.  The normal exceptions for invalid argument types
 102and out-of-memory conditions can be raised; errors related to socket or address
 103semantics raise :exc:`socket.error` or one of its subclasses.
 104
 105Non-blocking mode is supported through :meth:`~socket.setblocking`.  A
 106generalization of this based on timeouts is supported through
 107:meth:`~socket.settimeout`.
 108
 109
 110Module contents
 111---------------
 112
 113The module :mod:`socket` exports the following constants and functions:
 114
 115
 116.. exception:: error
 117
 118   .. index:: module: errno
 119
 120   A subclass of :exc:`IOError`, this exception is raised for socket-related
 121   errors.  It is recommended that you inspect its ``errno`` attribute to
 122   discriminate between different kinds of errors.
 123
 124   .. seealso::
 125      The :mod:`errno` module contains symbolic names for the error codes
 126      defined by the underlying operating system.
 127
 128
 129.. exception:: herror
 130
 131   A subclass of :exc:`socket.error`, this exception is raised for
 132   address-related errors, i.e. for functions that use *h_errno* in the POSIX
 133   C API, including :func:`gethostbyname_ex` and :func:`gethostbyaddr`.
 134   The accompanying value is a pair ``(h_errno, string)`` representing an
 135   error returned by a library call.  *h_errno* is a numeric value, while
 136   *string* represents the description of *h_errno*, as returned by the
 137   :c:func:`hstrerror` C function.
 138
 139
 140.. exception:: gaierror
 141
 142   A subclass of :exc:`socket.error`, this exception is raised for
 143   address-related errors by :func:`getaddrinfo` and :func:`getnameinfo`.
 144   The accompanying value is a pair ``(error, string)`` representing an error
 145   returned by a library call.  *string* represents the description of
 146   *error*, as returned by the :c:func:`gai_strerror` C function.  The
 147   numeric *error* value will match one of the :const:`EAI_\*` constants
 148   defined in this module.
 149
 150
 151.. exception:: timeout
 152
 153   A subclass of :exc:`socket.error`, this exception is raised when a timeout
 154   occurs on a socket which has had timeouts enabled via a prior call to
 155   :meth:`~socket.settimeout` (or implicitly through
 156   :func:`~socket.setdefaulttimeout`).  The accompanying value is a string
 157   whose value is currently always "timed out".
 158
 159
 160.. data:: AF_UNIX
 161          AF_INET
 162          AF_INET6
 163
 164   These constants represent the address (and protocol) families, used for the
 165   first argument to :func:`socket`.  If the :const:`AF_UNIX` constant is not
 166   defined then this protocol is unsupported.  More constants may be available
 167   depending on the system.
 168
 169
 170.. data:: SOCK_STREAM
 171          SOCK_DGRAM
 172          SOCK_RAW
 173          SOCK_RDM
 174          SOCK_SEQPACKET
 175
 176   These constants represent the socket types, used for the second argument to
 177   :func:`socket`.  More constants may be available depending on the system.
 178   (Only :const:`SOCK_STREAM` and :const:`SOCK_DGRAM` appear to be generally
 179   useful.)
 180
 181.. data:: SOCK_CLOEXEC
 182          SOCK_NONBLOCK
 183
 184   These two constants, if defined, can be combined with the socket types and
 185   allow you to set some flags atomically (thus avoiding possible race
 186   conditions and the need for separate calls).
 187
 188   .. seealso::
 189
 190      `Secure File Descriptor Handling <http://udrepper.livejournal.com/20407.html>`_
 191      for a more thorough explanation.
 192
 193   Availability: Linux >= 2.6.27.
 194
 195   .. versionadded:: 3.2
 196
 197.. data:: SO_*
 198          SOMAXCONN
 199          MSG_*
 200          SOL_*
 201          SCM_*
 202          IPPROTO_*
 203          IPPORT_*
 204          INADDR_*
 205          IP_*
 206          IPV6_*
 207          EAI_*
 208          AI_*
 209          NI_*
 210          TCP_*
 211
 212   Many constants of these forms, documented in the Unix documentation on sockets
 213   and/or the IP protocol, are also defined in the socket module. They are
 214   generally used in arguments to the :meth:`setsockopt` and :meth:`getsockopt`
 215   methods of socket objects.  In most cases, only those symbols that are defined
 216   in the Unix header files are defined; for a few symbols, default values are
 217   provided.
 218
 219.. data:: SIO_*
 220          RCVALL_*
 221
 222   Constants for Windows' WSAIoctl(). The constants are used as arguments to the
 223   :meth:`ioctl` method of socket objects.
 224
 225
 226.. data:: TIPC_*
 227
 228   TIPC related constants, matching the ones exported by the C socket API. See
 229   the TIPC documentation for more information.
 230
 231
 232.. data:: has_ipv6
 233
 234   This constant contains a boolean value which indicates if IPv6 is supported on
 235   this platform.
 236
 237
 238.. function:: create_connection(address[, timeout[, source_address]])
 239
 240   Convenience function.  Connect to *address* (a 2-tuple ``(host, port)``),
 241   and return the socket object.  Passing the optional *timeout* parameter will
 242   set the timeout on the socket instance before attempting to connect.  If no
 243   *timeout* is supplied, the global default timeout setting returned by
 244   :func:`getdefaulttimeout` is used.
 245
 246   If supplied, *source_address* must be a 2-tuple ``(host, port)`` for the
 247   socket to bind to as its source address before connecting.  If host or port
 248   are '' or 0 respectively the OS default behavior will be used.
 249
 250   .. versionchanged:: 3.2
 251      *source_address* was added.
 252
 253   .. versionchanged:: 3.2
 254      support for the :keyword:`with` statement was added.
 255
 256
 257.. function:: getaddrinfo(host, port, family=0, type=0, proto=0, flags=0)
 258
 259   Translate the *host*/*port* argument into a sequence of 5-tuples that contain
 260   all the necessary arguments for creating a socket connected to that service.
 261   *host* is a domain name, a string representation of an IPv4/v6 address
 262   or ``None``. *port* is a string service name such as ``'http'``, a numeric
 263   port number or ``None``.  By passing ``None`` as the value of *host*
 264   and *port*, you can pass ``NULL`` to the underlying C API.
 265
 266   The *family*, *type* and *proto* arguments can be optionally specified
 267   in order to narrow the list of addresses returned.  Passing zero as a
 268   value for each of these arguments selects the full range of results.
 269   The *flags* argument can be one or several of the ``AI_*`` constants,
 270   and will influence how results are computed and returned.
 271   For example, :const:`AI_NUMERICHOST` will disable domain name resolution
 272   and will raise an error if *host* is a domain name.
 273
 274   The function returns a list of 5-tuples with the following structure:
 275
 276   ``(family, type, proto, canonname, sockaddr)``
 277
 278   In these tuples, *family*, *type*, *proto* are all integers and are
 279   meant to be passed to the :func:`socket` function.  *canonname* will be
 280   a string representing the canonical name of the *host* if
 281   :const:`AI_CANONNAME` is part of the *flags* argument; else *canonname*
 282   will be empty.  *sockaddr* is a tuple describing a socket address, whose
 283   format depends on the returned *family* (a ``(address, port)`` 2-tuple for
 284   :const:`AF_INET`, a ``(address, port, flow info, scope id)`` 4-tuple for
 285   :const:`AF_INET6`), and is meant to be passed to the :meth:`socket.connect`
 286   method.
 287
 288   The following example fetches address information for a hypothetical TCP
 289   connection to ``www.python.org`` on port 80 (results may differ on your
 290   system if IPv6 isn't enabled)::
 291
 292      >>> socket.getaddrinfo("www.python.org", 80, proto=socket.SOL_TCP)
 293      [(2, 1, 6, '', ('82.94.164.162', 80)),
 294       (10, 1, 6, '', ('2001:888:2000:d::a2', 80, 0, 0))]
 295
 296   .. versionchanged:: 3.2
 297      parameters can now be passed as single keyword arguments.
 298
 299.. function:: getfqdn([name])
 300
 301   Return a fully qualified domain name for *name*. If *name* is omitted or empty,
 302   it is interpreted as the local host.  To find the fully qualified name, the
 303   hostname returned by :func:`gethostbyaddr` is checked, followed by aliases for the
 304   host, if available.  The first name which includes a period is selected.  In
 305   case no fully qualified domain name is available, the hostname as returned by
 306   :func:`gethostname` is returned.
 307
 308
 309.. function:: gethostbyname(hostname)
 310
 311   Translate a host name to IPv4 address format.  The IPv4 address is returned as a
 312   string, such as  ``'100.50.200.5'``.  If the host name is an IPv4 address itself
 313   it is returned unchanged.  See :func:`gethostbyname_ex` for a more complete
 314   interface. :func:`gethostbyname` does not support IPv6 name resolution, and
 315   :func:`getaddrinfo` should be used instead for IPv4/v6 dual stack support.
 316
 317
 318.. function:: gethostbyname_ex(hostname)
 319
 320   Translate a host name to IPv4 address format, extended interface. Return a
 321   triple ``(hostname, aliaslist, ipaddrlist)`` where *hostname* is the primary
 322   host name responding to the given *ip_address*, *aliaslist* is a (possibly
 323   empty) list of alternative host names for the same address, and *ipaddrlist* is
 324   a list of IPv4 addresses for the same interface on the same host (often but not
 325   always a single address). :func:`gethostbyname_ex` does not support IPv6 name
 326   resolution, and :func:`getaddrinfo` should be used instead for IPv4/v6 dual
 327   stack support.
 328
 329
 330.. function:: gethostname()
 331
 332   Return a string containing the hostname of the machine where  the Python
 333   interpreter is currently executing.
 334
 335   If you want to know the current machine's IP address, you may want to use
 336   ``gethostbyname(gethostname())``. This operation assumes that there is a
 337   valid address-to-host mapping for the host, and the assumption does not
 338   always hold.
 339
 340   Note: :func:`gethostname` doesn't always return the fully qualified domain
 341   name; use ``getfqdn()`` (see above).
 342
 343
 344.. function:: gethostbyaddr(ip_address)
 345
 346   Return a triple ``(hostname, aliaslist, ipaddrlist)`` where *hostname* is the
 347   primary host name responding to the given *ip_address*, *aliaslist* is a
 348   (possibly empty) list of alternative host names for the same address, and
 349   *ipaddrlist* is a list of IPv4/v6 addresses for the same interface on the same
 350   host (most likely containing only a single address). To find the fully qualified
 351   domain name, use the function :func:`getfqdn`. :func:`gethostbyaddr` supports
 352   both IPv4 and IPv6.
 353
 354
 355.. function:: getnameinfo(sockaddr, flags)
 356
 357   Translate a socket address *sockaddr* into a 2-tuple ``(host, port)``. Depending
 358   on the settings of *flags*, the result can contain a fully-qualified domain name
 359   or numeric address representation in *host*.  Similarly, *port* can contain a
 360   string port name or a numeric port number.
 361
 362
 363.. function:: getprotobyname(protocolname)
 364
 365   Translate an Internet protocol name (for example, ``'icmp'``) to a constant
 366   suitable for passing as the (optional) third argument to the :func:`socket`
 367   function.  This is usually only needed for sockets opened in "raw" mode
 368   (:const:`SOCK_RAW`); for the normal socket modes, the correct protocol is chosen
 369   automatically if the protocol is omitted or zero.
 370
 371
 372.. function:: getservbyname(servicename[, protocolname])
 373
 374   Translate an Internet service name and protocol name to a port number for that
 375   service.  The optional protocol name, if given, should be ``'tcp'`` or
 376   ``'udp'``, otherwise any protocol will match.
 377
 378
 379.. function:: getservbyport(port[, protocolname])
 380
 381   Translate an Internet port number and protocol name to a service name for that
 382   service.  The optional protocol name, if given, should be ``'tcp'`` or
 383   ``'udp'``, otherwise any protocol will match.
 384
 385
 386.. function:: socket([family[, type[, proto]]])
 387
 388   Create a new socket using the given address family, socket type and protocol
 389   number.  The address family should be :const:`AF_INET` (the default),
 390   :const:`AF_INET6` or :const:`AF_UNIX`.  The socket type should be
 391   :const:`SOCK_STREAM` (the default), :const:`SOCK_DGRAM` or perhaps one of the
 392   other ``SOCK_`` constants.  The protocol number is usually zero and may be
 393   omitted in that case.
 394
 395
 396.. function:: socketpair([family[, type[, proto]]])
 397
 398   Build a pair of connected socket objects using the given address family, socket
 399   type, and protocol number.  Address family, socket type, and protocol number are
 400   as for the :func:`socket` function above. The default family is :const:`AF_UNIX`
 401   if defined on the platform; otherwise, the default is :const:`AF_INET`.
 402   Availability: Unix.
 403
 404   .. versionchanged:: 3.2
 405      The returned socket objects now support the whole socket API, rather
 406      than a subset.
 407
 408
 409.. function:: fromfd(fd, family, type[, proto])
 410
 411   Duplicate the file descriptor *fd* (an integer as returned by a file object's
 412   :meth:`fileno` method) and build a socket object from the result.  Address
 413   family, socket type and protocol number are as for the :func:`socket` function
 414   above. The file descriptor should refer to a socket, but this is not checked ---
 415   subsequent operations on the object may fail if the file descriptor is invalid.
 416   This function is rarely needed, but can be used to get or set socket options on
 417   a socket passed to a program as standard input or output (such as a server
 418   started by the Unix inet daemon).  The socket is assumed to be in blocking mode.
 419
 420
 421.. function:: ntohl(x)
 422
 423   Convert 32-bit positive integers from network to host byte order.  On machines
 424   where the host byte order is the same as network byte order, this is a no-op;
 425   otherwise, it performs a 4-byte swap operation.
 426
 427
 428.. function:: ntohs(x)
 429
 430   Convert 16-bit positive integers from network to host byte order.  On machines
 431   where the host byte order is the same as network byte order, this is a no-op;
 432   otherwise, it performs a 2-byte swap operation.
 433
 434
 435.. function:: htonl(x)
 436
 437   Convert 32-bit positive integers from host to network byte order.  On machines
 438   where the host byte order is the same as network byte order, this is a no-op;
 439   otherwise, it performs a 4-byte swap operation.
 440
 441
 442.. function:: htons(x)
 443
 444   Convert 16-bit positive integers from host to network byte order.  On machines
 445   where the host byte order is the same as network byte order, this is a no-op;
 446   otherwise, it performs a 2-byte swap operation.
 447
 448
 449.. function:: inet_aton(ip_string)
 450
 451   Convert an IPv4 address from dotted-quad string format (for example,
 452   '123.45.67.89') to 32-bit packed binary format, as a bytes object four characters in
 453   length.  This is useful when conversing with a program that uses the standard C
 454   library and needs objects of type :c:type:`struct in_addr`, which is the C type
 455   for the 32-bit packed binary this function returns.
 456
 457   :func:`inet_aton` also accepts strings with less than three dots; see the
 458   Unix manual page :manpage:`inet(3)` for details.
 459
 460   If the IPv4 address string passed to this function is invalid,
 461   :exc:`socket.error` will be raised. Note that exactly what is valid depends on
 462   the underlying C implementation of :c:func:`inet_aton`.
 463
 464   :func:`inet_aton` does not support IPv6, and :func:`inet_pton` should be used
 465   instead for IPv4/v6 dual stack support.
 466
 467
 468.. function:: inet_ntoa(packed_ip)
 469
 470   Convert a 32-bit packed IPv4 address (a bytes object four characters in
 471   length) to its standard dotted-quad string representation (for example,
 472   '123.45.67.89').  This is useful when conversing with a program that uses the
 473   standard C library and needs objects of type :c:type:`struct in_addr`, which
 474   is the C type for the 32-bit packed binary data this function takes as an
 475   argument.
 476
 477   If the byte sequence passed to this function is not exactly 4 bytes in
 478   length, :exc:`socket.error` will be raised. :func:`inet_ntoa` does not
 479   support IPv6, and :func:`inet_ntop` should be used instead for IPv4/v6 dual
 480   stack support.
 481
 482
 483.. function:: inet_pton(address_family, ip_string)
 484
 485   Convert an IP address from its family-specific string format to a packed,
 486   binary format. :func:`inet_pton` is useful when a library or network protocol
 487   calls for an object of type :c:type:`struct in_addr` (similar to
 488   :func:`inet_aton`) or :c:type:`struct in6_addr`.
 489
 490   Supported values for *address_family* are currently :const:`AF_INET` and
 491   :const:`AF_INET6`. If the IP address string *ip_string* is invalid,
 492   :exc:`socket.error` will be raised. Note that exactly what is valid depends on
 493   both the value of *address_family* and the underlying implementation of
 494   :c:func:`inet_pton`.
 495
 496   Availability: Unix (maybe not all platforms).
 497
 498
 499.. function:: inet_ntop(address_family, packed_ip)
 500
 501   Convert a packed IP address (a bytes object of some number of characters) to its
 502   standard, family-specific string representation (for example, ``'7.10.0.5'`` or
 503   ``'5aef:2b::8'``). :func:`inet_ntop` is useful when a library or network protocol
 504   returns an object of type :c:type:`struct in_addr` (similar to :func:`inet_ntoa`)
 505   or :c:type:`struct in6_addr`.
 506
 507   Supported values for *address_family* are currently :const:`AF_INET` and
 508   :const:`AF_INET6`. If the string *packed_ip* is not the correct length for the
 509   specified address family, :exc:`ValueError` will be raised.  A
 510   :exc:`socket.error` is raised for errors from the call to :func:`inet_ntop`.
 511
 512   Availability: Unix (maybe not all platforms).
 513
 514
 515..
 516   XXX: Are sendmsg(), recvmsg() and CMSG_*() available on any
 517   non-Unix platforms?  The old (obsolete?) 4.2BSD form of the
 518   interface, in which struct msghdr has no msg_control or
 519   msg_controllen members, is not currently supported.
 520
 521.. function:: CMSG_LEN(length)
 522
 523   Return the total length, without trailing padding, of an ancillary
 524   data item with associated data of the given *length*.  This value
 525   can often be used as the buffer size for :meth:`~socket.recvmsg` to
 526   receive a single item of ancillary data, but :rfc:`3542` requires
 527   portable applications to use :func:`CMSG_SPACE` and thus include
 528   space for padding, even when the item will be the last in the
 529   buffer.  Raises :exc:`OverflowError` if *length* is outside the
 530   permissible range of values.
 531
 532   Availability: most Unix platforms, possibly others.
 533
 534   .. versionadded:: 3.3
 535
 536
 537.. function:: CMSG_SPACE(length)
 538
 539   Return the buffer size needed for :meth:`~socket.recvmsg` to
 540   receive an ancillary data item with associated data of the given
 541   *length*, along with any trailing padding.  The buffer space needed
 542   to receive multiple items is the sum of the :func:`CMSG_SPACE`
 543   values for their associated data lengths.  Raises
 544   :exc:`OverflowError` if *length* is outside the permissible range
 545   of values.
 546
 547   Note that some systems might support ancillary data without
 548   providing this function.  Also note that setting the buffer size
 549   using the results of this function may not precisely limit the
 550   amount of ancillary data that can be received, since additional
 551   data may be able to fit into the padding area.
 552
 553   Availability: most Unix platforms, possibly others.
 554
 555   .. versionadded:: 3.3
 556
 557
 558.. function:: getdefaulttimeout()
 559
 560   Return the default timeout in seconds (float) for new socket objects. A value
 561   of ``None`` indicates that new socket objects have no timeout. When the socket
 562   module is first imported, the default is ``None``.
 563
 564
 565.. function:: setdefaulttimeout(timeout)
 566
 567   Set the default timeout in seconds (float) for new socket objects.  When
 568   the socket module is first imported, the default is ``None``.  See
 569   :meth:`~socket.settimeout` for possible values and their respective
 570   meanings.
 571
 572
 573.. function:: sethostname(name)
 574
 575   Set the machine's hostname to *name*.  This will raise a
 576   :exc:`socket.error` if you don't have enough rights.
 577
 578   Availability: Unix.
 579
 580   .. versionadded:: 3.3
 581
 582
 583.. function:: if_nameindex()
 584
 585   Return a list of network interface information
 586   (index int, name string) tuples.
 587   :exc:`socket.error` if the system call fails.
 588
 589   Availability: Unix.
 590
 591   .. versionadded:: 3.3
 592
 593
 594.. function:: if_nametoindex(if_name)
 595
 596   Return a network interface index number corresponding to an
 597   interface name.
 598   :exc:`socket.error` if no interface with the given name exists.
 599
 600   Availability: Unix.
 601
 602   .. versionadded:: 3.3
 603
 604
 605.. function:: if_indextoname(if_index)
 606
 607   Return a network interface name corresponding to a
 608   interface index number.
 609   :exc:`socket.error` if no interface with the given index exists.
 610
 611   Availability: Unix.
 612
 613   .. versionadded:: 3.3
 614
 615
 616.. data:: SocketType
 617
 618   This is a Python type object that represents the socket object type. It is the
 619   same as ``type(socket(...))``.
 620
 621
 622.. _socket-objects:
 623
 624Socket Objects
 625--------------
 626
 627Socket objects have the following methods.  Except for :meth:`makefile` these
 628correspond to Unix system calls applicable to sockets.
 629
 630
 631.. method:: socket.accept()
 632
 633   Accept a connection. The socket must be bound to an address and listening for
 634   connections. The return value is a pair ``(conn, address)`` where *conn* is a
 635   *new* socket object usable to send and receive data on the connection, and
 636   *address* is the address bound to the socket on the other end of the connection.
 637
 638
 639.. method:: socket.bind(address)
 640
 641   Bind the socket to *address*.  The socket must not already be bound. (The format
 642   of *address* depends on the address family --- see above.)
 643
 644
 645.. method:: socket.close()
 646
 647   Close the socket.  All future operations on the socket object will fail. The
 648   remote end will receive no more data (after queued data is flushed). Sockets are
 649   automatically closed when they are garbage-collected.
 650
 651   .. note::
 652      :meth:`close()` releases the resource associated with a connection but
 653      does not necessarily close the connection immediately.  If you want
 654      to close the connection in a timely fashion, call :meth:`shutdown()`
 655      before :meth:`close()`.
 656
 657
 658.. method:: socket.connect(address)
 659
 660   Connect to a remote socket at *address*. (The format of *address* depends on the
 661   address family --- see above.)
 662
 663
 664.. method:: socket.connect_ex(address)
 665
 666   Like ``connect(address)``, but return an error indicator instead of raising an
 667   exception for errors returned by the C-level :c:func:`connect` call (other
 668   problems, such as "host not found," can still raise exceptions).  The error
 669   indicator is ``0`` if the operation succeeded, otherwise the value of the
 670   :c:data:`errno` variable.  This is useful to support, for example, asynchronous
 671   connects.
 672
 673
 674.. method:: socket.detach()
 675
 676   Put the socket object into closed state without actually closing the
 677   underlying file descriptor.  The file descriptor is returned, and can
 678   be reused for other purposes.
 679
 680   .. versionadded:: 3.2
 681
 682
 683.. method:: socket.fileno()
 684
 685   Return the socket's file descriptor (a small integer).  This is useful with
 686   :func:`select.select`.
 687
 688   Under Windows the small integer returned by this method cannot be used where a
 689   file descriptor can be used (such as :func:`os.fdopen`).  Unix does not have
 690   this limitation.
 691
 692
 693.. method:: socket.getpeername()
 694
 695   Return the remote address to which the socket is connected.  This is useful to
 696   find out the port number of a remote IPv4/v6 socket, for instance. (The format
 697   of the address returned depends on the address family --- see above.)  On some
 698   systems this function is not supported.
 699
 700
 701.. method:: socket.getsockname()
 702
 703   Return the socket's own address.  This is useful to find out the port number of
 704   an IPv4/v6 socket, for instance. (The format of the address returned depends on
 705   the address family --- see above.)
 706
 707
 708.. method:: socket.getsockopt(level, optname[, buflen])
 709
 710   Return the value of the given socket option (see the Unix man page
 711   :manpage:`getsockopt(2)`).  The needed symbolic constants (:const:`SO_\*` etc.)
 712   are defined in this module.  If *buflen* is absent, an integer option is assumed
 713   and its integer value is returned by the function.  If *buflen* is present, it
 714   specifies the maximum length of the buffer used to receive the option in, and
 715   this buffer is returned as a bytes object.  It is up to the caller to decode the
 716   contents of the buffer (see the optional built-in module :mod:`struct` for a way
 717   to decode C structures encoded as byte strings).
 718
 719
 720.. method:: socket.gettimeout()
 721
 722   Return the timeout in seconds (float) associated with socket operations,
 723   or ``None`` if no timeout is set.  This reflects the last call to
 724   :meth:`setblocking` or :meth:`settimeout`.
 725
 726
 727.. method:: socket.ioctl(control, option)
 728
 729   :platform: Windows
 730
 731   The :meth:`ioctl` method is a limited interface to the WSAIoctl system
 732   interface.  Please refer to the `Win32 documentation
 733   <http://msdn.microsoft.com/en-us/library/ms741621%28VS.85%29.aspx>`_ for more
 734   information.
 735
 736   On other platforms, the generic :func:`fcntl.fcntl` and :func:`fcntl.ioctl`
 737   functions may be used; they accept a socket object as their first argument.
 738
 739.. method:: socket.listen(backlog)
 740
 741   Listen for connections made to the socket.  The *backlog* argument specifies the
 742   maximum number of queued connections and should be at least 0; the maximum value
 743   is system-dependent (usually 5), the minimum value is forced to 0.
 744
 745
 746.. method:: socket.makefile(mode='r', buffering=None, *, encoding=None, \
 747                            errors=None, newline=None)
 748
 749   .. index:: single: I/O control; buffering
 750
 751   Return a :term:`file object` associated with the socket.  The exact returned
 752   type depends on the arguments given to :meth:`makefile`.  These arguments are
 753   interpreted the same way as by the built-in :func:`open` function.
 754
 755   Closing the file object won't close the socket unless there are no remaining
 756   references to the socket.  The socket must be in blocking mode; it can have
 757   a timeout, but the file object's internal buffer may end up in a inconsistent
 758   state if a timeout occurs.
 759
 760   .. note::
 761
 762      On Windows, the file-like object created by :meth:`makefile` cannot be
 763      used where a file object with a file descriptor is expected, such as the
 764      stream arguments of :meth:`subprocess.Popen`.
 765
 766
 767.. method:: socket.recv(bufsize[, flags])
 768
 769   Receive data from the socket.  The return value is a bytes object representing the
 770   data received.  The maximum amount of data to be received at once is specified
 771   by *bufsize*.  See the Unix manual page :manpage:`recv(2)` for the meaning of
 772   the optional argument *flags*; it defaults to zero.
 773
 774   .. note::
 775
 776      For best match with hardware and network realities, the value of  *bufsize*
 777      should be a relatively small power of 2, for example, 4096.
 778
 779
 780.. method:: socket.recvfrom(bufsize[, flags])
 781
 782   Receive data from the socket.  The return value is a pair ``(bytes, address)``
 783   where *bytes* is a bytes object representing the data received and *address* is the
 784   address of the socket sending the data.  See the Unix manual page
 785   :manpage:`recv(2)` for the meaning of the optional argument *flags*; it defaults
 786   to zero. (The format of *address* depends on the address family --- see above.)
 787
 788
 789.. method:: socket.recvmsg(bufsize[, ancbufsize[, flags]])
 790
 791   Receive normal data (up to *bufsize* bytes) and ancillary data from
 792   the socket.  The *ancbufsize* argument sets the size in bytes of
 793   the internal buffer used to receive the ancillary data; it defaults
 794   to 0, meaning that no ancillary data will be received.  Appropriate
 795   buffer sizes for ancillary data can be calculated using
 796   :func:`CMSG_SPACE` or :func:`CMSG_LEN`, and items which do not fit
 797   into the buffer might be truncated or discarded.  The *flags*
 798   argument defaults to 0 and has the same meaning as for
 799   :meth:`recv`.
 800
 801   The return value is a 4-tuple: ``(data, ancdata, msg_flags,
 802   address)``.  The *data* item is a :class:`bytes` object holding the
 803   non-ancillary data received.  The *ancdata* item is a list of zero
 804   or more tuples ``(cmsg_level, cmsg_type, cmsg_data)`` representing
 805   the ancillary data (control messages) received: *cmsg_level* and
 806   *cmsg_type* are integers specifying the protocol level and
 807   protocol-specific type respectively, and *cmsg_data* is a
 808   :class:`bytes` object holding the associated data.  The *msg_flags*
 809   item is the bitwise OR of various flags indicating conditions on
 810   the received message; see your system documentation for details.
 811   If the receiving socket is unconnected, *address* is the address of
 812   the sending socket, if available; otherwise, its value is
 813   unspecified.
 814
 815   On some systems, :meth:`sendmsg` and :meth:`recvmsg` can be used to
 816   pass file descriptors between processes over an :const:`AF_UNIX`
 817   socket.  When this facility is used (it is often restricted to
 818   :const:`SOCK_STREAM` sockets), :meth:`recvmsg` will return, in its
 819   ancillary data, items of the form ``(socket.SOL_SOCKET,
 820   socket.SCM_RIGHTS, fds)``, where *fds* is a :class:`bytes` object
 821   representing the new file descriptors as a binary array of the
 822   native C :c:type:`int` type.  If :meth:`recvmsg` raises an
 823   exception after the system call returns, it will first attempt to
 824   close any file descriptors received via this mechanism.
 825
 826   Some systems do not indicate the truncated length of ancillary data
 827   items which have been only partially received.  If an item appears
 828   to extend beyond the end of the buffer, :meth:`recvmsg` will issue
 829   a :exc:`RuntimeWarning`, and will return the part of it which is
 830   inside the buffer provided it has not been truncated before the
 831   start of its associated data.
 832
 833   On systems which support the :const:`SCM_RIGHTS` mechanism, the
 834   following function will receive up to *maxfds* file descriptors,
 835   returning the message data and a list containing the descriptors
 836   (while ignoring unexpected conditions such as unrelated control
 837   messages being received).  See also :meth:`sendmsg`. ::
 838
 839      import socket, array
 840
 841      def recv_fds(sock, msglen, maxfds):
 842          fds = array.array("i")   # Array of ints
 843          msg, ancdata, flags, addr = sock.recvmsg(msglen, socket.CMSG_LEN(maxfds * fds.itemsize))
 844          for cmsg_level, cmsg_type, cmsg_data in ancdata:
 845              if (cmsg_level == socket.SOL_SOCKET and cmsg_type == socket.SCM_RIGHTS):
 846                  # Append data, ignoring any truncated integers at the end.
 847                  fds.fromstring(cmsg_data[:len(cmsg_data) - (len(cmsg_data) % fds.itemsize)])
 848          return msg, list(fds)
 849
 850   Availability: most Unix platforms, possibly others.
 851
 852   .. versionadded:: 3.3
 853
 854
 855.. method:: socket.recvmsg_into(buffers[, ancbufsize[, flags]])
 856
 857   Receive normal data and ancillary data from the socket, behaving as
 858   :meth:`recvmsg` would, but scatter the non-ancillary data into a
 859   series of buffers instead of returning a new bytes object.  The
 860   *buffers* argument must be an iterable of objects that export
 861   writable buffers (e.g. :class:`bytearray` objects); these will be
 862   filled with successive chunks of the non-ancillary data until it
 863   has all been written or there are no more buffers.  The operating
 864   system may set a limit (:func:`~os.sysconf` value ``SC_IOV_MAX``)
 865   on the number of buffers that can be used.  The *ancbufsize* and
 866   *flags* arguments have the same meaning as for :meth:`recvmsg`.
 867
 868   The return value is a 4-tuple: ``(nbytes, ancdata, msg_flags,
 869   address)``, where *nbytes* is the total number of bytes of
 870   non-ancillary data written into the buffers, and *ancdata*,
 871   *msg_flags* and *address* are the same as for :meth:`recvmsg`.
 872
 873   Example::
 874
 875      >>> import socket
 876      >>> s1, s2 = socket.socketpair()
 877      >>> b1 = bytearray(b'----')
 878      >>> b2 = bytearray(b'0123456789')
 879      >>> b3 = bytearray(b'--------------')
 880      >>> s1.send(b'Mary had a little lamb')
 881      22
 882      >>> s2.recvmsg_into([b1, memoryview(b2)[2:9], b3])
 883      (22, [], 0, None)
 884      >>> [b1, b2, b3]
 885      [bytearray(b'Mary'), bytearray(b'01 had a 9'), bytearray(b'little lamb---')]
 886
 887   Availability: most Unix platforms, possibly others.
 888
 889   .. versionadded:: 3.3
 890
 891
 892.. method:: socket.recvfrom_into(buffer[, nbytes[, flags]])
 893
 894   Receive data from the socket, writing it into *buffer* instead of creating a
 895   new bytestring.  The return value is a pair ``(nbytes, address)`` where *nbytes* is
 896   the number of bytes received and *address* is the address of the socket sending
 897   the data.  See the Unix manual page :manpage:`recv(2)` for the meaning of the
 898   optional argument *flags*; it defaults to zero.  (The format of *address*
 899   depends on the address family --- see above.)
 900
 901
 902.. method:: socket.recv_into(buffer[, nbytes[, flags]])
 903
 904   Receive up to *nbytes* bytes from the socket, storing the data into a buffer
 905   rather than creating a new bytestring.  If *nbytes* is not specified (or 0),
 906   receive up to the size available in the given buffer.  Returns the number of
 907   bytes received.  See the Unix manual page :manpage:`recv(2)` for the meaning
 908   of the optional argument *flags*; it defaults to zero.
 909
 910
 911.. method:: socket.send(bytes[, flags])
 912
 913   Send data to the socket.  The socket must be connected to a remote socket.  The
 914   optional *flags* argument has the same meaning as for :meth:`recv` above.
 915   Returns the number of bytes sent. Applications are responsible for checking that
 916   all data has been sent; if only some of the data was transmitted, the
 917   application needs to attempt delivery of the remaining data.
 918
 919
 920.. method:: socket.sendall(bytes[, flags])
 921
 922   Send data to the socket.  The socket must be connected to a remote socket.  The
 923   optional *flags* argument has the same meaning as for :meth:`recv` above.
 924   Unlike :meth:`send`, this method continues to send data from *bytes* until
 925   either all data has been sent or an error occurs.  ``None`` is returned on
 926   success.  On error, an exception is raised, and there is no way to determine how
 927   much data, if any, was successfully sent.
 928
 929
 930.. method:: socket.sendto(bytes[, flags], address)
 931
 932   Send data to the socket.  The socket should not be connected to a remote socket,
 933   since the destination socket is specified by *address*.  The optional *flags*
 934   argument has the same meaning as for :meth:`recv` above.  Return the number of
 935   bytes sent. (The format of *address* depends on the address family --- see
 936   above.)
 937
 938
 939.. method:: socket.sendmsg(buffers[, ancdata[, flags[, address]]])
 940
 941   Send normal and ancillary data to the socket, gathering the
 942   non-ancillary data from a series of buffers and concatenating it
 943   into a single message.  The *buffers* argument specifies the
 944   non-ancillary data as an iterable of buffer-compatible objects
 945   (e.g. :class:`bytes` objects); the operating system may set a limit
 946   (:func:`~os.sysconf` value ``SC_IOV_MAX``) on the number of buffers
 947   that can be used.  The *ancdata* argument specifies the ancillary
 948   data (control messages) as an iterable of zero or more tuples
 949   ``(cmsg_level, cmsg_type, cmsg_data)``, where *cmsg_level* and
 950   *cmsg_type* are integers specifying the protocol level and
 951   protocol-specific type respectively, and *cmsg_data* is a
 952   buffer-compatible object holding the associated data.  Note that
 953   some systems (in particular, systems without :func:`CMSG_SPACE`)
 954   might support sending only one control message per call.  The
 955   *flags* argument defaults to 0 and has the same meaning as for
 956   :meth:`send`.  If *address* is supplied and not ``None``, it sets a
 957   destination address for the message.  The return value is the
 958   number of bytes of non-ancillary data sent.
 959
 960   The following function sends the list of file descriptors *fds*
 961   over an :const:`AF_UNIX` socket, on systems which support the
 962   :const:`SCM_RIGHTS` mechanism.  See also :meth:`recvmsg`. ::
 963
 964      import socket, array
 965
 966      def send_fds(sock, msg, fds):
 967          return sock.sendmsg([msg], [(socket.SOL_SOCKET, socket.SCM_RIGHTS, array.array("i", fds))])
 968
 969   Availability: most Unix platforms, possibly others.
 970
 971   .. versionadded:: 3.3
 972
 973
 974.. method:: socket.setblocking(flag)
 975
 976   Set blocking or non-blocking mode of the socket: if *flag* is false, the
 977   socket is set to non-blocking, else to blocking mode.
 978
 979   This method is a shorthand for certain :meth:`~socket.settimeout` calls:
 980
 981   * ``sock.setblocking(True)`` is equivalent to ``sock.settimeout(None)``
 982
 983   * ``sock.setblocking(False)`` is equivalent to ``sock.settimeout(0.0)``
 984
 985
 986.. method:: socket.settimeout(value)
 987
 988   Set a timeout on blocking socket operations.  The *value* argument can be a
 989   nonnegative floating point number expressing seconds, or ``None``.
 990   If a non-zero value is given, subsequent socket operations will raise a
 991   :exc:`timeout` exception if the timeout period *value* has elapsed before
 992   the operation has completed.  If zero is given, the socket is put in
 993   non-blocking mode. If ``None`` is given, the socket is put in blocking mode.
 994
 995   For further information, please consult the :ref:`notes on socket timeouts <socket-timeouts>`.
 996
 997
 998.. method:: socket.setsockopt(level, optname, value)
 999
1000   .. index:: module: struct
1001
1002   Set the value of the given socket option (see the Unix manual page
1003   :manpage:`setsockopt(2)`).  The needed symbolic constants are defined in the
1004   :mod:`socket` module (:const:`SO_\*` etc.).  The value can be an integer or a
1005   bytes object representing a buffer.  In the latter case it is up to the caller to
1006   ensure that the bytestring contains the proper bits (see the optional built-in
1007   module :mod:`struct` for a way to encode C structures as bytestrings).
1008
1009
1010.. method:: socket.shutdown(how)
1011
1012   Shut down one or both halves of the connection.  If *how* is :const:`SHUT_RD`,
1013   further receives are disallowed.  If *how* is :const:`SHUT_WR`, further sends
1014   are disallowed.  If *how* is :const:`SHUT_RDWR`, further sends and receives are
1015   disallowed.  Depending on the platform, shutting down one half of the connection
1016   can also close the opposite half (e.g. on Mac OS X, ``shutdown(SHUT_WR)`` does
1017   not allow further reads on the other end of the connection).
1018
1019Note that there are no methods :meth:`read` or :meth:`write`; use
1020:meth:`~socket.recv` and :meth:`~socket.send` without *flags* argument instead.
1021
1022Socket objects also have these (read-only) attributes that correspond to the
1023values given to the :class:`socket` constructor.
1024
1025
1026.. attribute:: socket.family
1027
1028   The socket family.
1029
1030
1031.. attribute:: socket.type
1032
1033   The socket type.
1034
1035
1036.. attribute:: socket.proto
1037
1038   The socket protocol.
1039
1040
1041
1042.. _socket-timeouts:
1043
1044Notes on socket timeouts
1045------------------------
1046
1047A socket object can be in one of three modes: blocking, non-blocking, or
1048timeout.  Sockets are by default always created in blocking mode, but this
1049can be changed by calling :func:`setdefaulttimeout`.
1050
1051* In *blocking mode*, operations block until complete or the system returns
1052  an error (such as connection timed out).
1053
1054* In *non-blocking mode*, operations fail (with an error that is unfortunately
1055  system-dependent) if they cannot be completed immediately: functions from the
1056  :mod:`select` can be used to know when and whether a socket is available for
1057  reading or writing.
1058
1059* In *timeout mode*, operations fail if they cannot be completed within the
1060  timeout specified for the socket (they raise a :exc:`timeout` exception)
1061  or if the system returns an error.
1062
1063.. note::
1064   At the operating system level, sockets in *timeout mode* are internally set
1065   in non-blocking mode.  Also, the blocking and timeout modes are shared between
1066   file descriptors and socket objects that refer to the same network endpoint.
1067   This implementation detail can have visible consequences if e.g. you decide
1068   to use the :meth:`~socket.fileno()` of a socket.
1069
1070Timeouts and the ``connect`` method
1071^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
1072
1073The :meth:`~socket.connect` operation is also subject to the timeout
1074setting, and in general it is recommended to call :meth:`~socket.settimeout`
1075before calling :meth:`~socket.connect` or pass a timeout parameter to
1076:meth:`create_connection`.  However, the system network stack may also
1077return a connection timeout error of its own regardless of any Python socket
1078timeout setting.
1079
1080Timeouts and the ``accept`` method
1081^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
1082
1083If :func:`getdefaulttimeout` is not :const:`None`, sockets returned by
1084the :meth:`~socket.accept` method inherit that timeout.  Otherwise, the
1085behaviour depends on settings of the listening socket:
1086
1087* if the listening socket is in *blocking mode* or in *timeout mode*,
1088  the socket returned by :meth:`~socket.accept` is in *blocking mode*;
1089
1090* if the listening socket is in *non-blocking mode*, whether the socket
1091  returned by :meth:`~socket.accept` is in blocking or non-blocking mode
1092  is operating system-dependent.  If you want to ensure cross-platform
1093  behaviour, it is recommended you manually override this setting.
1094
1095
1096.. _socket-example:
1097
1098Example
1099-------
1100
1101Here are four minimal example programs using the TCP/IP protocol: a server that
1102echoes all data that it receives back (servicing only one client), and a client
1103using it.  Note that a server must perform the sequence :func:`socket`,
1104:meth:`~socket.bind`, :meth:`~socket.listen`, :meth:`~socket.accept` (possibly
1105repeating the :meth:`~socket.accept` to service more than one client), while a
1106client only needs the sequence :func:`socket`, :meth:`~socket.connect`.  Also
1107note that the server does not :meth:`~socket.send`/:meth:`~socket.recv` on the
1108socket it is listening on but on the new socket returned by
1109:meth:`~socket.accept`.
1110
1111The first two examples support IPv4 only. ::
1112
1113   # Echo server program
1114   import socket
1115
1116   HOST = ''                 # Symbolic name meaning all available interfaces
1117   PORT = 50007              # Arbitrary non-privileged port
1118   s = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
1119   s.bind((HOST, PORT))
1120   s.listen(1)
1121   conn, addr = s.accept()
1122   print('Connected by', addr)
1123   while True:
1124       data = conn.recv(1024)
1125       if not data: break
1126       conn.send(data)
1127   conn.close()
1128
1129::
1130
1131   # Echo client program
1132   import socket
1133
1134   HOST = 'daring.cwi.nl'    # The remote host
1135   PORT = 50007              # The same port as used by the server
1136   s = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
1137   s.connect((HOST, PORT))
1138   s.send(b'Hello, world')
1139   data = s.recv(1024)
1140   s.close()
1141   print('Received', repr(data))
1142
1143The next two examples are identical to the above two, but support both IPv4 and
1144IPv6. The server side will listen to the first address family available (it
1145should listen to both instead). On most of IPv6-ready systems, IPv6 will take
1146precedence and the server may not accept IPv4 traffic. The client side will try
1147to connect to the all addresses returned as a result of the name resolution, and
1148sends traffic to the first one connected successfully. ::
1149
1150   # Echo server program
1151   import socket
1152   import sys
1153
1154   HOST = None               # Symbolic name meaning all available interfaces
1155   PORT = 50007              # Arbitrary non-privileged port
1156   s = None
1157   for res in socket.getaddrinfo(HOST, PORT, socket.AF_UNSPEC,
1158                                 socket.SOCK_STREAM, 0, socket.AI_PASSIVE):
1159       af, socktype, proto, canonname, sa = res
1160       try:
1161           s = socket.socket(af, socktype, proto)
1162       except socket.error as msg:
1163           s = None
1164           continue
1165       try:
1166           s.bind(sa)
1167           s.listen(1)
1168       except socket.error as msg:
1169           s.close()
1170           s = None
1171           continue
1172       break
1173   if s is None:
1174       print('could not open socket')
1175       sys.exit(1)
1176   conn, addr = s.accept()
1177   print('Connected by', addr)
1178   while True:
1179       data = conn.recv(1024)
1180       if not data: break
1181       conn.send(data)
1182   conn.close()
1183
1184::
1185
1186   # Echo client program
1187   import socket
1188   import sys
1189
1190   HOST = 'daring.cwi.nl'    # The remote host
1191   PORT = 50007              # The same port as used by the server
1192   s = None
1193   for res in socket.getaddrinfo(HOST, PORT, socket.AF_UNSPEC, socket.SOCK_STREAM):
1194       af, socktype, proto, canonname, sa = res
1195       try:
1196           s = socket.socket(af, socktype, proto)
1197       except socket.error as msg:
1198           s = None
1199           continue
1200       try:
1201           s.connect(sa)
1202       except socket.error as msg:
1203           s.close()
1204           s = None
1205           continue
1206       break
1207   if s is None:
1208       print('could not open socket')
1209       sys.exit(1)
1210   s.send(b'Hello, world')
1211   data = s.recv(1024)
1212   s.close()
1213   print('Received', repr(data))
1214
1215
1216The last example shows how to write a very simple network sniffer with raw
1217sockets on Windows. The example requires administrator privileges to modify
1218the interface::
1219
1220   import socket
1221
1222   # the public network interface
1223   HOST = socket.gethostbyname(socket.gethostname())
1224
1225   # create a raw socket and bind it to the public interface
1226   s = socket.socket(socket.AF_INET, socket.SOCK_RAW, socket.IPPROTO_IP)
1227   s.bind((HOST, 0))
1228
1229   # Include IP headers
1230   s.setsockopt(socket.IPPROTO_IP, socket.IP_HDRINCL, 1)
1231
1232   # receive all packages
1233   s.ioctl(socket.SIO_RCVALL, socket.RCVALL_ON)
1234
1235   # receive a package
1236   print(s.recvfrom(65565))
1237
1238   # disabled promiscuous mode
1239   s.ioctl(socket.SIO_RCVALL, socket.RCVALL_OFF)
1240
1241
1242Running an example several times with too small delay between executions, could
1243lead to this error::
1244
1245   socket.error: [Errno 98] Address already in use
1246
1247This is because the previous execution has left the socket in a ``TIME_WAIT``
1248state, and can't be immediately reused.
1249
1250There is a :mod:`socket` flag to set, in order to prevent this,
1251:data:`socket.SO_REUSEADDR`::
1252
1253   s = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
1254   s.setsockopt(socket.SOL_SOCKET, socket.SO_REUSEADDR, 1)
1255   s.bind((HOST, PORT))
1256
1257the :data:`SO_REUSEADDR` flag tells the kernel to reuse a local socket in
1258``TIME_WAIT`` state, without waiting for its natural timeout to expire.
1259
1260
1261.. seealso::
1262
1263   For an introduction to socket programming (in C), see the following papers:
1264
1265   - *An Introductory 4.3BSD Interprocess Communication Tutorial*, by Stuart Sechrest
1266
1267   - *An Advanced 4.3BSD Interprocess Communication Tutorial*, by Samuel J.  Leffler et
1268     al,
1269
1270   bo…

Large files files are truncated, but you can click here to view the full file