<?xml version="1.0" encoding="utf-8" ?> <!DOCTYPE erlref SYSTEM "erlref.dtd"> <erlref> <header> <copyright> <year>1997</year><year>2016</year> <holder>Ericsson AB. All Rights Reserved.</holder> </copyright> <legalnotice> Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License. </legalnotice> <title>gen_tcp</title> <prepared>tony@erix.ericsson.se</prepared> <docno></docno> <date>1997-10-24</date> <rev>A</rev> </header> <module>gen_tcp</module> <modulesummary>Interface to TCP/IP sockets.</modulesummary> <description> <p>This module provides functions for communicating with sockets using the TCP/IP protocol.</p> <p>The following code fragment is a simple example of a client connecting to a server at port 5678, transferring a binary, and closing the connection:</p> <code type="none"> client() -> SomeHostInNet = "localhost", % to make it runnable on one machine {ok, Sock} = gen_tcp:connect(SomeHostInNet, 5678, [binary, {packet, 0}]), ok = gen_tcp:send(Sock, "Some Data"), ok = gen_tcp:close(Sock).</code> <p>At the other end, a server is listening on port 5678, accepts the connection, and receives the binary:</p> <code type="none"> server() -> {ok, LSock} = gen_tcp:listen(5678, [binary, {packet, 0}, {active, false}]), {ok, Sock} = gen_tcp:accept(LSock), {ok, Bin} = do_recv(Sock, []), ok = gen_tcp:close(Sock), Bin. do_recv(Sock, Bs) -> case gen_tcp:recv(Sock, 0) of {ok, B} -> do_recv(Sock, [Bs, B]); {error, closed} -> {ok, list_to_binary(Bs)} end.</code> <p>For more examples, see section <seealso marker="#examples">Examples</seealso>.</p> </description> <datatypes> <datatype> <name name="option"/> </datatype> <datatype> <name name="option_name"/> </datatype> <datatype> <name name="connect_option"/> </datatype> <datatype> <name name="listen_option"/> </datatype> <datatype> <name>socket()</name> <desc><p>As returned by <seealso marker="#accept/1"><c>accept/1,2</c></seealso> and <seealso marker="#connect/3"><c>connect/3,4</c></seealso>.</p> <marker id="connect"></marker> </desc> </datatype> </datatypes> <funcs> <func> <name name="accept" arity="1"/> <name name="accept" arity="2"/> <fsummary>Accept an incoming connection request on a listening socket.</fsummary> <type_desc variable="ListenSocket">Returned by <seealso marker="#listen/2"><c>listen/2</c></seealso>. </type_desc> <desc> <p>Accepts an incoming connection request on a listening socket. <c><anno>Socket</anno></c> must be a socket returned from <seealso marker="#listen/2"><c>listen/2</c></seealso>. <c><anno>Timeout</anno></c> specifies a time-out value in milliseconds. Defaults to <c>infinity</c>.</p> <p>Returns:</p> <list type="bulleted"> <item><p><c>{ok, <anno>Socket</anno>}</c> if a connection is established</p></item> <item><p><c>{error, closed}</c> if <c><anno>ListenSocket</anno></c> is closed</p></item> <item><p><c>{error, timeout}</c> if no connection is established within the specified time</p></item> <item><p><c>{error, system_limit}</c> if all available ports in the Erlang emulator are in use</p></item> <item><p>A POSIX error value if something else goes wrong, see <seealso marker="inet"><c>inet(3)</c></seealso> for possible error values</p></item> </list> <p>Packets can be sent to the returned socket <c><anno>Socket</anno></c> using <seealso marker="#send/2"><c>send/2</c></seealso>. Packets sent from the peer are delivered as messages (unless <c>{active, false}</c> is specified in the option list for the listening socket, in which case packets are retrieved by calling <seealso marker="#recv/2"><c>recv/2</c></seealso>):</p> <code type="none"> {tcp, Socket, Data}</code> <note> <p>The <c>accept</c> call does <em>not</em> have to be issued from the socket owner process. Using version 5.5.3 and higher of the emulator, multiple simultaneous accept calls can be issued from different processes, which allows for a pool of acceptor processes handling incoming connections.</p> </note> </desc> </func> <func> <name name="close" arity="1"/> <fsummary>Close a TCP socket.</fsummary> <desc> <p>Closes a TCP socket.</p> </desc> </func> <func> <name name="connect" arity="3"/> <name name="connect" arity="4"/> <fsummary>Connect to a TCP port.</fsummary> <desc> <p>Connects to a server on TCP port <c><anno>Port</anno></c> on the host with IP address <c><anno>Address</anno></c>. Argument <c><anno>Address</anno></c> can be a hostname or an IP address.</p> <p>The following options are available:</p> <taglist> <tag><c>{ip, Address}</c></tag> <item><p>If the host has many network interfaces, this option specifies which one to use.</p></item> <tag><c>{ifaddr, Address}</c></tag> <item><p>Same as <c>{ip, Address}</c>. If the host has many network interfaces, this option specifies which one to use.</p> </item> <tag><c>{fd, integer() >= 0}</c></tag> <item><p>If a socket has somehow been connected without using <c>gen_tcp</c>, use this option to pass the file descriptor for it. If <c>{ip, Address}</c> and/or <c>{port, port_number()}</c> is combined with this option, the <c>fd</c> is bound to the specified interface and port before connecting. If these options are not specified, it is assumed that the <c>fd</c> is already bound appropriately.</p></item> <tag><c>inet</c></tag> <item><p>Sets up the socket for IPv4.</p></item> <tag><c>inet6</c></tag> <item><p>Sets up the socket for IPv6.</p></item> <tag><c>local</c></tag> <item> <p> Sets up a Unix Domain Socket. See <seealso marker="inet#type-local_address"> <c>inet:local_address()</c> </seealso> </p> </item> <tag><c>{port, Port}</c></tag> <item><p>Specifies which local port number to use.</p></item> <tag><c>{tcp_module, module()}</c></tag> <item><p>Overrides which callback module is used. Defaults to <c>inet_tcp</c> for IPv4 and <c>inet6_tcp</c> for IPv6.</p></item> <tag><c>Opt</c></tag> <item><p>See <seealso marker="inet#setopts/2"><c>inet:setopts/2</c></seealso>.</p> </item> </taglist> <p>Packets can be sent to the returned socket <c><anno>Socket</anno></c> using <seealso marker="#send/2"><c>send/2</c></seealso>. Packets sent from the peer are delivered as messages:</p> <code type="none"> {tcp, Socket, Data}</code> <p>If the socket is in <c>{active, N}</c> mode (see <seealso marker="inet#setopts/2"><c>inet:setopts/2</c></seealso> for details) and its message counter drops to <c>0</c>, the following message is delivered to indicate that the socket has transitioned to passive (<c>{active, false}</c>) mode:</p> <code type="none"> {tcp_passive, Socket}</code> <p>If the socket is closed, the following message is delivered:</p> <code type="none"> {tcp_closed, Socket}</code> <p>If an error occurs on the socket, the following message is delivered (unless <c>{active, false}</c> is specified in the option list for the socket, in which case packets are retrieved by calling <seealso marker="#recv/2"><c>recv/2</c></seealso>):</p> <code type="none"> {tcp_error, Socket, Reason}</code> <p>The optional <c><anno>Timeout</anno></c> parameter specifies a time-out in milliseconds. Defaults to <c>infinity</c>.</p> <note> <p>The default values for options specified to <c>connect</c> can be affected by the <c>Kernel</c> configuration parameter <c>inet_default_connect_options</c>. For details, see <seealso marker="inet"><c>inet(3)</c></seealso>.</p> </note> </desc> </func> <func> <name name="controlling_process" arity="2"/> <fsummary>Change controlling process of a socket.</fsummary> <desc> <p>Assigns a new controlling process <c><anno>Pid</anno></c> to <c><anno>Socket</anno></c>. The controlling process is the process that receives messages from the socket. If called by any other process than the current controlling process, <c>{error, not_owner}</c> is returned.</p> <p>If the socket is set in active mode, this function will transfer any messages in the mailbox of the caller to the new controlling process. If any other process is interacting with the socket while the transfer is happening, the transfer may not work correctly and messages may remain in the caller's mailbox. For instance changing the sockets active mode before the transfere is complete may cause this.</p> </desc> </func> <func> <name name="listen" arity="2"/> <fsummary>Set up a socket to listen on a port.</fsummary> <desc> <p>Sets up a socket to listen on port <c><anno>Port</anno></c> on the local host.</p> <p>If <c><anno>Port</anno> == 0</c>, the underlying OS assigns an available port number, use <seealso marker="inet#port/1"><c>inet:port/1</c></seealso> to retrieve it.</p> <p>The following options are available:</p> <taglist> <tag><c>list</c></tag> <item><p>Received <c>Packet</c> is delivered as a list.</p></item> <tag><c>binary</c></tag> <item><p>Received <c>Packet</c> is delivered as a binary.</p></item> <tag><c>{backlog, B}</c></tag> <item><p><c>B</c> is an integer >= <c>0</c>. The backlog value defines the maximum length that the queue of pending connections can grow to. Defaults to <c>5</c>.</p></item> <tag><c>{ip, Address}</c></tag> <item><p>If the host has many network interfaces, this option specifies which one to listen on.</p></item> <tag><c>{port, Port}</c></tag> <item><p>Specifies which local port number to use.</p></item> <tag><c>{fd, Fd}</c></tag> <item><p>If a socket has somehow been connected without using <c>gen_tcp</c>, use this option to pass the file descriptor for it.</p></item> <tag><c>{ifaddr, Address}</c></tag> <item><p>Same as <c>{ip, Address}</c>. If the host has many network interfaces, this option specifies which one to use.</p> </item> <tag><c>inet6</c></tag> <item><p>Sets up the socket for IPv6.</p></item> <tag><c>inet</c></tag> <item><p>Sets up the socket for IPv4.</p></item> <tag><c>{tcp_module, module()}</c></tag> <item><p>Overrides which callback module is used. Defaults to <c>inet_tcp</c> for IPv4 and <c>inet6_tcp</c> for IPv6.</p></item> <tag><c>Opt</c></tag> <item><p>See <seealso marker="inet#setopts/2"><c>inet:setopts/2</c></seealso>. </p></item> </taglist> <p>The returned socket <c><anno>ListenSocket</anno></c> can only be used in calls to <seealso marker="#accept/1"><c>accept/1,2</c></seealso>.</p> <note> <p>The default values for options specified to <c>listen</c> can be affected by the <c>Kernel</c> configuration parameter <c>inet_default_listen_options</c>. For details, see <seealso marker="inet"><c>inet(3)</c></seealso>.</p> </note> </desc> </func> <func> <name name="recv" arity="2"/> <name name="recv" arity="3"/> <fsummary>Receive a packet from a passive socket.</fsummary> <type_desc variable="HttpPacket">See the description of <c>HttpPacket</c> in <seealso marker="erts:erlang#decode_packet/3"><c>erlang:decode_packet/3</c></seealso> in <c>ERTS</c>. </type_desc> <desc> <p>Receives a packet from a socket in passive mode. A closed socket is indicated by return value <c>{error, closed}</c>.</p> <p>Argument <c><anno>Length</anno></c> is only meaningful when the socket is in <c>raw</c> mode and denotes the number of bytes to read. If <c><anno>Length</anno></c> is <c>0</c>, all available bytes are returned. If <c><anno>Length</anno></c> > <c>0</c>, exactly <c><anno>Length</anno></c> bytes are returned, or an error; possibly discarding less than <c><anno>Length</anno></c> bytes of data when the socket is closed from the other side.</p> <p>The optional <c><anno>Timeout</anno></c> parameter specifies a time-out in milliseconds. Defaults to <c>infinity</c>.</p> </desc> </func> <func> <name name="send" arity="2"/> <fsummary>Send a packet.</fsummary> <desc> <p>Sends a packet on a socket.</p> <p>There is no <c>send</c> call with a time-out option, use socket option <c>send_timeout</c> if time-outs are desired. See section <seealso marker="#examples">Examples</seealso>.</p> </desc> </func> <func> <name name="shutdown" arity="2"/> <fsummary>Asynchronously close a socket.</fsummary> <desc> <p>Closes a socket in one or two directions.</p> <p><c><anno>How</anno> == write</c> means closing the socket for writing, reading from it is still possible.</p> <p>If <c><anno>How</anno> == read</c> or there is no outgoing data buffered in the <c><anno>Socket</anno></c> port, the socket is shut down immediately and any error encountered is returned in <c><anno>Reason</anno></c>.</p> <p>If there is data buffered in the socket port, the attempt to shutdown the socket is postponed until that data is written to the kernel socket send buffer. If any errors are encountered, the socket is closed and <c>{error, closed}</c> is returned on the next <seealso marker="#recv/2"><c>recv/2</c></seealso> or <seealso marker="#send/2"><c>send/2</c></seealso>.</p> <p>Option <c>{exit_on_close, false}</c> is useful if the peer has done a shutdown on the write side.</p> </desc> </func> </funcs> <section> <title>Examples</title> <marker id="examples"></marker> <p>The following example illustrates use of option <c>{active,once}</c> and multiple accepts by implementing a server as a number of worker processes doing accept on a single listening socket. Function <c>start/2</c> takes the number of worker processes and the port number on which to listen for incoming connections. If <c>LPort</c> is specified as <c>0</c>, an ephemeral port number is used, which is why the start function returns the actual port number allocated:</p> <code type="none"> start(Num,LPort) -> case gen_tcp:listen(LPort,[{active, false},{packet,2}]) of {ok, ListenSock} -> start_servers(Num,ListenSock), {ok, Port} = inet:port(ListenSock), Port; {error,Reason} -> {error,Reason} end. start_servers(0,_) -> ok; start_servers(Num,LS) -> spawn(?MODULE,server,[LS]), start_servers(Num-1,LS). server(LS) -> case gen_tcp:accept(LS) of {ok,S} -> loop(S), server(LS); Other -> io:format("accept returned ~w - goodbye!~n",[Other]), ok end. loop(S) -> inet:setopts(S,[{active,once}]), receive {tcp,S,Data} -> Answer = process(Data), % Not implemented in this example gen_tcp:send(S,Answer), loop(S); {tcp_closed,S} -> io:format("Socket ~w closed [~w]~n",[S,self()]), ok end.</code> <p>Example of a simple client:</p> <code type="none"> client(PortNo,Message) -> {ok,Sock} = gen_tcp:connect("localhost",PortNo,[{active,false}, {packet,2}]), gen_tcp:send(Sock,Message), A = gen_tcp:recv(Sock,0), gen_tcp:close(Sock), A.</code> <p>The <c>send</c> call does not accept a time-out option because time-outs on send is handled through socket option <c>send_timeout</c>. The behavior of a send operation with no receiver is mainly defined by the underlying TCP stack and the network infrastructure. To write code that handles a hanging receiver that can eventually cause the sender to hang on a <c>send</c> do like the following.</p> <p>Consider a process that receives data from a client process to be forwarded to a server on the network. The process is connected to the server through TCP/IP and does not get any acknowledge for each message it sends, but has to rely on the send time-out option to detect that the other end is unresponsive. Option <c>send_timeout</c> can be used when connecting:</p> <code type="none"> ... {ok,Sock} = gen_tcp:connect(HostAddress, Port, [{active,false}, {send_timeout, 5000}, {packet,2}]), loop(Sock), % See below ...</code> <p>In the loop where requests are handled, send time-outs can now be detected:</p> <code type="none"> loop(Sock) -> receive {Client, send_data, Binary} -> case gen_tcp:send(Sock,[Binary]) of {error, timeout} -> io:format("Send timeout, closing!~n", []), handle_send_timeout(), % Not implemented here Client ! {self(),{error_sending, timeout}}, %% Usually, it's a good idea to give up in case of a %% send timeout, as you never know how much actually %% reached the server, maybe only a packet header?! gen_tcp:close(Sock); {error, OtherSendError} -> io:format("Some other error on socket (~p), closing", [OtherSendError]), Client ! {self(),{error_sending, OtherSendError}}, gen_tcp:close(Sock); ok -> Client ! {self(), data_sent}, loop(Sock) end end.</code> <p>Usually it suffices to detect time-outs on receive, as most protocols include some sort of acknowledgment from the server, but if the protocol is strictly one way, option <c>send_timeout</c> comes in handy.</p> </section> </erlref>