socket —- 底层网络接口

源代码:Lib/socket.py


这个模块提供了访问BSD套接字的接口。在所有现代Unix系统、Windows、macOS和其他一些平台上可用。

注解

一些行为可能因平台不同而异,因为调用的是操作系统的套接字API。

这个Python接口是用Python的面向对象风格对Unix系统调用和套接字库接口的直译:函数 socket() 返回一个 套接字对象 ,其方法是对各种套接字系统调用的实现。形参类型一般与C接口相比更高级:例如在Python文件 read()write() 操作中,接收操作的缓冲区分配是自动的,发送操作的缓冲区长度是隐式的。

参见

  • 模块 socketserver
  • 用于简化网络服务端编写的类。

  • 模块 ssl

  • 套接字对象的TLS/SSL封装。

套接字协议族

根据系统以及构建选项,此模块提供了各种套接字协议簇。

特定的套接字对象需要的地址格式将根据此套接字对象被创建时指定的地址族被自动选择。套接字地址表示如下:

  • 一个绑定在文件系统节点上的 AF_UNIX 套接字的地址表示为一个字符串,使用文件系统字符编码和 'surrogateescape' 错误回调方法(see PEP 383)。一个地址在 Linux 的抽象命名空间被返回为带有初始的 null 字节的 字节类对象 ;注意在这个命名空间种的套接字可能与普通文件系统套接字通信,所以打算运行在 Linux 上的程序可能需要解决两种地址类型。当传递为参数时,一个字符串或字节类对象可以用于任一类型的地址。

在 3.3 版更改: 之前,AF_UNIX 套接字路径被假设使用 UTF-8 编码。

在 3.5 版更改: 现在支持可写的 字节类对象

  • 一对 (host, port) 被用于 AF_INET 地址族,host 是一个表示为互联网域名表示法之内的主机名或者一个 IPv4 地址的字符串,例如 'daring.cwi.nl''100.50.200.5'port 是一个整数。

    • 对于 IPv4 地址,有两种可接受的特殊形式被用来代替一个主机地址: '' 代表 INADDR_ANY,用来绑定到所有接口;字符串 '<broadcast>' 代表 INADDR_BROADCAST。此行为不兼容 IPv6,因此,如果你的 Python 程序打算支持 IPv6,则可能需要避开这些。
  • 对于 AF_INET6 地址族,使用一个四元组 (host, port, flowinfo, scopeid)flowinfoscopeid 代表了 C 库里 struct sockaddrin6 中的 sin6_flowinfosin6_scope_id 成员。对于 socket 模块中的方法, _flowinfoscopeid 可以被省略,只为了向后兼容。注意,scopeid 的省略可能会导致 problems in manipulating scoped IPv6 addresses。

在 3.7 版更改: For multicast addresses (with scopeid meaningful) address may not contain%scope (or zone id) part. This information is superfluous and maybe safely omitted (recommended).

  • AF_NETLINK sockets are represented as pairs (pid, groups).

  • Linux-only support for TIPC is available using the AF_TIPCaddress family. TIPC is an open, non-IP based networked protocol designedfor use in clustered computer environments. Addresses are represented by atuple, and the fields depend on the address type. The general tuple form is(addr_type, v1, v2, v3 [, scope]), where:

    • addr_type is one of TIPC_ADDR_NAMESEQ, TIPC_ADDR_NAME,or TIPC_ADDR_ID.

    • scope is one of TIPC_ZONE_SCOPE, TIPC_CLUSTER_SCOPE, andTIPC_NODE_SCOPE.

    • If addr_type is TIPCADDR_NAME, then _v1 is the server type, v2 isthe port identifier, and v3 should be 0.

If addr_type is TIPCADDR_NAMESEQ, then _v1 is the server type, v2_is the lower port number, and _v3 is the upper port number.

If addr_type is TIPCADDR_ID, then _v1 is the node, v2 is thereference, and v3 should be set to 0.

  • A tuple (interface, ) is used for the AF_CAN address family,where interface is a string representing a network interface name like'can0'. The network interface name '' can be used to receive packetsfrom all network interfaces of this family.

    • CAN_ISOTP protocol require a tuple (interface, rx_addr, tx_addr)where both additional parameters are unsigned long integer that represent aCAN identifier (standard or extended).
  • A string or a tuple (id, unit) is used for the SYSPROTO_CONTROLprotocol of the PF_SYSTEM family. The string is the name of akernel control using a dynamically-assigned ID. The tuple can be used if IDand unit number of the kernel control are known or if a registered ID isused.

3.3 新版功能.

  • AF_BLUETOOTH supports the following protocols and addressformats:

    • BTPROTO_L2CAP accepts (bdaddr, psm) where bdaddr isthe Bluetooth address as a string and psm is an integer.

    • BTPROTO_RFCOMM accepts (bdaddr, channel) where bdaddris the Bluetooth address as a string and channel is an integer.

    • BTPROTO_HCI accepts (device_id,) where device_id iseither an integer or a string with the Bluetooth address of theinterface. (This depends on your OS; NetBSD and DragonFlyBSD expecta Bluetooth address while everything else expects an integer.)

在 3.2 版更改: NetBSD and DragonFlyBSD support added.

  • BTPROTO_SCO accepts bdaddr where bdaddr is abytes object containing the Bluetooth address in astring format. (ex. b'12:23:34:45:56:67') This protocol is notsupported under FreeBSD.
  • AF_ALG is a Linux-only socket based interface to Kernelcryptography. An algorithm socket is configured with a tuple of two to fourelements (type, name [, feat [, mask]]), where:

    • type is the algorithm type as string, e.g. aead, hash,skcipher or rng.

    • name is the algorithm name and operation mode as string, e.g.sha256, hmac(sha256), cbc(aes) or drbg_nopr_ctr_aes256.

    • feat and mask are unsigned 32bit integers.

Availability: Linux 2.6.38, some algorithm types require more recent Kernels.

3.6 新版功能.

  • AF_VSOCK allows communication between virtual machines andtheir hosts. The sockets are represented as a (CID, port) tuplewhere the context ID or CID and port are integers.

Availability: Linux >= 4.8 QEMU >= 2.8 ESX >= 4.0 ESX Workstation >= 6.5.

3.7 新版功能.

  • AF_PACKET is a low-level interface directly to network devices.The packets are represented by the tuple(ifname, proto[, pkttype[, hatype[, addr]]]) where:

    • ifname - String specifying the device name.

    • proto - An in network-byte-order integer specifying the Ethernetprotocol number.

    • pkttype - Optional integer specifying the packet type:

      • PACKET_HOST (the default) - Packet addressed to the local host.

      • PACKET_BROADCAST - Physical-layer broadcast packet.

      • PACKET_MULTIHOST - Packet sent to a physical-layer multicast address.

      • PACKET_OTHERHOST - Packet to some other host that has been caught bya device driver in promiscuous mode.

      • PACKET_OUTGOING - Packet originating from the local host that islooped back to a packet socket.

    • hatype - Optional integer specifying the ARP hardware address type.

    • addr - Optional bytes-like object specifying the hardware physicaladdress, whose interpretation depends on the device.

  • AF_QIPCRTR is a Linux-only socket based interface for communicatingwith services running on co-processors in Qualcomm platforms. The addressfamily is represented as a (node, port) tuple where the node and _port_are non-negative integers.

3.8 新版功能.

如果你在 IPv4/v6 套接字地址的 host 部分中使用了一个主机名,此程序可能会表现不确定行为,因为 Python 使用 DNS 解析返回的第一个地址。套接字地址在实际的 IPv4/v6 中以不同方式解析,根据 DNS 解析和/或 host 配置。为了确定行为,在 host 部分中使用数字的地址。

所有的错误都抛出异常。对于无效的参数类型和内存溢出异常情况可能抛出普通异常;从 Python 3.3 开始,与套接字或地址语义有关的错误抛出 OSError 或它的子类之一(常用 socket.error)。

可以用 setblocking() 设置非阻塞模式。一个基于超时的 generalization 通过 settimeout() 支持。

模块内容

socket 模块导出以下元素。

异常

  • exception socket.error
  • 一个被弃用的 OSError 的别名。

在 3.3 版更改: 根据 PEP 3151,这个类是 OSError 的别名。

  • exception socket.herror
  • A subclass of OSError, this exception is raised foraddress-related errors, i.e. for functions that use h_errno in the POSIXC API, including gethostbyname_ex() and gethostbyaddr().The accompanying value is a pair (herrno, string) representing anerror returned by a library call. _h_errno is a numeric value, whilestring represents the description of h_errno, as returned by thehstrerror() C function.

在 3.3 版更改: This class was made a subclass of OSError.

  • exception socket.gaierror
  • A subclass of OSError, this exception is raised foraddress-related errors by getaddrinfo() and getnameinfo().The accompanying value is a pair (error, string) representing an errorreturned by a library call. string represents the description oferror, as returned by the gaistrerror() C function. Thenumeric _error value will match one of the EAI_* constantsdefined in this module.

在 3.3 版更改: This class was made a subclass of OSError.

  • exception socket.timeout
  • A subclass of OSError, this exception is raised when a timeoutoccurs on a socket which has had timeouts enabled via a prior call tosettimeout() (or implicitly throughsetdefaulttimeout()). The accompanying value is a stringwhose value is currently always "timed out".

在 3.3 版更改: This class was made a subclass of OSError.

常数

The AF* and SOCK* constants are now AddressFamily andSocketKind IntEnum collections.

3.4 新版功能.

  • socket.AF_UNIX
  • socket.AF_INET
  • socket.AF_INET6
  • These constants represent the address (and protocol) families, used for thefirst argument to socket(). If the AF_UNIX constant is notdefined then this protocol is unsupported. More constants may be availabledepending on the system.

  • socket.SOCK_STREAM

  • socket.SOCK_DGRAM
  • socket.SOCK_RAW
  • socket.SOCK_RDM
  • socket.SOCK_SEQPACKET
  • These constants represent the socket types, used for the second argument tosocket(). More constants may be available depending on the system.(Only SOCK_STREAM and SOCK_DGRAM appear to be generallyuseful.)

  • socket.SOCK_CLOEXEC

  • socket.SOCK_NONBLOCK
  • These two constants, if defined, can be combined with the socket types andallow you to set some flags atomically (thus avoiding possible raceconditions and the need for separate calls).

参见

Secure File Descriptor Handlingfor a more thorough explanation.

Availability: Linux >= 2.6.27.

3.2 新版功能.

  • SO_*
  • socket.SOMAXCONN
  • MSG_*
  • SOL_*
  • SCM_*
  • IPPROTO_*
  • IPPORT_*
  • INADDR_*
  • IP_*
  • IPV6_*
  • EAI_*
  • AI_*
  • NI_*
  • TCP_*
  • Many constants of these forms, documented in the Unix documentation on socketsand/or the IP protocol, are also defined in the socket module. They aregenerally used in arguments to the setsockopt() and getsockopt()methods of socket objects. In most cases, only those symbols that are definedin the Unix header files are defined; for a few symbols, default values areprovided.

在 3.6 版更改: SO_DOMAIN, SO_PROTOCOL, SO_PEERSEC, SO_PASSSEC,TCP_USER_TIMEOUT, TCP_CONGESTION were added.

在 3.6.5 版更改: On Windows, TCP_FASTOPEN, TCP_KEEPCNT appear if run-time Windowssupports.

在 3.7 版更改: TCP_NOTSENT_LOWAT was added.

On Windows, TCP_KEEPIDLE, TCP_KEEPINTVL appear if run-time Windowssupports.

  • socket.AF_CAN
  • socket.PF_CAN
  • SOLCAN*
  • CAN_*
  • Many constants of these forms, documented in the Linux documentation, arealso defined in the socket module.

Availability: Linux >= 2.6.25.

3.3 新版功能.

  • socket.CAN_BCM
  • CANBCM*
  • CAN_BCM, in the CAN protocol family, is the broadcast manager (BCM) protocol.Broadcast manager constants, documented in the Linux documentation, are alsodefined in the socket module.

Availability: Linux >= 2.6.25.

注解

The CAN_BCM_CAN_FD_FRAME flag is only available on Linux >= 4.8.

3.4 新版功能.

  • socket.CAN_RAW_FD_FRAMES
  • Enables CAN FD support in a CAN_RAW socket. This is disabled by default.This allows your application to send both CAN and CAN FD frames; however,you must accept both CAN and CAN FD frames when reading from the socket.

This constant is documented in the Linux documentation.

Availability: Linux >= 3.6.

3.5 新版功能.

  • socket.CAN_ISOTP
  • CAN_ISOTP, in the CAN protocol family, is the ISO-TP (ISO 15765-2) protocol.ISO-TP constants, documented in the Linux documentation.

Availability: Linux >= 2.6.25.

3.7 新版功能.

  • socket.AF_PACKET
  • socket.PF_PACKET
  • PACKET_*
  • Many constants of these forms, documented in the Linux documentation, arealso defined in the socket module.

Availability: Linux >= 2.2.

  • socket.AF_RDS
  • socket.PF_RDS
  • socket.SOL_RDS
  • RDS_*
  • Many constants of these forms, documented in the Linux documentation, arealso defined in the socket module.

Availability: Linux >= 2.6.30.

3.3 新版功能.

  • socket.SIO_RCVALL
  • socket.SIO_KEEPALIVE_VALS
  • socket.SIO_LOOPBACK_FAST_PATH
  • RCVALL_*
  • Constants for Windows' WSAIoctl(). The constants are used as arguments to theioctl() method of socket objects.

在 3.6 版更改: SIO_LOOPBACK_FAST_PATH was added.

  • TIPC_*
  • TIPC related constants, matching the ones exported by the C socket API. Seethe TIPC documentation for more information.

  • socket.AF_ALG

  • socket.SOL_ALG
  • ALG_*
  • Constants for Linux Kernel cryptography.

Availability: Linux >= 2.6.38.

3.6 新版功能.

  • socket.AF_VSOCK
  • socket.IOCTL_VM_SOCKETS_GET_LOCAL_CID
  • VMADDR*
  • SO_VM*
  • Constants for Linux host/guest communication.

Availability: Linux >= 4.8.

3.7 新版功能.

3.4 新版功能.

  • socket.has_ipv6
  • This constant contains a boolean value which indicates if IPv6 is supported onthis platform.

  • socket.BDADDR_ANY

  • socket.BDADDR_LOCAL
  • These are string constants containing Bluetooth addresses with specialmeanings. For example, BDADDR_ANY can be used to indicateany address when specifying the binding socket withBTPROTO_RFCOMM.

  • socket.HCI_FILTER

  • socket.HCI_TIME_STAMP
  • socket.HCI_DATA_DIR
  • For use with BTPROTO_HCI. HCI_FILTER is notavailable for NetBSD or DragonFlyBSD. HCI_TIME_STAMP andHCI_DATA_DIR are not available for FreeBSD, NetBSD, orDragonFlyBSD.

  • socket.AF_QIPCRTR

  • Constant for Qualcomm's IPC router protocol, used to communicate withservice providing remote processors.

Availability: Linux >= 4.7.

函数

Creating sockets

The following functions all create socket objects.

  • socket.socket(family=AF_INET, type=SOCK_STREAM, proto=0, fileno=None)
  • Create a new socket using the given address family, socket type and protocolnumber. The address family should be AF_INET (the default),AF_INET6, AF_UNIX, AF_CAN, AF_PACKET,or AF_RDS. The socket type should be SOCK_STREAM (thedefault), SOCK_DGRAM, SOCK_RAW or perhaps one of the otherSOCK_ constants. The protocol number is usually zero and may be omittedor in the case where the address family is AF_CAN the protocolshould be one of CAN_RAW, CAN_BCM or CAN_ISOTP.

If fileno is specified, the values for family, type, and proto areauto-detected from the specified file descriptor. Auto-detection can beoverruled by calling the function with explicit family, type, or proto_arguments. This only affects how Python represents e.g. the return valueof socket.getpeername() but not the actual OS resource. Unlikesocket.fromfd(), _fileno will return the same socket and not aduplicate. This may help close a detached socket usingsocket.close().

The newly created socket is non-inheritable.

Raises an auditing event socket.new with arguments self, family, type, protocol.

在 3.3 版更改: The AF_CAN family was added.The AF_RDS family was added.

在 3.4 版更改: The CAN_BCM protocol was added.

在 3.4 版更改: The returned socket is now non-inheritable.

在 3.7 版更改: The CAN_ISOTP protocol was added.

在 3.7 版更改: When SOCK_NONBLOCK or SOCK_CLOEXECbit flags are applied to type they are cleared, andsocket.type will not reflect them. They are still passedto the underlying system socket() call. Therefore::

sock = socket.socket(

socket.AF_INET,socket.SOCK_STREAM | socket.SOCK_NONBLOCK)

will still create a non-blocking socket on OSes that supportSOCK_NONBLOCK, but sock.type will be set tosocket.SOCK_STREAM.

  • socket.socketpair([family[, type[, proto]]])
  • Build a pair of connected socket objects using the given address family, sockettype, and protocol number. Address family, socket type, and protocol number areas for the socket() function above. The default family is AF_UNIXif defined on the platform; otherwise, the default is AF_INET.

The newly created sockets are non-inheritable.

在 3.2 版更改: The returned socket objects now support the whole socket API, ratherthan a subset.

在 3.4 版更改: The returned sockets are now non-inheritable.

在 3.5 版更改: Windows support added.

  • socket.createconnection(_address[, timeout[, source_address]])
  • Connect to a TCP service listening on the Internet address (a 2-tuple(host, port)), and return the socket object. This is a higher-levelfunction than socket.connect(): if host is a non-numeric hostname,it will try to resolve it for both AF_INET and AF_INET6,and then try to connect to all possible addresses in turn until aconnection succeeds. This makes it easy to write clients that arecompatible to both IPv4 and IPv6.

Passing the optional timeout parameter will set the timeout on thesocket instance before attempting to connect. If no timeout issupplied, the global default timeout setting returned bygetdefaulttimeout() is used.

If supplied, source_address must be a 2-tuple (host, port) for thesocket to bind to as its source address before connecting. If host or portare '' or 0 respectively the OS default behavior will be used.

在 3.2 版更改: 添加了 source_address

  • socket.createserver(_address, *, family=AF_INET, backlog=None, reuse_port=False, dualstack_ipv6=False)
  • Convenience function which creates a TCP socket bound to address (a 2-tuple(host, port)) and return the socket object.

family should be either AF_INET or AF_INET6.backlog is the queue size passed to socket.listen(); when 0a default reasonable value is chosen.reuse_port dictates whether to set the SO_REUSEPORT socket option.

If dualstack_ipv6 is true and the platform supports it the socket willbe able to accept both IPv4 and IPv6 connections, else it will raiseValueError. Most POSIX platforms and Windows are supposed to supportthis functionality.When this functionality is enabled the address returned bysocket.getpeername() when an IPv4 connection occurs will be an IPv6address represented as an IPv4-mapped IPv6 address.If dualstack_ipv6 is false it will explicitly disable this functionalityon platforms that enable it by default (e.g. Linux).This parameter can be used in conjunction with has_dualstack_ipv6():

  1. import socket
  2.  
  3. addr = ("", 8080) # all interfaces, port 8080
  4. if socket.has_dualstack_ipv6():
  5. s = socket.create_server(addr, family=socket.AF_INET6, dualstack_ipv6=True)
  6. else:
  7. s = socket.create_server(addr)

注解

On POSIX platforms the SO_REUSEADDR socket option is set in order toimmediately reuse previous sockets which were bound on the same _address_and remained in TIME_WAIT state.

3.8 新版功能.

  • socket.has_dualstack_ipv6()
  • Return True if the platform supports creating a TCP socket which canhandle both IPv4 and IPv6 connections.

3.8 新版功能.

  • socket.fromfd(fd, family, type, proto=0)
  • Duplicate the file descriptor fd (an integer as returned by a file object'sfileno() method) and build a socket object from the result. Addressfamily, socket type and protocol number are as for the socket() functionabove. The file descriptor should refer to a socket, but this is not checked —-subsequent operations on the object may fail if the file descriptor is invalid.This function is rarely needed, but can be used to get or set socket options ona socket passed to a program as standard input or output (such as a serverstarted by the Unix inet daemon). The socket is assumed to be in blocking mode.

The newly created socket is non-inheritable.

在 3.4 版更改: The returned socket is now non-inheritable.

  • socket.fromshare(data)
  • Instantiate a socket from data obtained from the socket.share()method. The socket is assumed to be in blocking mode.

可用性: Windows。

3.3 新版功能.

  • socket.SocketType
  • This is a Python type object that represents the socket object type. It is thesame as type(socket(…)).

其他功能

The socket module also offers various network-related services:

  • socket.close(fd)
  • Close a socket file descriptor. This is like os.close(), but forsockets. On some platforms (most noticeable Windows) os.close()does not work for socket file descriptors.

3.7 新版功能.

  • socket.getaddrinfo(host, port, family=0, type=0, proto=0, flags=0)
  • Translate the host/port argument into a sequence of 5-tuples that containall the necessary arguments for creating a socket connected to that service.host is a domain name, a string representation of an IPv4/v6 addressor None. port is a string service name such as 'http', a numericport number or None. By passing None as the value of host_and _port, you can pass NULL to the underlying C API.

The family, type and proto arguments can be optionally specifiedin order to narrow the list of addresses returned. Passing zero as avalue for each of these arguments selects the full range of results.The flags argument can be one or several of the AI* constants,and will influence how results are computed and returned.For example, AI_NUMERICHOST will disable domain name resolutionand will raise an error if _host is a domain name.

The function returns a list of 5-tuples with the following structure:

(family, type, proto, canonname, sockaddr)

In these tuples, family, type, proto are all integers and aremeant to be passed to the socket() function. canonname will bea string representing the canonical name of the host ifAICANONNAME is part of the _flags argument; else canonname_will be empty. _sockaddr is a tuple describing a socket address, whoseformat depends on the returned family (a (address, port) 2-tuple forAF_INET, a (address, port, flow info, scope id) 4-tuple forAF_INET6), and is meant to be passed to the socket.connect()method.

Raises an auditing event socket.getaddrinfo with arguments host, port, family, type, protocol.

The following example fetches address information for a hypothetical TCPconnection to example.org on port 80 (results may differ on yoursystem if IPv6 isn't enabled):

  1. >>> socket.getaddrinfo("example.org", 80, proto=socket.IPPROTO_TCP)
  2. [(<AddressFamily.AF_INET6: 10>, <SocketType.SOCK_STREAM: 1>,
  3. 6, '', ('2606:2800:220:1:248:1893:25c8:1946', 80, 0, 0)),
  4. (<AddressFamily.AF_INET: 2>, <SocketType.SOCK_STREAM: 1>,
  5. 6, '', ('93.184.216.34', 80))]

在 3.2 版更改: parameters can now be passed using keyword arguments.

在 3.7 版更改: for IPv6 multicast addresses, string representing an address will notcontain %scope part.

  • socket.getfqdn([name])
  • Return a fully qualified domain name for name. If name is omitted or empty,it is interpreted as the local host. To find the fully qualified name, thehostname returned by gethostbyaddr() is checked, followed by aliases for thehost, if available. The first name which includes a period is selected. Incase no fully qualified domain name is available, the hostname as returned bygethostname() is returned.

  • socket.gethostbyname(hostname)

  • Translate a host name to IPv4 address format. The IPv4 address is returned as astring, such as '100.50.200.5'. If the host name is an IPv4 address itselfit is returned unchanged. See gethostbyname_ex() for a more completeinterface. gethostbyname() does not support IPv6 name resolution, andgetaddrinfo() should be used instead for IPv4/v6 dual stack support.

Raises an auditing event socket.gethostbyname with argument hostname.

  • socket.gethostbynameex(_hostname)
  • Translate a host name to IPv4 address format, extended interface. Return atriple (hostname, aliaslist, ipaddrlist) where hostname is the primaryhost name responding to the given ip_address, aliaslist is a (possiblyempty) list of alternative host names for the same address, and ipaddrlist isa list of IPv4 addresses for the same interface on the same host (often but notalways a single address). gethostbyname_ex() does not support IPv6 nameresolution, and getaddrinfo() should be used instead for IPv4/v6 dualstack support.

Raises an auditing event socket.gethostbyname with argument hostname.

  • socket.gethostname()
  • Return a string containing the hostname of the machine where the Pythoninterpreter is currently executing.

Raises an auditing event socket.gethostname with no arguments.

Note: gethostname() doesn't always return the fully qualified domainname; use getfqdn() for that.

  • socket.gethostbyaddr(ip_address)
  • Return a triple (hostname, aliaslist, ipaddrlist) where hostname is theprimary host name responding to the given ip_address, aliaslist is a(possibly empty) list of alternative host names for the same address, andipaddrlist is a list of IPv4/v6 addresses for the same interface on the samehost (most likely containing only a single address). To find the fully qualifieddomain name, use the function getfqdn(). gethostbyaddr() supportsboth IPv4 and IPv6.

Raises an auditing event socket.gethostbyaddr with argument ip_address.

  • socket.getnameinfo(sockaddr, flags)
  • Translate a socket address sockaddr into a 2-tuple (host, port). Dependingon the settings of flags, the result can contain a fully-qualified domain nameor numeric address representation in host. Similarly, port can contain astring port name or a numeric port number.

For IPv6 addresses, %scope is appended to the host part if sockaddr_contains meaningful _scopeid. Usually this happens for multicast addresses.

For more information about flags you can consult getnameinfo(3)).

Raises an auditing event socket.getnameinfo with argument sockaddr.

  • socket.getprotobyname(protocolname)
  • Translate an Internet protocol name (for example, 'icmp') to a constantsuitable for passing as the (optional) third argument to the socket()function. This is usually only needed for sockets opened in "raw" mode(SOCK_RAW); for the normal socket modes, the correct protocol is chosenautomatically if the protocol is omitted or zero.

  • socket.getservbyname(servicename[, protocolname])

  • Translate an Internet service name and protocol name to a port number for thatservice. The optional protocol name, if given, should be 'tcp' or'udp', otherwise any protocol will match.

Raises an auditing event socket.getservbyname with arguments servicename, protocolname.

  • socket.getservbyport(port[, protocolname])
  • Translate an Internet port number and protocol name to a service name for thatservice. The optional protocol name, if given, should be 'tcp' or'udp', otherwise any protocol will match.

Raises an auditing event socket.getservbyport with arguments port, protocolname.

  • socket.ntohl(x)
  • Convert 32-bit positive integers from network to host byte order. On machineswhere the host byte order is the same as network byte order, this is a no-op;otherwise, it performs a 4-byte swap operation.

  • socket.ntohs(x)

  • Convert 16-bit positive integers from network to host byte order. On machineswhere the host byte order is the same as network byte order, this is a no-op;otherwise, it performs a 2-byte swap operation.

3.7 版后已移除: In case x does not fit in 16-bit unsigned integer, but does fit in apositive C int, it is silently truncated to 16-bit unsigned integer.This silent truncation feature is deprecated, and will raise anexception in future versions of Python.

  • socket.htonl(x)
  • Convert 32-bit positive integers from host to network byte order. On machineswhere the host byte order is the same as network byte order, this is a no-op;otherwise, it performs a 4-byte swap operation.

  • socket.htons(x)

  • Convert 16-bit positive integers from host to network byte order. On machineswhere the host byte order is the same as network byte order, this is a no-op;otherwise, it performs a 2-byte swap operation.

3.7 版后已移除: In case x does not fit in 16-bit unsigned integer, but does fit in apositive C int, it is silently truncated to 16-bit unsigned integer.This silent truncation feature is deprecated, and will raise anexception in future versions of Python.

  • socket.inetaton(_ip_string)
  • Convert an IPv4 address from dotted-quad string format (for example,'123.45.67.89') to 32-bit packed binary format, as a bytes object four characters inlength. This is useful when conversing with a program that uses the standard Clibrary and needs objects of type struct in_addr, which is the C typefor the 32-bit packed binary this function returns.

inet_aton() also accepts strings with less than three dots; see theUnix manual page inet(3)) for details.

If the IPv4 address string passed to this function is invalid,OSError will be raised. Note that exactly what is valid depends onthe underlying C implementation of inet_aton().

inet_aton() does not support IPv6, and inet_pton() should be usedinstead for IPv4/v6 dual stack support.

  • socket.inetntoa(_packed_ip)
  • Convert a 32-bit packed IPv4 address (a bytes-like object fourbytes in length) to its standard dotted-quad string representation (for example,'123.45.67.89'). This is useful when conversing with a program that uses thestandard C library and needs objects of type struct in_addr, whichis the C type for the 32-bit packed binary data this function takes as anargument.

If the byte sequence passed to this function is not exactly 4 bytes inlength, OSError will be raised. inet_ntoa() does notsupport IPv6, and inet_ntop() should be used instead for IPv4/v6 dualstack support.

在 3.5 版更改: 现在支持可写的 字节类对象

  • socket.inetpton(_address_family, ip_string)
  • Convert an IP address from its family-specific string format to a packed,binary format. inet_pton() is useful when a library or network protocolcalls for an object of type struct in_addr (similar toinet_aton()) or struct in6_addr.

Supported values for address_family are currently AF_INET andAF_INET6. If the IP address string ip_string is invalid,OSError will be raised. Note that exactly what is valid depends onboth the value of address_family and the underlying implementation ofinet_pton().

Availability: Unix (maybe not all platforms), Windows.

在 3.4 版更改: Windows support added

  • socket.inetntop(_address_family, packed_ip)
  • Convert a packed IP address (a bytes-like object of some number ofbytes) to its standard, family-specific string representation (forexample, '7.10.0.5' or '5aef:2b::8').inet_ntop() is useful when a library or network protocol returns anobject of type struct in_addr (similar to inet_ntoa()) orstruct in6_addr.

Supported values for address_family are currently AF_INET andAF_INET6. If the bytes object packed_ip is not the correctlength for the specified address family, ValueError will be raised.OSError is raised for errors from the call to inet_ntop().

Availability: Unix (maybe not all platforms), Windows.

在 3.4 版更改: Windows support added

在 3.5 版更改: 现在支持可写的 字节类对象

  • socket.CMSGLEN(_length)
  • Return the total length, without trailing padding, of an ancillarydata item with associated data of the given length. This valuecan often be used as the buffer size for recvmsg() toreceive a single item of ancillary data, but RFC 3542 requiresportable applications to use CMSG_SPACE() and thus includespace for padding, even when the item will be the last in thebuffer. Raises OverflowError if length is outside thepermissible range of values.

Availability: most Unix platforms, possibly others.

3.3 新版功能.

  • socket.CMSGSPACE(_length)
  • Return the buffer size needed for recvmsg() toreceive an ancillary data item with associated data of the givenlength, along with any trailing padding. The buffer space neededto receive multiple items is the sum of the CMSG_SPACE()values for their associated data lengths. RaisesOverflowError if length is outside the permissible rangeof values.

Note that some systems might support ancillary data withoutproviding this function. Also note that setting the buffer sizeusing the results of this function may not precisely limit theamount of ancillary data that can be received, since additionaldata may be able to fit into the padding area.

Availability: most Unix platforms, possibly others.

3.3 新版功能.

  • socket.getdefaulttimeout()
  • Return the default timeout in seconds (float) for new socket objects. A valueof None indicates that new socket objects have no timeout. When the socketmodule is first imported, the default is None.

  • socket.setdefaulttimeout(timeout)

  • Set the default timeout in seconds (float) for new socket objects. Whenthe socket module is first imported, the default is None. Seesettimeout() for possible values and their respectivemeanings.

  • socket.sethostname(name)

  • Set the machine's hostname to name. This will raise anOSError if you don't have enough rights.

Raises an auditing event socket.sethostname with argument name.

Availability: Unix.

3.3 新版功能.

  • socket.if_nameindex()
  • Return a list of network interface information(index int, name string) tuples.OSError if the system call fails.

可用性: Unix, Windows。

3.3 新版功能.

在 3.8 版更改: Windows support was added.

  • socket.ifnametoindex(_if_name)
  • Return a network interface index number corresponding to aninterface name.OSError if no interface with the given name exists.

可用性: Unix, Windows。

3.3 新版功能.

在 3.8 版更改: Windows support was added.

  • socket.ifindextoname(_if_index)
  • Return a network interface name corresponding to aninterface index number.OSError if no interface with the given index exists.

可用性: Unix, Windows。

3.3 新版功能.

在 3.8 版更改: Windows support was added.

Socket Objects

Socket objects have the following methods. Except formakefile(), these correspond to Unix system calls applicableto sockets.

在 3.2 版更改: Support for the context manager protocol was added. Exiting thecontext manager is equivalent to calling close().

  • socket.accept()
  • Accept a connection. The socket must be bound to an address and listening forconnections. The return value is a pair (conn, address) where conn is anew socket object usable to send and receive data on the connection, andaddress is the address bound to the socket on the other end of the connection.

The newly created socket is non-inheritable.

在 3.4 版更改: The socket is now non-inheritable.

在 3.5 版更改: If the system call is interrupted and the signal handler does not raisean exception, the method now retries the system call instead of raisingan InterruptedError exception (see PEP 475 for the rationale).

  • socket.bind(address)
  • Bind the socket to address. The socket must not already be bound. (The formatof address depends on the address family —- see above.)

Raises an auditing event socket.bind with arguments self, address.

  • socket.close()
  • Mark the socket closed. The underlying system resource (e.g. a filedescriptor) is also closed when all file objects from makefile()are closed. Once that happens, all future operations on the socketobject will fail. The remote end will receive no more data (afterqueued data is flushed).

Sockets are automatically closed when they are garbage-collected, butit is recommended to close() them explicitly, or to use awith statement around them.

在 3.6 版更改: OSError is now raised if an error occurs when the underlyingclose() call is made.

注解

close() releases the resource associated with a connection butdoes not necessarily close the connection immediately. If you wantto close the connection in a timely fashion, call shutdown()before close().

  • socket.connect(address)
  • Connect to a remote socket at address. (The format of address depends on theaddress family —- see above.)

If the connection is interrupted by a signal, the method waits until theconnection completes, or raise a socket.timeout on timeout, if thesignal handler doesn't raise an exception and the socket is blocking or hasa timeout. For non-blocking sockets, the method raises anInterruptedError exception if the connection is interrupted by asignal (or the exception raised by the signal handler).

Raises an auditing event socket.connect with arguments self, address.

在 3.5 版更改: The method now waits until the connection completes instead of raising anInterruptedError exception if the connection is interrupted by asignal, the signal handler doesn't raise an exception and the socket isblocking or has a timeout (see the PEP 475 for the rationale).

  • socket.connectex(_address)
  • Like connect(address), but return an error indicator instead of raising anexception for errors returned by the C-level connect() call (otherproblems, such as "host not found," can still raise exceptions). The errorindicator is 0 if the operation succeeded, otherwise the value of theerrno variable. This is useful to support, for example, asynchronousconnects.

Raises an auditing event socket.connect with arguments self, address.

  • socket.detach()
  • Put the socket object into closed state without actually closing theunderlying file descriptor. The file descriptor is returned, and canbe reused for other purposes.

3.2 新版功能.

  • socket.dup()
  • Duplicate the socket.

The newly created socket is non-inheritable.

在 3.4 版更改: The socket is now non-inheritable.

  • socket.fileno()
  • Return the socket's file descriptor (a small integer), or -1 on failure. Thisis useful with select.select().

Under Windows the small integer returned by this method cannot be used where afile descriptor can be used (such as os.fdopen()). Unix does not havethis limitation.

  • socket.get_inheritable()
  • Get the inheritable flag of the socket's filedescriptor or socket's handle: True if the socket can be inherited inchild processes, False if it cannot.

3.4 新版功能.

  • socket.getpeername()
  • Return the remote address to which the socket is connected. This is useful tofind out the port number of a remote IPv4/v6 socket, for instance. (The formatof the address returned depends on the address family —- see above.) On somesystems this function is not supported.

  • socket.getsockname()

  • Return the socket's own address. This is useful to find out the port number ofan IPv4/v6 socket, for instance. (The format of the address returned depends onthe address family —- see above.)

  • socket.getsockopt(level, optname[, buflen])

  • Return the value of the given socket option (see the Unix man pagegetsockopt(2))). The needed symbolic constants (SO* etc.)are defined in this module. If _buflen is absent, an integer option is assumedand its integer value is returned by the function. If buflen is present, itspecifies the maximum length of the buffer used to receive the option in, andthis buffer is returned as a bytes object. It is up to the caller to decode thecontents of the buffer (see the optional built-in module struct for a wayto decode C structures encoded as byte strings).

  • socket.getblocking()

  • Return True if socket is in blocking mode, False if innon-blocking.

This is equivalent to checking socket.gettimeout() == 0.

3.7 新版功能.

  • socket.gettimeout()
  • Return the timeout in seconds (float) associated with socket operations,or None if no timeout is set. This reflects the last call tosetblocking() or settimeout().

  • socket.ioctl(control, option)

    • Platform
    • Windows

The ioctl() method is a limited interface to the WSAIoctl systeminterface. Please refer to the Win32 documentation for moreinformation.

On other platforms, the generic fcntl.fcntl() and fcntl.ioctl()functions may be used; they accept a socket object as their first argument.

Currently only the following control codes are supported:SIO_RCVALL, SIO_KEEPALIVE_VALS, and SIO_LOOPBACK_FAST_PATH.

在 3.6 版更改: SIO_LOOPBACK_FAST_PATH was added.

  • socket.listen([backlog])
  • Enable a server to accept connections. If backlog is specified, it mustbe at least 0 (if it is lower, it is set to 0); it specifies the number ofunaccepted connections that the system will allow before refusing newconnections. If not specified, a default reasonable value is chosen.

在 3.5 版更改: The backlog parameter is now optional.

  • socket.makefile(mode='r', buffering=None, *, encoding=None, errors=None, newline=None)
  • Return a file object associated with the socket. The exact returnedtype depends on the arguments given to makefile(). These arguments areinterpreted the same way as by the built-in open() function, exceptthe only supported mode values are 'r' (default), 'w' and 'b'.

The socket must be in blocking mode; it can have a timeout, but the fileobject's internal buffer may end up in an inconsistent state if a timeoutoccurs.

Closing the file object returned by makefile() won't close theoriginal socket unless all other file objects have been closed andsocket.close() has been called on the socket object.

注解

On Windows, the file-like object created by makefile() cannot beused where a file object with a file descriptor is expected, such as thestream arguments of subprocess.Popen().

  • socket.recv(bufsize[, flags])
  • Receive data from the socket. The return value is a bytes object representing thedata received. The maximum amount of data to be received at once is specifiedby bufsize. See the Unix manual page recv(2)) for the meaning ofthe optional argument flags; it defaults to zero.

注解

For best match with hardware and network realities, the value of _bufsize_should be a relatively small power of 2, for example, 4096.

在 3.5 版更改: If the system call is interrupted and the signal handler does not raisean exception, the method now retries the system call instead of raisingan InterruptedError exception (see PEP 475 for the rationale).

  • socket.recvfrom(bufsize[, flags])
  • Receive data from the socket. The return value is a pair (bytes, address)where bytes is a bytes object representing the data received and address is theaddress of the socket sending the data. See the Unix manual pagerecv(2)) for the meaning of the optional argument flags; it defaultsto zero. (The format of address depends on the address family —- see above.)

在 3.5 版更改: If the system call is interrupted and the signal handler does not raisean exception, the method now retries the system call instead of raisingan InterruptedError exception (see PEP 475 for the rationale).

在 3.7 版更改: For multicast IPv6 address, first item of address does not contain%scope part anymore. In order to get full IPv6 address usegetnameinfo().

  • socket.recvmsg(bufsize[, ancbufsize[, flags]])
  • Receive normal data (up to bufsize bytes) and ancillary data fromthe socket. The ancbufsize argument sets the size in bytes ofthe internal buffer used to receive the ancillary data; it defaultsto 0, meaning that no ancillary data will be received. Appropriatebuffer sizes for ancillary data can be calculated usingCMSG_SPACE() or CMSG_LEN(), and items which do not fitinto the buffer might be truncated or discarded. The _flags_argument defaults to 0 and has the same meaning as forrecv().

The return value is a 4-tuple: (data, ancdata, msgflags,address). The _data item is a bytes object holding thenon-ancillary data received. The ancdata item is a list of zeroor more tuples (cmsglevel, cmsg_type, cmsg_data) representingthe ancillary data (control messages) received: _cmsg_level andcmsg_type are integers specifying the protocol level andprotocol-specific type respectively, and cmsg_data is abytes object holding the associated data. The msg_flags_item is the bitwise OR of various flags indicating conditions onthe received message; see your system documentation for details.If the receiving socket is unconnected, _address is the address ofthe sending socket, if available; otherwise, its value isunspecified.

On some systems, sendmsg() and recvmsg() can be used topass file descriptors between processes over an AF_UNIXsocket. When this facility is used (it is often restricted toSOCK_STREAM sockets), recvmsg() will return, in itsancillary data, items of the form (socket.SOLSOCKET,socket.SCM_RIGHTS, fds), where _fds is a bytes objectrepresenting the new file descriptors as a binary array of thenative C int type. If recvmsg() raises anexception after the system call returns, it will first attempt toclose any file descriptors received via this mechanism.

Some systems do not indicate the truncated length of ancillary dataitems which have been only partially received. If an item appearsto extend beyond the end of the buffer, recvmsg() will issuea RuntimeWarning, and will return the part of it which isinside the buffer provided it has not been truncated before thestart of its associated data.

On systems which support the SCMRIGHTS mechanism, thefollowing function will receive up to _maxfds file descriptors,returning the message data and a list containing the descriptors(while ignoring unexpected conditions such as unrelated controlmessages being received). See also sendmsg().

  1. import socket, array
  2.  
  3. def recv_fds(sock, msglen, maxfds):
  4. fds = array.array("i") # Array of ints
  5. msg, ancdata, flags, addr = sock.recvmsg(msglen, socket.CMSG_LEN(maxfds * fds.itemsize))
  6. for cmsg_level, cmsg_type, cmsg_data in ancdata:
  7. if cmsg_level == socket.SOL_SOCKET and cmsg_type == socket.SCM_RIGHTS:
  8. # Append data, ignoring any truncated integers at the end.
  9. fds.frombytes(cmsg_data[:len(cmsg_data) - (len(cmsg_data) % fds.itemsize)])
  10. return msg, list(fds)

Availability: most Unix platforms, possibly others.

3.3 新版功能.

在 3.5 版更改: If the system call is interrupted and the signal handler does not raisean exception, the method now retries the system call instead of raisingan InterruptedError exception (see PEP 475 for the rationale).

  • socket.recvmsginto(_buffers[, ancbufsize[, flags]])
  • Receive normal data and ancillary data from the socket, behaving asrecvmsg() would, but scatter the non-ancillary data into aseries of buffers instead of returning a new bytes object. Thebuffers argument must be an iterable of objects that exportwritable buffers (e.g. bytearray objects); these will befilled with successive chunks of the non-ancillary data until ithas all been written or there are no more buffers. The operatingsystem may set a limit (sysconf() value SCIOV_MAX)on the number of buffers that can be used. The _ancbufsize andflags arguments have the same meaning as for recvmsg().

The return value is a 4-tuple: (nbytes, ancdata, msgflags,address), where _nbytes is the total number of bytes ofnon-ancillary data written into the buffers, and ancdata,msg_flags and address are the same as for recvmsg().

示例:

  1. >>> import socket
  2. >>> s1, s2 = socket.socketpair()
  3. >>> b1 = bytearray(b'----')
  4. >>> b2 = bytearray(b'0123456789')
  5. >>> b3 = bytearray(b'--------------')
  6. >>> s1.send(b'Mary had a little lamb')
  7. 22
  8. >>> s2.recvmsg_into([b1, memoryview(b2)[2:9], b3])
  9. (22, [], 0, None)
  10. >>> [b1, b2, b3]
  11. [bytearray(b'Mary'), bytearray(b'01 had a 9'), bytearray(b'little lamb---')]

Availability: most Unix platforms, possibly others.

3.3 新版功能.

  • socket.recvfrominto(_buffer[, nbytes[, flags]])
  • Receive data from the socket, writing it into buffer instead of creating anew bytestring. The return value is a pair (nbytes, address) where nbytes isthe number of bytes received and address is the address of the socket sendingthe data. See the Unix manual page recv(2)) for the meaning of theoptional argument flags; it defaults to zero. (The format of _address_depends on the address family —- see above.)

  • socket.recvinto(_buffer[, nbytes[, flags]])

  • Receive up to nbytes bytes from the socket, storing the data into a bufferrather than creating a new bytestring. If nbytes is not specified (or 0),receive up to the size available in the given buffer. Returns the number ofbytes received. See the Unix manual page recv(2)) for the meaningof the optional argument flags; it defaults to zero.

  • socket.send(bytes[, flags])

  • Send data to the socket. The socket must be connected to a remote socket. Theoptional flags argument has the same meaning as for recv() above.Returns the number of bytes sent. Applications are responsible for checking thatall data has been sent; if only some of the data was transmitted, theapplication needs to attempt delivery of the remaining data. For furtherinformation on this topic, consult the 套接字编程指南.

在 3.5 版更改: If the system call is interrupted and the signal handler does not raisean exception, the method now retries the system call instead of raisingan InterruptedError exception (see PEP 475 for the rationale).

  • socket.sendall(bytes[, flags])
  • Send data to the socket. The socket must be connected to a remote socket. Theoptional flags argument has the same meaning as for recv() above.Unlike send(), this method continues to send data from bytes untileither all data has been sent or an error occurs. None is returned onsuccess. On error, an exception is raised, and there is no way to determine howmuch data, if any, was successfully sent.

在 3.5 版更改: The socket timeout is no more reset each time data is sent successfully.The socket timeout is now the maximum total duration to send all data.

在 3.5 版更改: If the system call is interrupted and the signal handler does not raisean exception, the method now retries the system call instead of raisingan InterruptedError exception (see PEP 475 for the rationale).

  • socket.sendto(bytes, address)
  • socket.sendto(bytes, flags, address)
  • Send data to the socket. The socket should not be connected to a remote socket,since the destination socket is specified by address. The optional flags_argument has the same meaning as for recv() above. Return the number ofbytes sent. (The format of _address depends on the address family —- seeabove.)

Raises an auditing event socket.sendto with arguments self, address.

在 3.5 版更改: If the system call is interrupted and the signal handler does not raisean exception, the method now retries the system call instead of raisingan InterruptedError exception (see PEP 475 for the rationale).

  • socket.sendmsg(buffers[, ancdata[, flags[, address]]])
  • Send normal and ancillary data to the socket, gathering thenon-ancillary data from a series of buffers and concatenating itinto a single message. The buffers argument specifies thenon-ancillary data as an iterable ofbytes-like objects(e.g. bytes objects); the operating system may set a limit(sysconf() value SCIOV_MAX) on the number of buffersthat can be used. The _ancdata argument specifies the ancillarydata (control messages) as an iterable of zero or more tuples(cmsglevel, cmsg_type, cmsg_data), where _cmsg_level andcmsg_type are integers specifying the protocol level andprotocol-specific type respectively, and cmsg_data is abytes-like object holding the associated data. Note thatsome systems (in particular, systems without CMSG_SPACE())might support sending only one control message per call. Theflags argument defaults to 0 and has the same meaning as forsend(). If address is supplied and not None, it sets adestination address for the message. The return value is thenumber of bytes of non-ancillary data sent.

The following function sends the list of file descriptors _fds_over an AF_UNIX socket, on systems which support theSCM_RIGHTS mechanism. See also recvmsg().

  1. import socket, array
  2.  
  3. def send_fds(sock, msg, fds):
  4. return sock.sendmsg([msg], [(socket.SOL_SOCKET, socket.SCM_RIGHTS, array.array("i", fds))])

Availability: most Unix platforms, possibly others.

Raises an auditing event socket.sendmsg with arguments self, address.

3.3 新版功能.

在 3.5 版更改: If the system call is interrupted and the signal handler does not raisean exception, the method now retries the system call instead of raisingan InterruptedError exception (see PEP 475 for the rationale).

  • socket.sendmsgafalg([_msg, ]*, op[, iv[, assoclen[, flags]]])
  • Specialized version of sendmsg() for AF_ALG socket.Set mode, IV, AEAD associated data length and flags for AF_ALG socket.

Availability: Linux >= 2.6.38.

3.6 新版功能.

  • socket.sendfile(file, offset=0, count=None)
  • Send a file until EOF is reached by using high-performanceos.sendfile and return the total number of bytes which were sent.file must be a regular file object opened in binary mode. Ifos.sendfile is not available (e.g. Windows) or file is not aregular file send() will be used instead. offset tells from where tostart reading the file. If specified, count is the total number of bytesto transmit as opposed to sending the file until EOF is reached. Fileposition is updated on return or also in case of error in which casefile.tell() can be used to figure out the number ofbytes which were sent. The socket must be of SOCK_STREAM type.Non-blocking sockets are not supported.

3.5 新版功能.

  • socket.setinheritable(_inheritable)
  • Set the inheritable flag of the socket's filedescriptor or socket's handle.

3.4 新版功能.

  • socket.setblocking(flag)
  • Set blocking or non-blocking mode of the socket: if flag is false, thesocket is set to non-blocking, else to blocking mode.

This method is a shorthand for certain settimeout() calls:

  • sock.setblocking(True) is equivalent to sock.settimeout(None)

  • sock.setblocking(False) is equivalent to sock.settimeout(0.0)

在 3.7 版更改: The method no longer applies SOCK_NONBLOCK flag onsocket.type.

  • socket.settimeout(value)
  • Set a timeout on blocking socket operations. The value argument can be anonnegative floating point number expressing seconds, or None.If a non-zero value is given, subsequent socket operations will raise atimeout exception if the timeout period value has elapsed beforethe operation has completed. If zero is given, the socket is put innon-blocking mode. If None is given, the socket is put in blocking mode.

For further information, please consult the notes on socket timeouts.

在 3.7 版更改: The method no longer toggles SOCK_NONBLOCK flag onsocket.type.

  • socket.setsockopt(level, optname, value: int)
  • socket.setsockopt(level, optname, value: buffer)
  • socket.setsockopt(level, optname, None, optlen: int)
  • Set the value of the given socket option (see the Unix manual pagesetsockopt(2))). The needed symbolic constants are defined in thesocket module (SO* etc.). The value can be an integer,None or a bytes-like object representing a buffer. In the latercase it is up to the caller to ensure that the bytestring contains theproper bits (see the optional built-in module struct for a way toencode C structures as bytestrings). When _value is set to None,optlen argument is required. It's equivalent to call setsockopt() Cfunction with optval=NULL and optlen=optlen.

在 3.5 版更改: 现在支持可写的 字节类对象

在 3.6 版更改: setsockopt(level, optname, None, optlen: int) form added.

  • socket.shutdown(how)
  • Shut down one or both halves of the connection. If how is SHUTRD,further receives are disallowed. If _how is SHUTWR, further sendsare disallowed. If _how is SHUT_RDWR, further sends and receives aredisallowed.

  • socket.share(process_id)

  • Duplicate a socket and prepare it for sharing with a target process. Thetarget process must be provided with process_id. The resulting bytes objectcan then be passed to the target process using some form of interprocesscommunication and the socket can be recreated there using fromshare().Once this method has been called, it is safe to close the socket sincethe operating system has already duplicated it for the target process.

可用性: Windows。

3.3 新版功能.

Note that there are no methods read() or write(); userecv() and send() without flags argument instead.

Socket objects also have these (read-only) attributes that correspond to thevalues given to the socket constructor.

  • socket.family
  • The socket family.

  • socket.type

  • The socket type.

  • socket.proto

  • The socket protocol.

Notes on socket timeouts

A socket object can be in one of three modes: blocking, non-blocking, ortimeout. Sockets are by default always created in blocking mode, but thiscan be changed by calling setdefaulttimeout().

  • In blocking mode, operations block until complete or the system returnsan error (such as connection timed out).

  • In non-blocking mode, operations fail (with an error that is unfortunatelysystem-dependent) if they cannot be completed immediately: functions from theselect can be used to know when and whether a socket is available forreading or writing.

  • In timeout mode, operations fail if they cannot be completed within thetimeout specified for the socket (they raise a timeout exception)or if the system returns an error.

注解

At the operating system level, sockets in timeout mode are internally setin non-blocking mode. Also, the blocking and timeout modes are shared betweenfile descriptors and socket objects that refer to the same network endpoint.This implementation detail can have visible consequences if e.g. you decideto use the fileno() of a socket.

Timeouts and the connect method

The connect() operation is also subject to the timeoutsetting, and in general it is recommended to call settimeout()before calling connect() or pass a timeout parameter tocreate_connection(). However, the system network stack may alsoreturn a connection timeout error of its own regardless of any Python sockettimeout setting.

Timeouts and the accept method

If getdefaulttimeout() is not None, sockets returned bythe accept() method inherit that timeout. Otherwise, thebehaviour depends on settings of the listening socket:

  • if the listening socket is in blocking mode or in timeout mode,the socket returned by accept() is in blocking mode;

  • if the listening socket is in non-blocking mode, whether the socketreturned by accept() is in blocking or non-blocking modeis operating system-dependent. If you want to ensure cross-platformbehaviour, it is recommended you manually override this setting.

示例

Here are four minimal example programs using the TCP/IP protocol: a server thatechoes all data that it receives back (servicing only one client), and a clientusing it. Note that a server must perform the sequence socket(),bind(), listen(), accept() (possiblyrepeating the accept() to service more than one client), while aclient only needs the sequence socket(), connect(). Alsonote that the server does not sendall()/recv() onthe socket it is listening on but on the new socket returned byaccept().

The first two examples support IPv4 only.

  1. # Echo server program
  2. import socket
  3.  
  4. HOST = '' # Symbolic name meaning all available interfaces
  5. PORT = 50007 # Arbitrary non-privileged port
  6. with socket.socket(socket.AF_INET, socket.SOCK_STREAM) as s:
  7. s.bind((HOST, PORT))
  8. s.listen(1)
  9. conn, addr = s.accept()
  10. with conn:
  11. print('Connected by', addr)
  12. while True:
  13. data = conn.recv(1024)
  14. if not data: break
  15. conn.sendall(data)
  1. # Echo client program
  2. import socket
  3.  
  4. HOST = 'daring.cwi.nl' # The remote host
  5. PORT = 50007 # The same port as used by the server
  6. with socket.socket(socket.AF_INET, socket.SOCK_STREAM) as s:
  7. s.connect((HOST, PORT))
  8. s.sendall(b'Hello, world')
  9. data = s.recv(1024)
  10. print('Received', repr(data))

The next two examples are identical to the above two, but support both IPv4 andIPv6. The server side will listen to the first address family available (itshould listen to both instead). On most of IPv6-ready systems, IPv6 will takeprecedence and the server may not accept IPv4 traffic. The client side will tryto connect to the all addresses returned as a result of the name resolution, andsends traffic to the first one connected successfully.

  1. # Echo server program
  2. import socket
  3. import sys
  4.  
  5. HOST = None # Symbolic name meaning all available interfaces
  6. PORT = 50007 # Arbitrary non-privileged port
  7. s = None
  8. for res in socket.getaddrinfo(HOST, PORT, socket.AF_UNSPEC,
  9. socket.SOCK_STREAM, 0, socket.AI_PASSIVE):
  10. af, socktype, proto, canonname, sa = res
  11. try:
  12. s = socket.socket(af, socktype, proto)
  13. except OSError as msg:
  14. s = None
  15. continue
  16. try:
  17. s.bind(sa)
  18. s.listen(1)
  19. except OSError as msg:
  20. s.close()
  21. s = None
  22. continue
  23. break
  24. if s is None:
  25. print('could not open socket')
  26. sys.exit(1)
  27. conn, addr = s.accept()
  28. with conn:
  29. print('Connected by', addr)
  30. while True:
  31. data = conn.recv(1024)
  32. if not data: break
  33. conn.send(data)
  1. # Echo client program
  2. import socket
  3. import sys
  4.  
  5. HOST = 'daring.cwi.nl' # The remote host
  6. PORT = 50007 # The same port as used by the server
  7. s = None
  8. for res in socket.getaddrinfo(HOST, PORT, socket.AF_UNSPEC, socket.SOCK_STREAM):
  9. af, socktype, proto, canonname, sa = res
  10. try:
  11. s = socket.socket(af, socktype, proto)
  12. except OSError as msg:
  13. s = None
  14. continue
  15. try:
  16. s.connect(sa)
  17. except OSError as msg:
  18. s.close()
  19. s = None
  20. continue
  21. break
  22. if s is None:
  23. print('could not open socket')
  24. sys.exit(1)
  25. with s:
  26. s.sendall(b'Hello, world')
  27. data = s.recv(1024)
  28. print('Received', repr(data))

The next example shows how to write a very simple network sniffer with rawsockets on Windows. The example requires administrator privileges to modifythe interface:

  1. import socket
  2.  
  3. # the public network interface
  4. HOST = socket.gethostbyname(socket.gethostname())
  5.  
  6. # create a raw socket and bind it to the public interface
  7. s = socket.socket(socket.AF_INET, socket.SOCK_RAW, socket.IPPROTO_IP)
  8. s.bind((HOST, 0))
  9.  
  10. # Include IP headers
  11. s.setsockopt(socket.IPPROTO_IP, socket.IP_HDRINCL, 1)
  12.  
  13. # receive all packages
  14. s.ioctl(socket.SIO_RCVALL, socket.RCVALL_ON)
  15.  
  16. # receive a package
  17. print(s.recvfrom(65565))
  18.  
  19. # disabled promiscuous mode
  20. s.ioctl(socket.SIO_RCVALL, socket.RCVALL_OFF)

The next example shows how to use the socket interface to communicate to a CANnetwork using the raw socket protocol. To use CAN with the broadcastmanager protocol instead, open a socket with:

  1. socket.socket(socket.AF_CAN, socket.SOCK_DGRAM, socket.CAN_BCM)

After binding (CAN_RAW) or connecting (CAN_BCM) the socket, youcan use the socket.send(), and the socket.recv() operations (andtheir counterparts) on the socket object as usual.

This last example might require special privileges:

  1. import socket
  2. import struct
  3.  
  4.  
  5. # CAN frame packing/unpacking (see 'struct can_frame' in <linux/can.h>)
  6.  
  7. can_frame_fmt = "=IB3x8s"
  8. can_frame_size = struct.calcsize(can_frame_fmt)
  9.  
  10. def build_can_frame(can_id, data):
  11. can_dlc = len(data)
  12. data = data.ljust(8, b'\x00')
  13. return struct.pack(can_frame_fmt, can_id, can_dlc, data)
  14.  
  15. def dissect_can_frame(frame):
  16. can_id, can_dlc, data = struct.unpack(can_frame_fmt, frame)
  17. return (can_id, can_dlc, data[:can_dlc])
  18.  
  19.  
  20. # create a raw socket and bind it to the 'vcan0' interface
  21. s = socket.socket(socket.AF_CAN, socket.SOCK_RAW, socket.CAN_RAW)
  22. s.bind(('vcan0',))
  23.  
  24. while True:
  25. cf, addr = s.recvfrom(can_frame_size)
  26.  
  27. print('Received: can_id=%x, can_dlc=%x, data=%s' % dissect_can_frame(cf))
  28.  
  29. try:
  30. s.send(cf)
  31. except OSError:
  32. print('Error sending CAN frame')
  33.  
  34. try:
  35. s.send(build_can_frame(0x01, b'\x01\x02\x03'))
  36. except OSError:
  37. print('Error sending CAN frame')

Running an example several times with too small delay between executions, couldlead to this error:

  1. OSError: [Errno 98] Address already in use

This is because the previous execution has left the socket in a TIME_WAITstate, and can't be immediately reused.

There is a socket flag to set, in order to prevent this,socket.SO_REUSEADDR:

  1. s = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
  2. s.setsockopt(socket.SOL_SOCKET, socket.SO_REUSEADDR, 1)
  3. s.bind((HOST, PORT))

the SO_REUSEADDR flag tells the kernel to reuse a local socket inTIME_WAIT state, without waiting for its natural timeout to expire.

参见

For an introduction to socket programming (in C), see the following papers:

  • An Introductory 4.3BSD Interprocess Communication Tutorial, by Stuart Sechrest

  • An Advanced 4.3BSD Interprocess Communication Tutorial, by Samuel J. Leffler etal,

both in the UNIX Programmer's Manual, Supplementary Documents 1 (sectionsPS1:7 and PS1:8). The platform-specific reference material for the varioussocket-related system calls are also a valuable source of information on thedetails of socket semantics. For Unix, refer to the manual pages; for Windows,see the WinSock (or Winsock 2) specification. For IPv6-ready APIs, readers maywant to refer to RFC 3493 titled Basic Socket Interface Extensions for IPv6.