home / documentation / v18 / http

HTTP

目录

Added in: v0.10.0

稳定性: 2 - Stable

To use the HTTP server and client one must require('node:http').

The HTTP interfaces in Node.js are designed to support many features of the protocol which have been traditionally difficult to use. In particular, large, possibly chunk-encoded, messages. The interface is careful to never buffer entire requests or responses, so the user is able to stream data.

HTTP message headers are represented by an object like this:

JS

Keys are lowercased. Values are not modified.

In order to support the full spectrum of possible HTTP applications, the Node.js HTTP API is very low-level. It deals with stream handling and message parsing only. It parses a message into headers and body but it does not parse the actual headers or the body.

See message.headers for details on how duplicate headers are handled.

The raw headers as they were received are retained in the rawHeaders property, which is an array of [key, value, key2, value2, ...]. For example, the previous message header object might have a rawHeaders list like the following:

JS

C http.Agent

Added in: v0.3.4

An Agent is responsible for managing connection persistence and reuse for HTTP clients. It maintains a queue of pending requests for a given host and port, reusing a single socket connection for each until the queue is empty, at which time the socket is either destroyed or put into a pool where it is kept to be used again for requests to the same host and port. Whether it is destroyed or pooled depends on the keepAlive option.

Pooled connections have TCP Keep-Alive enabled for them, but servers may still close idle connections, in which case they will be removed from the pool and a new connection will be made when a new HTTP request is made for that host and port. Servers may also refuse to allow multiple requests over the same connection, in which case the connection will have to be remade for every request and cannot be pooled. The Agent will still make the requests to that server, but each one will occur over a new connection.

When a connection is closed by the client or the server, it is removed from the pool. Any unused sockets in the pool will be unrefed so as not to keep the Node.js process running when there are no outstanding requests. (see socket.unref()).

It is good practice, to destroy() an Agent instance when it is no longer in use, because unused sockets consume OS resources.

Sockets are removed from an agent when the socket emits either a 'close' event or an 'agentRemove' event. When intending to keep one HTTP request open for a long time without keeping it in the agent, something like the following may be done:

JS

An agent may also be used for an individual request. By providing {agent: false} as an option to the http.get() or http.request() functions, a one-time use Agent with default options will be used for the client connection.

agent:false:

JS

M new Agent([options])

历史
版本更改
v15.6.0, v14.17.0Change the default scheduling from 'fifo' to 'lifo'.
v14.5.0, v12.19.0Add `maxTotalSockets` option to agent constructor.
v14.5.0, v12.20.0Add `scheduling` option to specify the free socket scheduling strategy.
v0.3.4Added in: v0.3.4
  • options Object Set of configurable options to set on the agent. Can have the following fields:
    • keepAlive boolean Keep sockets around even when there are no outstanding requests, so they can be used for future requests without having to reestablish a TCP connection. Not to be confused with the keep-alive value of the Connection header. The Connection: keep-alive header is always sent when using an agent except when the Connection header is explicitly specified or when the keepAlive and maxSockets options are respectively set to false and Infinity, in which case Connection: close will be used. Default: false.
    • keepAliveMsecs number When using the keepAlive option, specifies the initial delay for TCP Keep-Alive packets. Ignored when the keepAlive option is false or undefined. Default: 1000.
    • maxSockets number Maximum number of sockets to allow per host. If the same host opens multiple concurrent connections, each request will use new socket until the maxSockets value is reached. If the host attempts to open more connections than maxSockets, the additional requests will enter into a pending request queue, and will enter active connection state when an existing connection terminates. This makes sure there are at most maxSockets active connections at any point in time, from a given host. Default: Infinity.
    • maxTotalSockets number Maximum number of sockets allowed for all hosts in total. Each request will use a new socket until the maximum is reached. Default: Infinity.
    • maxFreeSockets number Maximum number of sockets per host to leave open in a free state. Only relevant if keepAlive is set to true. Default: 256.
    • scheduling string Scheduling strategy to apply when picking the next free socket to use. It can be 'fifo' or 'lifo'. The main difference between the two scheduling strategies is that 'lifo' selects the most recently used socket, while 'fifo' selects the least recently used socket. In case of a low rate of request per second, the 'lifo' scheduling will lower the risk of picking a socket that might have been closed by the server due to inactivity. In case of a high rate of request per second, the 'fifo' scheduling will maximize the number of open sockets, while the 'lifo' scheduling will keep it as low as possible. Default: 'lifo'.
    • timeout number Socket timeout in milliseconds. This will set the timeout when the socket is created.

options in socket.connect() are also supported.

The default http.globalAgent that is used by http.request() has all of these values set to their respective defaults.

To configure any of them, a custom http.Agent instance must be created.

JS

M agent.createConnection(options[, callback])

Added in: v0.11.4

Produces a socket/stream to be used for HTTP requests.

By default, this function is the same as net.createConnection(). However, custom agents may override this method in case greater flexibility is desired.

A socket/stream can be supplied in one of two ways: by returning the socket/stream from this function, or by passing the socket/stream to callback.

This method is guaranteed to return an instance of the net.Socket class, a subclass of stream.Duplex, unless the user specifies a socket type other than net.Socket.

callback has a signature of (err, stream).

M agent.keepSocketAlive(socket)

Added in: v8.1.0

Called when socket is detached from a request and could be persisted by the Agent. Default behavior is to:

JS

This method can be overridden by a particular Agent subclass. If this method returns a falsy value, the socket will be destroyed instead of persisting it for use with the next request.

The socket argument can be an instance of net.Socket, a subclass of stream.Duplex.

M agent.reuseSocket(socket, request)

Added in: v8.1.0

Called when socket is attached to request after being persisted because of the keep-alive options. Default behavior is to:

JS

This method can be overridden by a particular Agent subclass.

The socket argument can be an instance of net.Socket, a subclass of stream.Duplex.

M agent.destroy()

Added in: v0.11.4

Destroy any sockets that are currently in use by the agent.

It is usually not necessary to do this. However, if using an agent with keepAlive enabled, then it is best to explicitly shut down the agent when it is no longer needed. Otherwise, sockets might stay open for quite a long time before the server terminates them.

M agent.freeSockets

历史
版本更改
v16.0.0The property now has a `null` prototype.
v0.11.4Added in: v0.11.4

An object which contains arrays of sockets currently awaiting use by the agent when keepAlive is enabled. Do not modify.

Sockets in the freeSockets list will be automatically destroyed and removed from the array on 'timeout'.

M agent.getName([options])

历史
版本更改
v17.7.0The `options` parameter is now optional.
v0.11.4Added in: v0.11.4
  • options Object A set of options providing information for name generation
    • host string A domain name or IP address of the server to issue the request to
    • port number Port of remote server
    • localAddress string Local interface to bind for network connections when issuing the request
    • family integer Must be 4 or 6 if this doesn't equal undefined.
  • Returns: string

Get a unique name for a set of request options, to determine whether a connection can be reused. For an HTTP agent, this returns host:port:localAddress or host:port:localAddress:family. For an HTTPS agent, the name includes the CA, cert, ciphers, and other HTTPS/TLS-specific options that determine socket reusability.

M agent.maxFreeSockets

Added in: v0.11.7

By default set to 256. For agents with keepAlive enabled, this sets the maximum number of sockets that will be left open in the free state.

M agent.maxSockets

Added in: v0.3.6

By default set to Infinity. Determines how many concurrent sockets the agent can have open per origin. Origin is the returned value of agent.getName().

M agent.maxTotalSockets

Added in: v14.5.0, v12.19.0

By default set to Infinity. Determines how many concurrent sockets the agent can have open. Unlike maxSockets, this parameter applies across all origins.

M agent.requests

历史
版本更改
v16.0.0The property now has a `null` prototype.
v0.5.9Added in: v0.5.9

An object which contains queues of requests that have not yet been assigned to sockets. Do not modify.

M agent.sockets

历史
版本更改
v16.0.0The property now has a `null` prototype.
v0.3.6Added in: v0.3.6

An object which contains arrays of sockets currently in use by the agent. Do not modify.

C http.ClientRequest

Added in: v0.1.17

This object is created internally and returned from http.request(). It represents an in-progress request whose header has already been queued. The header is still mutable using the setHeader(name, value), getHeader(name), removeHeader(name) API. The actual header will be sent along with the first data chunk or when calling request.end().

To get the response, add a listener for 'response' to the request object. 'response' will be emitted from the request object when the response headers have been received. The 'response' event is executed with one argument which is an instance of http.IncomingMessage.

During the 'response' event, one can add listeners to the response object; particularly to listen for the 'data' event.

If no 'response' handler is added, then the response will be entirely discarded. However, if a 'response' event handler is added, then the data from the response object must be consumed, either by calling response.read() whenever there is a 'readable' event, or by adding a 'data' handler, or by calling the .resume() method. Until the data is consumed, the 'end' event will not fire. Also, until the data is read it will consume memory that can eventually lead to a 'process out of memory' error.

For backward compatibility, res will only emit 'error' if there is an 'error' listener registered.

Set Content-Length header to limit the response body size. Mismatching the Content-Length header value will result in an [Error][] being thrown, identified by code: 'ERR_HTTP_CONTENT_LENGTH_MISMATCH'.

Content-Length value should be in bytes, not characters. Use Buffer.byteLength() to determine the length of the body in bytes.

E 'abort'

Deprecated in: v17.0.0, v16.12.0

稳定性: 0 - Deprecated. Listen for the `'close'` event instead.

Emitted when the request has been aborted by the client. This event is only emitted on the first call to abort().

E 'close'

Added in: v0.5.4

Indicates that the request is completed, or its underlying connection was terminated prematurely (before the response completion).

E 'connect'

Added in: v0.7.0

Emitted each time a server responds to a request with a CONNECT method. If this event is not being listened for, clients receiving a CONNECT method will have their connections closed.

This event is guaranteed to be passed an instance of the net.Socket class, a subclass of stream.Duplex, unless the user specifies a socket type other than net.Socket.

A client and server pair demonstrating how to listen for the 'connect' event:

JS

E 'continue'

Added in: v0.3.2

Emitted when the server sends a '100 Continue' HTTP response, usually because the request contained 'Expect: 100-continue'. This is an instruction that the client should send the request body.

E 'finish'

Added in: v0.3.6

Emitted when the request has been sent. More specifically, this event is emitted when the last segment of the response headers and body have been handed off to the operating system for transmission over the network. It does not imply that the server has received anything yet.

E 'information'

Added in: v10.0.0

Emitted when the server sends a 1xx intermediate response (excluding 101 Upgrade). The listeners of this event will receive an object containing the HTTP version, status code, status message, key-value headers object, and array with the raw header names followed by their respective values.

JS

101 Upgrade statuses do not fire this event due to their break from the traditional HTTP request/response chain, such as web sockets, in-place TLS upgrades, or HTTP 2.0. To be notified of 101 Upgrade notices, listen for the 'upgrade' event instead.

E 'response'

Added in: v0.1.0

Emitted when a response is received to this request. This event is emitted only once.

E 'socket'

Added in: v0.5.3

This event is guaranteed to be passed an instance of the net.Socket class, a subclass of stream.Duplex, unless the user specifies a socket type other than net.Socket.

E 'timeout'

Added in: v0.7.8

Emitted when the underlying socket times out from inactivity. This only notifies that the socket has been idle. The request must be destroyed manually.

See also: request.setTimeout().

E 'upgrade'

Added in: v0.1.94

Emitted each time a server responds to a request with an upgrade. If this event is not being listened for and the response status code is 101 Switching Protocols, clients receiving an upgrade header will have their connections closed.

This event is guaranteed to be passed an instance of the net.Socket class, a subclass of stream.Duplex, unless the user specifies a socket type other than net.Socket.

A client server pair demonstrating how to listen for the 'upgrade' event.

JS

M request.abort()

Deprecated in: v14.1.0, v13.14.0

稳定性: 0 - Deprecated: Use `request.destroy()` instead.

Marks the request as aborting. Calling this will cause remaining data in the response to be dropped and the socket to be destroyed.

M request.aborted

历史
版本更改
v11.0.0The `aborted` property is no longer a timestamp number.
v17.0.0, v16.12.0Added in: v17.0.0, v16.12.0
稳定性: 0 - Deprecated. Check `request.destroyed` instead.

The request.aborted property will be true if the request has been aborted.

M request.connection

Deprecated in: v13.0.0

稳定性: 0 - Deprecated. Use `request.socket`.

See request.socket.

M request.cork()

Added in: v13.2.0, v12.16.0

See writable.cork().

M request.end([data[, encoding]][, callback])

历史
版本更改
v10.0.0This method now returns a reference to `ClientRequest`.
v0.1.90Added in: v0.1.90

Finishes sending the request. If any parts of the body are unsent, it will flush them to the stream. If the request is chunked, this will send the terminating '0\r\n\r\n'.

If data is specified, it is equivalent to calling request.write(data, encoding) followed by request.end(callback).

If callback is specified, it will be called when the request stream is finished.

M request.destroy([error])

历史
版本更改
v14.5.0The function returns `this` for consistency with other Readable streams.
v0.3.0Added in: v0.3.0
  • error Error Optional, an error to emit with 'error' event.
  • Returns: this

Destroy the request. Optionally emit an 'error' event, and emit a 'close' event. Calling this will cause remaining data in the response to be dropped and the socket to be destroyed.

See writable.destroy() for further details.

M request.destroyed

Added in: v14.1.0, v13.14.0

Is true after request.destroy() has been called.

See writable.destroyed for further details.

M request.finished

Deprecated in: v13.4.0, v12.16.0

稳定性: 0 - Deprecated. Use `request.writableEnded`.

The request.finished property will be true if request.end() has been called. request.end() will automatically be called if the request was initiated via http.get().

M request.flushHeaders()

Added in: v1.6.0

Flushes the request headers.

For efficiency reasons, Node.js normally buffers the request headers until request.end() is called or the first chunk of request data is written. It then tries to pack the request headers and data into a single TCP packet.

That's usually desired (it saves a TCP round-trip), but not when the first data is not sent until possibly much later. request.flushHeaders() bypasses the optimization and kickstarts the request.

M request.getHeader(name)

Added in: v1.6.0

Reads out a header on the request. The name is case-insensitive. The type of the return value depends on the arguments provided to request.setHeader().

JS

M request.getHeaderNames()

Added in: v7.7.0

  • Returns: string[]

Returns an array containing the unique names of the current outgoing headers. All header names are lowercase.

JS

M request.getHeaders()

Added in: v7.7.0

Returns a shallow copy of the current outgoing headers. Since a shallow copy is used, array values may be mutated without additional calls to various header-related http module methods. The keys of the returned object are the header names and the values are the respective header values. All header names are lowercase.

The object returned by the request.getHeaders() method does not prototypically inherit from the JavaScript Object. This means that typical Object methods such as obj.toString(), obj.hasOwnProperty(), and others are not defined and will not work.

JS

M request.getRawHeaderNames()

Added in: v15.13.0, v14.17.0

  • Returns: string[]

Returns an array containing the unique names of the current outgoing raw headers. Header names are returned with their exact casing being set.

JS

M request.hasHeader(name)

Added in: v7.7.0

Returns true if the header identified by name is currently set in the outgoing headers. The header name matching is case-insensitive.

JS

M request.maxHeadersCount

Limits maximum response headers count. If set to 0, no limit will be applied.

M request.path

Added in: v0.4.0

M request.method

Added in: v0.1.97

M request.host

Added in: v14.5.0, v12.19.0

M request.protocol

Added in: v14.5.0, v12.19.0

M request.removeHeader(name)

Added in: v1.6.0

Removes a header that's already defined into headers object.

JS

M request.reusedSocket

Added in: v13.0.0, v12.16.0

  • boolean Whether the request is send through a reused socket.

When sending request through a keep-alive enabled agent, the underlying socket might be reused. But if server closes connection at unfortunate time, client may run into a 'ECONNRESET' error.

JS

By marking a request whether it reused socket or not, we can do automatic error retry base on it.

JS

M request.setHeader(name, value)

Added in: v1.6.0

Sets a single header value for headers object. If this header already exists in the to-be-sent headers, its value will be replaced. Use an array of strings here to send multiple headers with the same name. Non-string values will be stored without modification. Therefore, request.getHeader() may return non-string values. However, the non-string values will be converted to strings for network transmission.

JS

or

JS

When the value is a string an exception will be thrown if it contains characters outside the latin1 encoding.

If you need to pass UTF-8 characters in the value please encode the value using the RFC 8187 standard.

JS

M request.setNoDelay([noDelay])

Added in: v0.5.9

Once a socket is assigned to this request and is connected socket.setNoDelay() will be called.

M request.setSocketKeepAlive([enable][, initialDelay])

Added in: v0.5.9

Once a socket is assigned to this request and is connected socket.setKeepAlive() will be called.

M request.setTimeout(timeout[, callback])

历史
版本更改
v9.0.0Consistently set socket timeout only when the socket connects.
v0.5.9Added in: v0.5.9
  • timeout number Milliseconds before a request times out.
  • callback Function Optional function to be called when a timeout occurs. Same as binding to the 'timeout' event.
  • Returns: http.ClientRequest

Once a socket is assigned to this request and is connected socket.setTimeout() will be called.

M request.socket

Added in: v0.3.0

Reference to the underlying socket. Usually users will not want to access this property. In particular, the socket will not emit 'readable' events because of how the protocol parser attaches to the socket.

JS

This property is guaranteed to be an instance of the net.Socket class, a subclass of stream.Duplex, unless the user specified a socket type other than net.Socket.

M request.uncork()

Added in: v13.2.0, v12.16.0

See writable.uncork().

M request.writableEnded

Added in: v12.9.0

Is true after request.end() has been called. This property does not indicate whether the data has been flushed, for this use request.writableFinished instead.

M request.writableFinished

Added in: v12.7.0

Is true if all data has been flushed to the underlying system, immediately before the 'finish' event is emitted.

M request.write(chunk[, encoding][, callback])

Added in: v0.1.29

Sends a chunk of the body. This method can be called multiple times. If no Content-Length is set, data will automatically be encoded in HTTP Chunked transfer encoding, so that server knows when the data ends. The Transfer-Encoding: chunked header is added. Calling request.end() is necessary to finish sending the request.

The encoding argument is optional and only applies when chunk is a string. Defaults to 'utf8'.

The callback argument is optional and will be called when this chunk of data is flushed, but only if the chunk is non-empty.

Returns true if the entire data was flushed successfully to the kernel buffer. Returns false if all or part of the data was queued in user memory. 'drain' will be emitted when the buffer is free again.

When write function is called with empty string or buffer, it does nothing and waits for more input.

C http.Server

Added in: v0.1.17

E 'checkContinue'

Added in: v0.3.0

Emitted each time a request with an HTTP Expect: 100-continue is received. If this event is not listened for, the server will automatically respond with a 100 Continue as appropriate.

Handling this event involves calling response.writeContinue() if the client should continue to send the request body, or generating an appropriate HTTP response (e.g. 400 Bad Request) if the client should not continue to send the request body.

When this event is emitted and handled, the 'request' event will not be emitted.

E 'checkExpectation'

Added in: v5.5.0

Emitted each time a request with an HTTP Expect header is received, where the value is not 100-continue. If this event is not listened for, the server will automatically respond with a 417 Expectation Failed as appropriate.

When this event is emitted and handled, the 'request' event will not be emitted.

E 'clientError'

历史
版本更改
v12.0.0The default behavior will return a 431 Request Header Fields Too Large if a HPE_HEADER_OVERFLOW error occurs.
v9.4.0The `rawPacket` is the current buffer that just parsed. Adding this buffer to the error object of `'clientError'` event is to make it possible that developers can log the broken packet.
v6.0.0The default action of calling `.destroy()` on the `socket` will no longer take place if there are listeners attached for `'clientError'`.
v0.1.94Added in: v0.1.94

If a client connection emits an 'error' event, it will be forwarded here. Listener of this event is responsible for closing/destroying the underlying socket. For example, one may wish to more gracefully close the socket with a custom HTTP response instead of abruptly severing the connection.

This event is guaranteed to be passed an instance of the net.Socket class, a subclass of stream.Duplex, unless the user specifies a socket type other than net.Socket.

Default behavior is to try close the socket with a HTTP '400 Bad Request', or a HTTP '431 Request Header Fields Too Large' in the case of a HPE_HEADER_OVERFLOW error. If the socket is not writable or headers of the current attached http.ServerResponse has been sent, it is immediately destroyed.

socket is the net.Socket object that the error originated from.

JS

When the 'clientError' event occurs, there is no request or response object, so any HTTP response sent, including response headers and payload, must be written directly to the socket object. Care must be taken to ensure the response is a properly formatted HTTP response message.

err is an instance of Error with two extra columns:

  • bytesParsed: the bytes count of request packet that Node.js may have parsed correctly;
  • rawPacket: the raw packet of current request.

In some cases, the client has already received the response and/or the socket has already been destroyed, like in case of ECONNRESET errors. Before trying to send data to the socket, it is better to check that it is still writable.

JS

E 'close'

Added in: v0.1.4

Emitted when the server closes.

E 'connect'

Added in: v0.7.0

Emitted each time a client requests an HTTP CONNECT method. If this event is not listened for, then clients requesting a CONNECT method will have their connections closed.

This event is guaranteed to be passed an instance of the net.Socket class, a subclass of stream.Duplex, unless the user specifies a socket type other than net.Socket.

After this event is emitted, the request's socket will not have a 'data' event listener, meaning it will need to be bound in order to handle data sent to the server on that socket.

E 'connection'

Added in: v0.1.0

This event is emitted when a new TCP stream is established. socket is typically an object of type net.Socket. Usually users will not want to access this event. In particular, the socket will not emit 'readable' events because of how the protocol parser attaches to the socket. The socket can also be accessed at request.socket.

This event can also be explicitly emitted by users to inject connections into the HTTP server. In that case, any Duplex stream can be passed.

If socket.setTimeout() is called here, the timeout will be replaced with server.keepAliveTimeout when the socket has served a request (if server.keepAliveTimeout is non-zero).

This event is guaranteed to be passed an instance of the net.Socket class, a subclass of stream.Duplex, unless the user specifies a socket type other than net.Socket.

E 'dropRequest'

Added in: v18.7.0

When the number of requests on a socket reaches the threshold of server.maxRequestsPerSocket, the server will drop new requests and emit 'dropRequest' event instead, then send 503 to client.

E 'request'

Added in: v0.1.0

Emitted each time there is a request. There may be multiple requests per connection (in the case of HTTP Keep-Alive connections).

E 'upgrade'

历史
版本更改
v10.0.0Not listening to this event no longer causes the socket to be destroyed if a client sends an Upgrade header.
v0.1.94Added in: v0.1.94

Emitted each time a client requests an HTTP upgrade. Listening to this event is optional and clients cannot insist on a protocol change.

After this event is emitted, the request's socket will not have a 'data' event listener, meaning it will need to be bound in order to handle data sent to the server on that socket.

This event is guaranteed to be passed an instance of the net.Socket class, a subclass of stream.Duplex, unless the user specifies a socket type other than net.Socket.

M server.close([callback])

Added in: v0.1.90

Stops the server from accepting new connections. See net.Server.close().

M server.closeAllConnections()

Added in: v18.2.0

Closes all connections connected to this server.

M server.closeIdleConnections()

Added in: v18.2.0

Closes all connections connected to this server which are not sending a request or waiting for a response.

M server.headersTimeout

Added in: v11.3.0, v10.14.0

Limit the amount of time the parser will wait to receive the complete HTTP headers.

If the timeout expires, the server responds with status 408 without forwarding the request to the request listener and then closes the connection.

It must be set to a non-zero value (e.g. 120 seconds) to protect against potential Denial-of-Service attacks in case the server is deployed without a reverse proxy in front.

M server.listen()

Starts the HTTP server listening for connections. This method is identical to server.listen() from net.Server.

M server.listening

Added in: v5.7.0

  • boolean Indicates whether or not the server is listening for connections.

M server.maxHeadersCount

Added in: v0.7.0

Limits maximum incoming headers count. If set to 0, no limit will be applied.

M server.requestTimeout

历史
版本更改
v18.0.0The default request timeout changed from no timeout to 300s (5 minutes).
v14.11.0Added in: v14.11.0

Sets the timeout value in milliseconds for receiving the entire request from the client.

If the timeout expires, the server responds with status 408 without forwarding the request to the request listener and then closes the connection.

It must be set to a non-zero value (e.g. 120 seconds) to protect against potential Denial-of-Service attacks in case the server is deployed without a reverse proxy in front.

M server.setTimeout([msecs][, callback])

历史
版本更改
v13.0.0The default timeout changed from 120s to 0 (no timeout).
v0.9.12Added in: v0.9.12

Sets the timeout value for sockets, and emits a 'timeout' event on the Server object, passing the socket as an argument, if a timeout occurs.

If there is a 'timeout' event listener on the Server object, then it will be called with the timed-out socket as an argument.

By default, the Server does not timeout sockets. However, if a callback is assigned to the Server's 'timeout' event, timeouts must be handled explicitly.

M server.maxRequestsPerSocket

Added in: v16.10.0

  • number Requests per socket. Default: 0 (no limit)

The maximum number of requests socket can handle before closing keep alive connection.

A value of 0 will disable the limit.

When the limit is reached it will set the Connection header value to close, but will not actually close the connection, subsequent requests sent after the limit is reached will get 503 Service Unavailable as a response.

M server.timeout

历史
版本更改
v13.0.0The default timeout changed from 120s to 0 (no timeout).
v0.9.12Added in: v0.9.12
  • number Timeout in milliseconds. Default: 0 (no timeout)

The number of milliseconds of inactivity before a socket is presumed to have timed out.

A value of 0 will disable the timeout behavior on incoming connections.

The socket timeout logic is set up on connection, so changing this value only affects new connections to the server, not any existing connections.

M server.keepAliveTimeout

Added in: v8.0.0

  • number Timeout in milliseconds. Default: 5000 (5 seconds).

The number of milliseconds of inactivity a server needs to wait for additional incoming data, after it has finished writing the last response, before a socket will be destroyed. If the server receives new data before the keep-alive timeout has fired, it will reset the regular inactivity timeout, i.e., server.timeout.

A value of 0 will disable the keep-alive timeout behavior on incoming connections. A value of 0 makes the http server behave similarly to Node.js versions prior to 8.0.0, which did not have a keep-alive timeout.

The socket timeout logic is set up on connection, so changing this value only affects new connections to the server, not any existing connections.

C http.ServerResponse

Added in: v0.1.17

This object is created internally by an HTTP server, not by the user. It is passed as the second parameter to the 'request' event.

E 'close'

Added in: v0.6.7

Indicates that the response is completed, or its underlying connection was terminated prematurely (before the response completion).

E 'finish'

Added in: v0.3.6

Emitted when the response has been sent. More specifically, this event is emitted when the last segment of the response headers and body have been handed off to the operating system for transmission over the network. It does not imply that the client has received anything yet.

M response.addTrailers(headers)

Added in: v0.3.0

This method adds HTTP trailing headers (a header but at the end of the message) to the response.

Trailers will only be emitted if chunked encoding is used for the response; if it is not (e.g. if the request was HTTP/1.0), they will be silently discarded.

HTTP requires the Trailer header to be sent in order to emit trailers, with a list of the header fields in its value. E.g.,

JS

Attempting to set a header field name or value that contains invalid characters will result in a TypeError being thrown.

M response.connection

Deprecated in: v13.0.0

稳定性: 0 - Deprecated. Use `response.socket`.

See response.socket.

M response.cork()

Added in: v13.2.0, v12.16.0

See writable.cork().

M response.end([data[, encoding]][, callback])

历史
版本更改
v10.0.0This method now returns a reference to `ServerResponse`.
v0.1.90Added in: v0.1.90

This method signals to the server that all of the response headers and body have been sent; that server should consider this message complete. The method, response.end(), MUST be called on each response.

If data is specified, it is similar in effect to calling response.write(data, encoding) followed by response.end(callback).

If callback is specified, it will be called when the response stream is finished.

M response.finished

Deprecated in: v13.4.0, v12.16.0

稳定性: 0 - Deprecated. Use `response.writableEnded`.

The response.finished property will be true if response.end() has been called.

M response.flushHeaders()

Added in: v1.6.0

Flushes the response headers. See also: request.flushHeaders().

M response.getHeader(name)

Added in: v0.4.0

Reads out a header that's already been queued but not sent to the client. The name is case-insensitive. The type of the return value depends on the arguments provided to response.setHeader().

JS

M response.getHeaderNames()

Added in: v7.7.0

  • Returns: string[]

Returns an array containing the unique names of the current outgoing headers. All header names are lowercase.

JS

M response.getHeaders()

Added in: v7.7.0

Returns a shallow copy of the current outgoing headers. Since a shallow copy is used, array values may be mutated without additional calls to various header-related http module methods. The keys of the returned object are the header names and the values are the respective header values. All header names are lowercase.

The object returned by the response.getHeaders() method does not prototypically inherit from the JavaScript Object. This means that typical Object methods such as obj.toString(), obj.hasOwnProperty(), and others are not defined and will not work.

JS

M response.hasHeader(name)

Added in: v7.7.0

Returns true if the header identified by name is currently set in the outgoing headers. The header name matching is case-insensitive.

JS

M response.headersSent

Added in: v0.9.3

Boolean (read-only). True if headers were sent, false otherwise.

M response.removeHeader(name)

Added in: v0.4.0

Removes a header that's queued for implicit sending.

JS

M response.req

Added in: v15.7.0

A reference to the original HTTP request object.

M response.sendDate

Added in: v0.7.5

When true, the Date header will be automatically generated and sent in the response if it is not already present in the headers. Defaults to true.

This should only be disabled for testing; HTTP requires the Date header in responses.

M response.setHeader(name, value)

Added in: v0.4.0

Returns the response object.

Sets a single header value for implicit headers. If this header already exists in the to-be-sent headers, its value will be replaced. Use an array of strings here to send multiple headers with the same name. Non-string values will be stored without modification. Therefore, response.getHeader() may return non-string values. However, the non-string values will be converted to strings for network transmission. The same response object is returned to the caller, to enable call chaining.

JS

or

JS

Attempting to set a header field name or value that contains invalid characters will result in a TypeError being thrown.

When headers have been set with response.setHeader(), they will be merged with any headers passed to response.writeHead(), with the headers passed to response.writeHead() given precedence.

JS

If response.writeHead() method is called and this method has not been called, it will directly write the supplied header values onto the network channel without caching internally, and the response.getHeader() on the header will not yield the expected result. If progressive population of headers is desired with potential future retrieval and modification, use response.setHeader() instead of response.writeHead().

M response.setTimeout(msecs[, callback])

Added in: v0.9.12

Sets the Socket's timeout value to msecs. If a callback is provided, then it is added as a listener on the 'timeout' event on the response object.

If no 'timeout' listener is added to the request, the response, or the server, then sockets are destroyed when they time out. If a handler is assigned to the request, the response, or the server's 'timeout' events, timed out sockets must be handled explicitly.

M response.socket

Added in: v0.3.0

Reference to the underlying socket. Usually users will not want to access this property. In particular, the socket will not emit 'readable' events because of how the protocol parser attaches to the socket. After response.end(), the property is nulled.

JS

This property is guaranteed to be an instance of the net.Socket class, a subclass of stream.Duplex, unless the user specified a socket type other than net.Socket.

M response.statusCode

Added in: v0.4.0

When using implicit headers (not calling response.writeHead() explicitly), this property controls the status code that will be sent to the client when the headers get flushed.

JS

After response header was sent to the client, this property indicates the status code which was sent out.

M response.statusMessage

Added in: v0.11.8

When using implicit headers (not calling response.writeHead() explicitly), this property controls the status message that will be sent to the client when the headers get flushed. If this is left as undefined then the standard message for the status code will be used.

JS

After response header was sent to the client, this property indicates the status message which was sent out.

M response.uncork()

Added in: v13.2.0, v12.16.0

See writable.uncork().

M response.writableEnded

Added in: v12.9.0

Is true after response.end() has been called. This property does not indicate whether the data has been flushed, for this use response.writableFinished instead.

M response.writableFinished

Added in: v12.7.0

Is true if all data has been flushed to the underlying system, immediately before the 'finish' event is emitted.

M response.write(chunk[, encoding][, callback])

Added in: v0.1.29

If this method is called and response.writeHead() has not been called, it will switch to implicit header mode and flush the implicit headers.

This sends a chunk of the response body. This method may be called multiple times to provide successive parts of the body.

In the node:http module, the response body is omitted when the request is a HEAD request. Similarly, the 204 and 304 responses must not include a message body.

chunk can be a string or a buffer. If chunk is a string, the second parameter specifies how to encode it into a byte stream. callback will be called when this chunk of data is flushed.

This is the raw HTTP body and has nothing to do with higher-level multi-part body encodings that may be used.

The first time response.write() is called, it will send the buffered header information and the first chunk of the body to the client. The second time response.write() is called, Node.js assumes data will be streamed, and sends the new data separately. That is, the response is buffered up to the first chunk of the body.

Returns true if the entire data was flushed successfully to the kernel buffer. Returns false if all or part of the data was queued in user memory. 'drain' will be emitted when the buffer is free again.

M response.writeContinue()

Added in: v0.3.0

Sends an HTTP/1.1 100 Continue message to the client, indicating that the request body should be sent. See the 'checkContinue' event on Server.

M response.writeEarlyHints(hints[, callback])

历史
版本更改
v18.11.0Allow passing hints as an object.
v18.11.0Added in: v18.11.0

Sends an HTTP/1.1 103 Early Hints message to the client with a Link header, indicating that the user agent can preload/preconnect the linked resources. The hints is an object containing the values of headers to be sent with early hints message. The optional callback argument will be called when the response message has been written.

Example

JS

M response.writeHead(statusCode[, statusMessage][, headers])

历史
版本更改
v14.14.0Allow passing headers as an array.
v11.10.0, v10.17.0Return `this` from `writeHead()` to allow chaining with `end()`.
v5.11.0, v4.4.5A `RangeError` is thrown if `statusCode` is not a number in the range `[100, 999]`.
v0.1.30Added in: v0.1.30

Sends a response header to the request. The status code is a 3-digit HTTP status code, like 404. The last argument, headers, are the response headers. Optionally one can give a human-readable statusMessage as the second argument.

headers may be an Array where the keys and values are in the same list. It is not a list of tuples. So, the even-numbered offsets are key values, and the odd-numbered offsets are the associated values. The array is in the same format as request.rawHeaders.

Returns a reference to the ServerResponse, so that calls can be chained.

JS

This method must only be called once on a message and it must be called before response.end() is called.

If response.write() or response.end() are called before calling this, the implicit/mutable headers will be calculated and call this function.

When headers have been set with response.setHeader(), they will be merged with any headers passed to response.writeHead(), with the headers passed to response.writeHead() given precedence.

If this method is called and response.setHeader() has not been called, it will directly write the supplied header values onto the network channel without caching internally, and the response.getHeader() on the header will not yield the expected result. If progressive population of headers is desired with potential future retrieval and modification, use response.setHeader() instead.

JS

Content-Length is read in bytes, not characters. Use Buffer.byteLength() to determine the length of the body in bytes. Node.js will check whether Content-Length and the length of the body which has been transmitted are equal or not.

Attempting to set a header field name or value that contains invalid characters will result in a [Error][] being thrown.

M response.writeProcessing()

Added in: v10.0.0

Sends a HTTP/1.1 102 Processing message to the client, indicating that the request body should be sent.

C http.IncomingMessage

历史
版本更改
v15.5.0The `destroyed` value returns `true` after the incoming data is consumed.
v13.1.0, v12.16.0The `readableHighWaterMark` value mirrors that of the socket.
v0.1.17Added in: v0.1.17

An IncomingMessage object is created by http.Server or http.ClientRequest and passed as the first argument to the 'request' and 'response' event respectively. It may be used to access response status, headers, and data.

Different from its socket value which is a subclass of stream.Duplex, the IncomingMessage itself extends stream.Readable and is created separately to parse and emit the incoming HTTP headers and payload, as the underlying socket may be reused multiple times in case of keep-alive.

E 'aborted'

Deprecated in: v17.0.0, v16.12.0

稳定性: 0 - Deprecated. Listen for `'close'` event instead.

Emitted when the request has been aborted.

E 'close'

历史
版本更改
v16.0.0The close event is now emitted when the request has been completed and not when the underlying socket is closed.
v0.4.2Added in: v0.4.2

Emitted when the request has been completed.

M message.aborted

Deprecated in: v17.0.0, v16.12.0

稳定性: 0 - Deprecated. Check `message.destroyed` from {stream.Readable}.

The message.aborted property will be true if the request has been aborted.

M message.complete

Added in: v0.3.0

The message.complete property will be true if a complete HTTP message has been received and successfully parsed.

This property is particularly useful as a means of determining if a client or server fully transmitted a message before a connection was terminated:

JS

M message.connection

Deprecated in: v16.0.0

稳定性: 0 - Deprecated. Use `message.socket`.

Alias for message.socket.

M message.destroy([error])

历史
版本更改
v14.5.0, v12.19.0The function returns `this` for consistency with other Readable streams.
v0.3.0Added in: v0.3.0

Calls destroy() on the socket that received the IncomingMessage. If error is provided, an 'error' event is emitted on the socket and error is passed as an argument to any listeners on the event.

M message.headers

历史
版本更改
v15.1.0`message.headers` is now lazily computed using an accessor property on the prototype and is no longer enumerable.
v0.1.5Added in: v0.1.5

The request/response headers object.

Key-value pairs of header names and values. Header names are lower-cased.

JS

Duplicates in raw headers are handled in the following ways, depending on the header name:

  • Duplicates of age, authorization, content-length, content-type, etag, expires, from, host, if-modified-since, if-unmodified-since, last-modified, location, max-forwards, proxy-authorization, referer, retry-after, server, or user-agent are discarded.
  • set-cookie is always an array. Duplicates are added to the array.
  • For duplicate cookie headers, the values are joined together with ; .
  • For all other headers, the values are joined together with , .

M message.headersDistinct

Added in: v18.3.0

Similar to message.headers, but there is no join logic and the values are always arrays of strings, even for headers received just once.

JS

M message.httpVersion

Added in: v0.1.1

In case of server request, the HTTP version sent by the client. In the case of client response, the HTTP version of the connected-to server. Probably either '1.1' or '1.0'.

Also message.httpVersionMajor is the first integer and message.httpVersionMinor is the second.

M message.method

Added in: v0.1.1

Only valid for request obtained from http.Server.

The request method as a string. Read only. Examples: 'GET', 'DELETE'.

M message.rawHeaders

Added in: v0.11.6

  • string[]

The raw request/response headers list exactly as they were received.

The keys and values are in the same list. It is not a list of tuples. So, the even-numbered offsets are key values, and the odd-numbered offsets are the associated values.

Header names are not lowercased, and duplicates are not merged.

JS

M message.rawTrailers

Added in: v0.11.6

  • string[]

The raw request/response trailer keys and values exactly as they were received. Only populated at the 'end' event.

M message.setTimeout(msecs[, callback])

Added in: v0.5.9

Calls message.socket.setTimeout(msecs, callback).

M message.socket

Added in: v0.3.0

The net.Socket object associated with the connection.

With HTTPS support, use request.socket.getPeerCertificate() to obtain the client's authentication details.

This property is guaranteed to be an instance of the net.Socket class, a subclass of stream.Duplex, unless the user specified a socket type other than net.Socket or internally nulled.

M message.statusCode

Added in: v0.1.1

Only valid for response obtained from http.ClientRequest.

The 3-digit HTTP response status code. E.G. 404.

M message.statusMessage

Added in: v0.11.10

Only valid for response obtained from http.ClientRequest.

The HTTP response status message (reason phrase). E.G. OK or Internal Server Error.

M message.trailers

Added in: v0.3.0

The request/response trailers object. Only populated at the 'end' event.

M message.trailersDistinct

Added in: v18.3.0

Similar to message.trailers, but there is no join logic and the values are always arrays of strings, even for headers received just once. Only populated at the 'end' event.

M message.url

Added in: v0.1.90

Only valid for request obtained from http.Server.

Request URL string. This contains only the URL that is present in the actual HTTP request. Take the following request:

HTTP

To parse the URL into its parts:

JS

When request.url is '/status?name=ryan' and request.headers.host is 'localhost:3000':

BASH

C http.OutgoingMessage

Added in: v0.1.17

This class serves as the parent class of http.ClientRequest and http.ServerResponse. It is an abstract outgoing message from the perspective of the participants of an HTTP transaction.

E 'drain'

Added in: v0.3.6

Emitted when the buffer of the message is free again.

E 'finish'

Added in: v0.1.17

Emitted when the transmission is finished successfully.

E 'prefinish'

Added in: v0.11.6

Emitted after outgoingMessage.end() is called. When the event is emitted, all data has been processed but not necessarily completely flushed.

M outgoingMessage.addTrailers(headers)

Added in: v0.3.0

Adds HTTP trailers (headers but at the end of the message) to the message.

Trailers will only be emitted if the message is chunked encoded. If not, the trailers will be silently discarded.

HTTP requires the Trailer header to be sent to emit trailers, with a list of header field names in its value, e.g.

JS

Attempting to set a header field name or value that contains invalid characters will result in a TypeError being thrown.

M outgoingMessage.appendHeader(name, value)

Added in: v18.3.0

Append a single header value for the header object.

If the value is an array, this is equivalent of calling this method multiple times.

If there were no previous value for the header, this is equivalent of calling outgoingMessage.setHeader(name, value).

Depending of the value of options.uniqueHeaders when the client request or the server were created, this will end up in the header being sent multiple times or a single time with values joined using ; .

M outgoingMessage.connection

Deprecated in: v15.12.0, v14.17.1

稳定性: 0 - Deprecated: Use `outgoingMessage.socket` instead.

Alias of outgoingMessage.socket.

M outgoingMessage.cork()

Added in: v13.2.0, v12.16.0

See writable.cork().

M outgoingMessage.destroy([error])

Added in: v0.3.0

  • error Error Optional, an error to emit with error event
  • Returns: this

Destroys the message. Once a socket is associated with the message and is connected, that socket will be destroyed as well.

M outgoingMessage.end(chunk[, encoding][, callback])

历史
版本更改
v0.11.6add `callback` argument.
v0.1.90Added in: v0.1.90

Finishes the outgoing message. If any parts of the body are unsent, it will flush them to the underlying system. If the message is chunked, it will send the terminating chunk 0\r\n\r\n, and send the trailers (if any).

If chunk is specified, it is equivalent to calling outgoingMessage.write(chunk, encoding), followed by outgoingMessage.end(callback).

If callback is provided, it will be called when the message is finished (equivalent to a listener of the 'finish' event).

M outgoingMessage.flushHeaders()

Added in: v1.6.0

Flushes the message headers.

For efficiency reason, Node.js normally buffers the message headers until outgoingMessage.end() is called or the first chunk of message data is written. It then tries to pack the headers and data into a single TCP packet.

It is usually desired (it saves a TCP round-trip), but not when the first data is not sent until possibly much later. outgoingMessage.flushHeaders() bypasses the optimization and kickstarts the message.

M outgoingMessage.getHeader(name)

Added in: v0.4.0

Gets the value of the HTTP header with the given name. If that header is not set, the returned value will be undefined.

M outgoingMessage.getHeaderNames()

Added in: v7.7.0

  • Returns string[]

Returns an array containing the unique names of the current outgoing headers. All names are lowercase.

M outgoingMessage.getHeaders()

Added in: v7.7.0

Returns a shallow copy of the current outgoing headers. Since a shallow copy is used, array values may be mutated without additional calls to various header-related HTTP module methods. The keys of the returned object are the header names and the values are the respective header values. All header names are lowercase.

The object returned by the outgoingMessage.getHeaders() method does not prototypically inherit from the JavaScript Object. This means that typical Object methods such as obj.toString(), obj.hasOwnProperty(), and others are not defined and will not work.

JS

M outgoingMessage.hasHeader(name)

Added in: v7.7.0

Returns true if the header identified by name is currently set in the outgoing headers. The header name is case-insensitive.

JS

M outgoingMessage.headersSent

Added in: v0.9.3

Read-only. true if the headers were sent, otherwise false.

M outgoingMessage.pipe()

Added in: v9.0.0

Overrides the stream.pipe() method inherited from the legacy Stream class which is the parent class of http.OutgoingMessage.

Calling this method will throw an Error because outgoingMessage is a write-only stream.

M outgoingMessage.removeHeader(name)

Added in: v0.4.0

Removes a header that is queued for implicit sending.

JS

M outgoingMessage.setHeader(name, value)

Added in: v0.4.0

Sets a single header value. If the header already exists in the to-be-sent headers, its value will be replaced. Use an array of strings to send multiple headers with the same name.

M outgoingMessage.setTimeout(msesc[, callback])

Added in: v0.9.12

  • msesc number
  • callback Function Optional function to be called when a timeout occurs. Same as binding to the timeout event.
  • Returns: this

Once a socket is associated with the message and is connected, socket.setTimeout() will be called with msecs as the first parameter.

M outgoingMessage.socket

Added in: v0.3.0

Reference to the underlying socket. Usually, users will not want to access this property.

After calling outgoingMessage.end(), this property will be nulled.

M outgoingMessage.uncork()

Added in: v13.2.0, v12.16.0

See writable.uncork()

M outgoingMessage.writableCorked

Added in: v13.2.0, v12.16.0

The number of times outgoingMessage.cork() has been called.

M outgoingMessage.writableEnded

Added in: v12.9.0

Is true if outgoingMessage.end() has been called. This property does not indicate whether the data has been flushed. For that purpose, use message.writableFinished instead.

M outgoingMessage.writableFinished

Added in: v12.7.0

Is true if all data has been flushed to the underlying system.

M outgoingMessage.writableHighWaterMark

Added in: v12.9.0

The highWaterMark of the underlying socket if assigned. Otherwise, the default buffer level when writable.write() starts returning false (16384).

M outgoingMessage.writableLength

Added in: v12.9.0

The number of buffered bytes.

M outgoingMessage.writableObjectMode

Added in: v12.9.0

Always false.

M outgoingMessage.write(chunk[, encoding][, callback])

历史
版本更改
v0.11.6The `callback` argument was added.
v0.1.29Added in: v0.1.29

Sends a chunk of the body. This method can be called multiple times.

The encoding argument is only relevant when chunk is a string. Defaults to 'utf8'.

The callback argument is optional and will be called when this chunk of data is flushed.

Returns true if the entire data was flushed successfully to the kernel buffer. Returns false if all or part of the data was queued in the user memory. The 'drain' event will be emitted when the buffer is free again.

M http.METHODS

Added in: v0.11.8

  • string[]

A list of the HTTP methods that are supported by the parser.

M http.STATUS_CODES

Added in: v0.1.22

A collection of all the standard HTTP response status codes, and the short description of each. For example, http.STATUS_CODES[404] === 'Not Found'.

M http.createServer([options][, requestListener])

历史
版本更改
v18.0.0The `requestTimeout`, `headersTimeout`, `keepAliveTimeout`, and `connectionsCheckingInterval` options are supported now.
v18.0.0The `noDelay` option now defaults to `true`.
v17.7.0The `noDelay`, `keepAlive` and `keepAliveInitialDelay` options are supported now.
v13.8.0, v12.15.0, v10.19.0The `insecureHTTPParser` option is supported now.
v13.3.0The `maxHeaderSize` option is supported now.
v9.6.0, v8.12.0The `options` argument is supported now.
v0.1.13Added in: v0.1.13
  • options Object

    • IncomingMessage http.IncomingMessage Specifies the IncomingMessage class to be used. Useful for extending the original IncomingMessage. Default: IncomingMessage.
    • ServerResponse http.ServerResponse Specifies the ServerResponse class to be used. Useful for extending the original ServerResponse. Default:ServerResponse.
    • requestTimeout: Sets the timeout value in milliseconds for receiving the entire request from the client. See server.requestTimeout for more information. Default: 300000.
    • headersTimeout: Sets the timeout value in milliseconds for receiving the complete HTTP headers from the client. See server.headersTimeout for more information. Default: 60000.
    • keepAliveTimeout: The number of milliseconds of inactivity a server needs to wait for additional incoming data, after it has finished writing the last response, before a socket will be destroyed. See server.keepAliveTimeout for more information. Default: 5000.
    • connectionsCheckingInterval: Sets the interval value in milliseconds to check for request and headers timeout in incomplete requests. Default: 30000.
    • insecureHTTPParser boolean Use an insecure HTTP parser that accepts invalid HTTP headers when true. Using the insecure parser should be avoided. See --insecure-http-parser for more information. Default: false
    • maxHeaderSize number Optionally overrides the value of --max-http-header-size for requests received by this server, i.e. the maximum length of request headers in bytes. Default: 16384 (16 KiB).
    • noDelay boolean If set to true, it disables the use of Nagle's algorithm immediately after a new incoming connection is received. Default: true.
    • keepAlive boolean If set to true, it enables keep-alive functionality on the socket immediately after a new incoming connection is received, similarly on what is done in [socket.setKeepAlive([enable][, initialDelay])][socket.setKeepAlive(enable, initialDelay)]. Default: false.
    • keepAliveInitialDelay number If set to a positive number, it sets the initial delay before the first keepalive probe is sent on an idle socket. Default: 0.
    • uniqueHeaders Array A list of response headers that should be sent only once. If the header's value is an array, the items will be joined using ; .
  • requestListener Function

  • Returns: http.Server

Returns a new instance of http.Server.

The requestListener is a function which is automatically added to the 'request' event.

CJS
CJS

M http.get(options[, callback])

M http.get(url[, options][, callback])

历史
版本更改
v10.9.0The `url` parameter can now be passed along with a separate `options` object.
v7.5.0The `options` parameter can be a WHATWG `URL` object.
v0.3.6Added in: v0.3.6

Since most requests are GET requests without bodies, Node.js provides this convenience method. The only difference between this method and http.request() is that it sets the method to GET and calls req.end() automatically. The callback must take care to consume the response data for reasons stated in http.ClientRequest section.

The callback is invoked with a single argument that is an instance of http.IncomingMessage.

JSON fetching example:

JS

M http.globalAgent

Added in: v0.5.9

Global instance of Agent which is used as the default for all HTTP client requests.

M http.maxHeaderSize

Added in: v11.6.0, v10.15.0

Read-only property specifying the maximum allowed size of HTTP headers in bytes. Defaults to 16 KiB. Configurable using the --max-http-header-size CLI option.

This can be overridden for servers and client requests by passing the maxHeaderSize option.

M http.request(options[, callback])

M http.request(url[, options][, callback])

历史
版本更改
v16.7.0, v14.18.0When using a `URL` object parsed username and password will now be properly URI decoded.
v15.3.0, v14.17.0It is possible to abort a request with an AbortSignal.
v13.8.0, v12.15.0, v10.19.0The `insecureHTTPParser` option is supported now.
v13.3.0The `maxHeaderSize` option is supported now.
v10.9.0The `url` parameter can now be passed along with a separate `options` object.
v7.5.0The `options` parameter can be a WHATWG `URL` object.
v0.3.6Added in: v0.3.6
  • url string | URL
  • options Object
    • agent http.Agent | boolean Controls Agent behavior. Possible values:
      • undefined (default): use http.globalAgent for this host and port.
      • Agent object: explicitly use the passed in Agent.
      • false: causes a new Agent with default values to be used.
    • auth string Basic authentication ('user:password') to compute an Authorization header.
    • createConnection Function A function that produces a socket/stream to use for the request when the agent option is not used. This can be used to avoid creating a custom Agent class just to override the default createConnection function. See agent.createConnection() for more details. Any Duplex stream is a valid return value.
    • defaultPort number Default port for the protocol. Default:agent.defaultPort if an Agent is used, else undefined.
    • family number IP address family to use when resolving host or hostname. Valid values are 4 or 6. When unspecified, both IP v4 and v6 will be used.
    • headers Object An object containing request headers.
    • hints number Optional dns.lookup() hints.
    • host string A domain name or IP address of the server to issue the request to. Default: 'localhost'.
    • hostname string Alias for host. To support url.parse(), hostname will be used if both host and hostname are specified.
    • insecureHTTPParser boolean Use an insecure HTTP parser that accepts invalid HTTP headers when true. Using the insecure parser should be avoided. See --insecure-http-parser for more information. Default: false
    • localAddress string Local interface to bind for network connections.
    • localPort number Local port to connect from.
    • lookup Function Custom lookup function. Default: dns.lookup().
    • maxHeaderSize number Optionally overrides the value of --max-http-header-size (the maximum length of response headers in bytes) for responses received from the server. Default: 16384 (16 KiB).
    • method string A string specifying the HTTP request method. Default:'GET'.
    • path string Request path. Should include query string if any. E.G. '/index.html?page=12'. An exception is thrown when the request path contains illegal characters. Currently, only spaces are rejected but that may change in the future. Default: '/'.
    • port number Port of remote server. Default: defaultPort if set, else 80.
    • protocol string Protocol to use. Default: 'http:'.
    • setHost boolean: Specifies whether or not to automatically add the Host header. Defaults to true.
    • signal AbortSignal: An AbortSignal that may be used to abort an ongoing request.
    • socketPath string Unix domain socket. Cannot be used if one of host or port is specified, as those specify a TCP Socket.
    • timeout number: A number specifying the socket timeout in milliseconds. This will set the timeout before the socket is connected.
    • uniqueHeaders Array A list of request headers that should be sent only once. If the header's value is an array, the items will be joined using ; .
  • callback Function
  • Returns: http.ClientRequest

options in socket.connect() are also supported.

Node.js maintains several connections per server to make HTTP requests. This function allows one to transparently issue requests.

url can be a string or a URL object. If url is a string, it is automatically parsed with new URL(). If it is a URL object, it will be automatically converted to an ordinary options object.

If both url and options are specified, the objects are merged, with the options properties taking precedence.

The optional callback parameter will be added as a one-time listener for the 'response' event.

http.request() returns an instance of the http.ClientRequest class. The ClientRequest instance is a writable stream. If one needs to upload a file with a POST request, then write to the ClientRequest object.

JS

In the example req.end() was called. With http.request() one must always call req.end() to signify the end of the request - even if there is no data being written to the request body.

If any error is encountered during the request (be that with DNS resolution, TCP level errors, or actual HTTP parse errors) an 'error' event is emitted on the returned request object. As with all 'error' events, if no listeners are registered the error will be thrown.

There are a few special headers that should be noted.

  • Sending a 'Connection: keep-alive' will notify Node.js that the connection to the server should be persisted until the next request.

  • Sending a 'Content-Length' header will disable the default chunked encoding.

  • Sending an 'Expect' header will immediately send the request headers. Usually, when sending 'Expect: 100-continue', both a timeout and a listener for the 'continue' event should be set. See RFC 2616 Section 8.2.3 for more information.

  • Sending an Authorization header will override using the auth option to compute basic authentication.

Example using a URL as options:

JS

In a successful request, the following events will be emitted in the following order:

  • 'socket'
  • 'response'
    • 'data' any number of times, on the res object ('data' will not be emitted at all if the response body is empty, for instance, in most redirects)
    • 'end' on the res object
  • 'close'

In the case of a connection error, the following events will be emitted:

  • 'socket'
  • 'error'
  • 'close'

In the case of a premature connection close before the response is received, the following events will be emitted in the following order:

  • 'socket'
  • 'error' with an error with message 'Error: socket hang up' and code 'ECONNRESET'
  • 'close'

In the case of a premature connection close after the response is received, the following events will be emitted in the following order:

  • 'socket'
  • 'response'
    • 'data' any number of times, on the res object
  • (connection closed here)
  • 'aborted' on the res object
  • 'error' on the res object with an error with message 'Error: aborted' and code 'ECONNRESET'.
  • 'close'
  • 'close' on the res object

If req.destroy() is called before a socket is assigned, the following events will be emitted in the following order:

  • (req.destroy() called here)
  • 'error' with an error with message 'Error: socket hang up' and code 'ECONNRESET'
  • 'close'

If req.destroy() is called before the connection succeeds, the following events will be emitted in the following order:

  • 'socket'
  • (req.destroy() called here)
  • 'error' with an error with message 'Error: socket hang up' and code 'ECONNRESET'
  • 'close'

If req.destroy() is called after the response is received, the following events will be emitted in the following order:

  • 'socket'
  • 'response'
    • 'data' any number of times, on the res object
  • (req.destroy() called here)
  • 'aborted' on the res object
  • 'error' on the res object with an error with message 'Error: aborted' and code 'ECONNRESET'.
  • 'close'
  • 'close' on the res object

If req.abort() is called before a socket is assigned, the following events will be emitted in the following order:

  • (req.abort() called here)
  • 'abort'
  • 'close'

If req.abort() is called before the connection succeeds, the following events will be emitted in the following order:

  • 'socket'
  • (req.abort() called here)
  • 'abort'
  • 'error' with an error with message 'Error: socket hang up' and code 'ECONNRESET'
  • 'close'

If req.abort() is called after the response is received, the following events will be emitted in the following order:

  • 'socket'
  • 'response'
    • 'data' any number of times, on the res object
  • (req.abort() called here)
  • 'abort'
  • 'aborted' on the res object
  • 'error' on the res object with an error with message 'Error: aborted' and code 'ECONNRESET'.
  • 'close'
  • 'close' on the res object

Setting the timeout option or using the setTimeout() function will not abort the request or do anything besides add a 'timeout' event.

Passing an AbortSignal and then calling abort on the corresponding AbortController will behave the same way as calling .destroy() on the request itself.

M http.validateHeaderName(name)

Added in: v14.3.0

Performs the low-level validations on the provided name that are done when res.setHeader(name, value) is called.

Passing illegal value as name will result in a TypeError being thrown, identified by code: 'ERR_INVALID_HTTP_TOKEN'.

It is not necessary to use this method before passing headers to an HTTP request or response. The HTTP module will automatically validate such headers. Examples:

Example:

JS

M http.validateHeaderValue(name, value)

Added in: v14.3.0

Performs the low-level validations on the provided value that are done when res.setHeader(name, value) is called.

Passing illegal value as value will result in a TypeError being thrown.

  • Undefined value error is identified by code: 'ERR_HTTP_INVALID_HEADER_VALUE'.
  • Invalid value character error is identified by code: 'ERR_INVALID_CHAR'.

It is not necessary to use this method before passing headers to an HTTP request or response. The HTTP module will automatically validate such headers.

Examples:

JS

M http.setMaxIdleHTTPParsers

Added in: v18.8.0

Set the maximum number of idle HTTP parsers. Default: 1000.