:mod:`usocket` -- socket module
.. module:: usocket :synopsis: socket module
|see_cpython_module| :mod:`python:socket`.
This module provides access to the BSD socket interface.
Difference to CPython
For efficiency and consistency, socket objects in Pycopy implement a stream (file-like) interface directly. In CPython, you need to convert a socket to a file-like object using makefile() method. This method is still supported by Pycopy (but is a no-op), so where compatibility with CPython matters, be sure to use it.
The native socket address format of the usocket module is an opaque data type
returned by getaddrinfo function, which must be used to resolve textual address
(including numeric addresses):
sockaddr = usocket.getaddrinfo('google.com', 80)[0][-1]
# You must use getaddrinfo() even for numeric addresses
sockaddr = usocket.getaddrinfo('127.0.0.1', 80)[0][-1]
# Now you can use that address
sock.connect(addr)
Using getaddrinfo is the most efficient (both in terms of memory and processing power) and portable way to work with addresses.
However, socket module (note the difference with native Pycopy
usocket module described here) provides CPython-compatible way to specify
addresses using tuples, as described below. Note that depending on a
Pycopy port, socket module can be builtin or need to be
installed from pycopy-lib (as in the case of Pycopy Unix port),
and some ports still accept only numeric addresses in the tuple format,
and require to use getaddrinfo function to resolve domain names.
Summing up:
- Always use getaddrinfo when writing portable applications.
- Tuple addresses described below can be used as a shortcut for quick hacks and interactive use, if your port supports them.
Tuple address format for socket module:
- IPv4: (ipv4_address, port), where ipv4_address is a string with
dot-notation numeric IPv4 address, e.g.
"8.8.8.8", and port is and integer port number in the range 1-65535. Note the domain names are not accepted as ipv4_address, they should be resolved first using usocket.getaddrinfo(). - IPv6: (ipv6_address, port, flowinfo, scopeid), where ipv6_address
is a string with colon-notation numeric IPv6 address, e.g.
"2001:db8::1", and port is an integer port number in the range 1-65535. flowinfo must be 0. scopeid is the interface scope identifier for link-local addresses. Note the domain names are not accepted as ipv6_address, they should be resolved first using usocket.getaddrinfo(). Availability of IPv6 support depends on a Pycopy port.
.. function:: socket(af=AF_INET, type=SOCK_STREAM, proto=IPPROTO_TCP, /)
Create a new socket using the given address family, socket type and
protocol number. Note that specifying *proto* in most cases is not
required (and not recommended, as some Pycopy ports may omit
``IPPROTO_*`` constants). Instead, *type* argument will select needed
protocol automatically::
# Create STREAM TCP socket
socket(AF_INET, SOCK_STREAM)
# Create DGRAM UDP socket
socket(AF_INET, SOCK_DGRAM)
.. function:: getaddrinfo(host, port, af=0, type=0, proto=0, flags=0, /)
Translate the host/port argument into a sequence of 5-tuples that contain all the
necessary arguments for creating a socket connected to that service. Arguments
*af*, *type*, and *proto* (which have the same meaning as for the `socket()` function)
can be used to filter which kind of addresses are returned. If a parameter is not
specified or zero, all combinations of addresses can be returned (requiring
filtering on the user side).
The resulting list of 5-tuples has the following structure::
(family, type, proto, canonname, sockaddr)
The following example shows how to connect to a given url::
s = usocket.socket()
# This assumes that if "type" is not specified, an address for
# SOCK_STREAM will be returned, which may be not true
s.connect(usocket.getaddrinfo('google.com', 80)[0][-1])
Recommended use of filtering params::
s = usocket.socket()
# Guaranteed to return an address which can be connect'ed to for
# stream operation.
s.connect(usocket.getaddrinfo('google.com', 80, 0, SOCK_STREAM)[0][-1])
.. admonition:: Difference to CPython
:class: attention
CPython raises a ``socket.gaierror`` exception (`OSError` subclass) in case
of error in this function. Pycopy doesn't have ``socket.gaierror``
and raises OSError directly. Note that error numbers of `getaddrinfo()`
form a separate namespace and may not match error numbers from
the :mod:`uerrno` module. To distinguish `getaddrinfo()` errors, they are
represented by negative numbers, whereas standard system errors are
positive numbers (error numbers are accessible using ``e.args[0]`` property
from an exception object). The use of negative values is a provisional
detail which may change in the future.
.. function:: inet_ntop(af, bin_addr)
Convert a binary network address *bin_addr* of the given address family *af*
to a textual representation::
>>> usocket.inet_ntop(usocket.AF_INET, b"\x7f\0\0\1")
'127.0.0.1'
.. function:: inet_pton(af, txt_addr)
Convert a textual network address *txt_addr* of the given address family *af*
to a binary representation::
>>> usocket.inet_pton(usocket.AF_INET, "1.2.3.4")
b'\x01\x02\x03\x04'
.. data:: AF_INET
AF_INET6
Address family types. Availability depends on a particular `Pycopy port`.
.. data:: SOCK_STREAM
SOCK_DGRAM
Socket types.
.. data:: IPPROTO_UDP
IPPROTO_TCP
IP protocol numbers. Availability depends on a particular `Pycopy port`.
Note that you don't need to specify these in a call to `usocket.socket()`,
because `SOCK_STREAM` socket type automatically selects `IPPROTO_TCP`, and
`SOCK_DGRAM` - `IPPROTO_UDP`. Thus, the only real use of these constants
is as an argument to `setsockopt()`.
.. data:: usocket.SOL_* Socket option levels (an argument to `setsockopt()`). The exact inventory depends on a `Pycopy port`.
.. data:: usocket.SO_* Socket options (an argument to `setsockopt()`). The exact inventory depends on a `Pycopy port`.
Constants specific to WiPy:
.. data:: IPPROTO_SEC
Special protocol value to create SSL-compatible socket.
.. method:: socket.close() Mark the socket closed and release all resources. Once that happens, all future operations on the socket object will fail. The remote end will receive EOF indication if supported by protocol. Sockets are automatically closed when they are garbage-collected, but it is recommended to `close()` them explicitly as soon you finished working with them.
.. method:: socket.bind(address) Bind the socket to *address*. The socket must not already be bound.
.. method:: socket.listen([backlog]) Enable a server to accept connections. If *backlog* is specified, it must be at least 0 (if it's lower, it will be set to 0); and specifies the number of unaccepted connections that the system will allow before refusing new connections. If not specified, a default reasonable value is chosen.
.. method:: socket.accept() Accept a connection. The socket must be bound to an address and listening for connections. The return value is a pair (conn, address) where conn is a new socket object usable to send and receive data on the connection, and address is the address bound to the socket on the other end of the connection.
.. method:: socket.connect(address) Connect to a remote socket at *address*.
.. method:: socket.send(bytes)
Send data to the socket. The socket must be connected to a remote socket.
Returns number of bytes sent, which may be smaller than the length of data
("short write").
.. method:: socket.sendall(bytes) Send all data to the socket. The socket must be connected to a remote socket. Unlike `send()`, this method will try to send all of data, by sending data chunk by chunk consecutively. The behavior of this method on non-blocking sockets is undefined. Due to this, on Pycopy, it's recommended to use `write()` method instead, which has the same "no short writes" policy for blocking sockets, and will return number of bytes sent on non-blocking sockets.
.. method:: socket.recv(bufsize) Receive data from the socket. The return value is a bytes object representing the data received. The maximum amount of data to be received at once is specified by bufsize.
.. method:: socket.sendto(bytes, address) Send data to the socket. The socket should not be connected to a remote socket, since the destination socket is specified by *address*.
.. method:: socket.recvfrom(bufsize) 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 the address of the socket sending the data.
.. method:: socket.setsockopt(level, optname, value) Set the value of the given socket option. The needed symbolic constants are defined in the socket module (SO_* etc.). The *value* can be an integer or a bytes-like object representing a buffer.
.. method:: socket.settimeout(value)
**Note**: Not every port supports this method, see below.
Set a timeout on blocking socket operations. The value argument can be a nonnegative floating
point number expressing seconds, or None. If a non-zero value is given, subsequent socket operations
will raise an `OSError` exception if the timeout period value has elapsed before the operation has
completed. If zero is given, the socket is put in non-blocking mode. If None is given, the socket
is put in blocking mode.
Not every `Pycopy port` supports this method. A more portable and
generic solution is to use `uselect.poll` object. This allows to wait on
multiple objects at the same time (and not just on sockets, but on generic
`stream` objects which support polling). Example::
# Instead of:
s.settimeout(1.0) # time in seconds
s.read(10) # may timeout
# Use:
poller = uselect.poll()
poller.register(s, uselect.POLLIN)
res = poller.poll(1000) # time in milliseconds
if not res:
# s is still not ready for input, i.e. operation timed out
.. admonition:: Difference to CPython
:class: attention
CPython raises a ``socket.timeout`` exception in case of timeout,
which is an `OSError` subclass. Pycopy raises an OSError directly
instead. If you use ``except OSError:`` to catch the exception,
your code will work both in Pycopy and CPython.
.. method:: socket.setblocking(flag) Set blocking or non-blocking mode of the socket: if flag is false, the socket 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)``
.. method:: socket.makefile(mode='rb', buffering=0, /)
Return a file object associated with the socket. The exact returned type depends on the arguments
given to makefile(). The support is limited to binary modes only ('rb', 'wb', and 'rwb').
CPython's arguments: *encoding*, *errors* and *newline* are not supported.
.. admonition:: Difference to CPython
:class: attention
As Pycopy doesn't support buffered streams, values of *buffering*
parameter is ignored and treated as if it was 0 (unbuffered).
.. admonition:: Difference to CPython
:class: attention
Closing the file object returned by makefile() WILL close the
original socket as well.
.. method:: socket.read([size]) Read up to size bytes from the socket. Return a bytes object. If *size* is not given, it reads all data available from the socket until EOF; as such the method will not return until the socket is closed. This function tries to read as much data as requested (no "short reads"). This may be not possible with non-blocking socket though, and then less data will be returned.
.. method:: socket.readinto(buf[, nbytes]) Read bytes into the *buf*. If *nbytes* is specified then read at most that many bytes. Otherwise, read at most *len(buf)* bytes. Just as `read()`, this method follows "no short reads" policy. Return value: number of bytes read and stored into *buf*.
.. method:: socket.readline() Read a line, ending in a newline character. Return value: the line read.
.. method:: socket.write(buf) Write the buffer of bytes to the socket. This function will try to write all data to a socket (no "short writes"). This may be not possible with a non-blocking socket though, and returned value will be less than the length of *buf*. Return value: number of bytes written.
.. exception:: usocket.error
Pycopy does NOT have this exception.
.. admonition:: Difference to CPython
:class: attention
CPython used to have a ``socket.error`` exception which is now deprecated,
and is an alias of `OSError`. In Pycopy, use `OSError` directly.