Google Git
Sign in
apache / httpcomponents-core / refs/heads/4.4.x / . / src / docbkx / nio.xml
blob: beb3963cff0be7a5b122282b758b9fcad8e5fd36 [file] [log] [blame]
<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE preface PUBLIC "-//OASIS//DTD DocBook XML V4.4//EN"
"http://www.oasis-open.org/docbook/xml/4.4/docbookx.dtd">
<!--
====================================================================
Licensed to the Apache Software Foundation (ASF) under one
or more contributor license agreements. See the NOTICE file
distributed with this work for additional information
regarding copyright ownership. The ASF licenses this file
to you 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.
====================================================================
-->
<chapter id="nio">
<title>Asynchronous I/O based on NIO</title>
<para>
Asynchronous I/O model may be more appropriate for those scenarios where raw data throughput
is less important than the ability to handle thousands of simultaneous connections in
a scalable, resource efficient manner. Asynchronous I/O is arguably more complex and usually
requires a special care when dealing with large message payloads.
</para>
<section>
<title>Differences from other I/O frameworks</title>
<para>
Solves similar problems as other frameworks, but has certain distinct features:
</para>
<itemizedlist>
<listitem>
<para>
minimalistic, optimized for data volume intensive protocols such as HTTP.
</para>
</listitem>
<listitem>
<para>
efficient memory management: data consumer can read is only as much input data as it
can process without having to allocate more memory.
</para>
</listitem>
<listitem>
<para>
direct access to the NIO channels where possible.
</para>
</listitem>
</itemizedlist>
</section>
<section>
<title>I/O reactor</title>
<para>
HttpCore NIO is based on the Reactor pattern as described by Doug Lea. The purpose of I/O
reactors is to react to I/O events and to dispatch event notifications to individual I/O
sessions. The main idea of I/O reactor pattern is to break away from the one thread per
connection model imposed by the classic blocking I/O model. The <interfacename>IOReactor
</interfacename> interface represents an abstract object which implements the Reactor pattern.
Internally, <interfacename>IOReactor</interfacename> implementations encapsulate
functionality of the NIO <classname>java.nio.channels.Selector</classname>.
</para>
<para>
I/O reactors usually employ a small number of dispatch threads (often as few as one) to
dispatch I/O event notifications to a much greater number (often as many as several
thousands) of I/O sessions or connections. It is generally recommended to have one dispatch
thread per CPU core.
</para>
<programlisting><![CDATA[
IOReactorConfig config = IOReactorConfig.DEFAULT;
IOReactor ioreactor = new DefaultConnectingIOReactor(config);
]]></programlisting>
<section>
<title>I/O dispatchers</title>
<para>
<interfacename>IOReactor</interfacename> implementations make use of the
<interfacename>IOEventDispatch</interfacename> interface to notify clients of events
pending for a particular session. All methods of the <interfacename>IOEventDispatch
</interfacename> are executed on a dispatch thread of the I/O reactor. Therefore, it is
important that processing that takes place in the event methods will not block the
dispatch thread for too long, as the I/O reactor will be unable to react to other
events.
</para>
<programlisting><![CDATA[
IOReactor ioreactor = new DefaultConnectingIOReactor();
IOEventDispatch eventDispatch = <...>
ioreactor.execute(eventDispatch);
]]></programlisting>
<para>
Generic I/O events as defined by the <interfacename>IOEventDispatch</interfacename>
interface:
</para>
<itemizedlist>
<listitem>
<formalpara>
<title><methodname>connected</methodname>:</title>
<para>
Triggered when a new session has been created.
</para>
</formalpara>
</listitem>
<listitem>
<formalpara>
<title><methodname>inputReady</methodname>:</title>
<para>
Triggered when the session has pending input.
</para>
</formalpara>
</listitem>
<listitem>
<formalpara>
<title><methodname>outputReady</methodname>:</title>
<para>
Triggered when the session is ready for output.
</para>
</formalpara>
</listitem>
<listitem>
<formalpara>
<title><methodname>timeout</methodname>:</title>
<para>
Triggered when the session has timed out.
</para>
</formalpara>
</listitem>
<listitem>
<formalpara>
<title><methodname>disconnected</methodname>:</title>
<para>
Triggered when the session has been terminated.
</para>
</formalpara>
</listitem>
</itemizedlist>
</section>
<section>
<title>I/O reactor shutdown</title>
<para>
The shutdown of I/O reactors is a complex process and may usually take a while to
complete. I/O reactors will attempt to gracefully terminate all active I/O sessions and
dispatch threads approximately within the specified grace period. If any of the I/O
sessions fails to terminate correctly, the I/O reactor will forcibly shut down
remaining sessions.
</para>
<programlisting><![CDATA[
IOReactor ioreactor = <...>
long gracePeriod = 3000L; // milliseconds
ioreactor.shutdown(gracePeriod);
]]></programlisting>
<para>
The <methodname>IOReactor#shutdown(long)</methodname> method is safe to call from any
thread.
</para>
</section>
<section>
<title>I/O sessions</title>
<para>
The <interfacename>IOSession</interfacename> interface represents a sequence of
logically related data exchanges between two end points. <interfacename>IOSession
</interfacename> encapsulates functionality of NIO <classname>
java.nio.channels.SelectionKey</classname> and <classname>
java.nio.channels.SocketChannel</classname>. The channel associated with the
<interfacename>IOSession</interfacename> can be used to read data from and write data
to the session.
</para>
<programlisting><![CDATA[
IOSession iosession = <...>
ReadableByteChannel ch = (ReadableByteChannel) iosession.channel();
ByteBuffer dst = ByteBuffer.allocate(2048);
ch.read(dst);
]]></programlisting>
</section>
<section>
<title>I/O session state management</title>
<para>
I/O sessions are not bound to an execution thread, therefore one cannot use the context
of the thread to store a session's state. All details about a particular session must
be stored within the session itself.
</para>
<programlisting><![CDATA[
IOSession iosession = <...>
Object someState = <...>
iosession.setAttribute("state", someState);
...
IOSession iosession = <...>
Object currentState = iosession.getAttribute("state");
]]></programlisting>
<para>
Please note that if several sessions make use of shared objects, access to those
objects must be made thread-safe.
</para>
</section>
<section>
<title>I/O session event mask</title>
<para>
One can declare an interest in a particular type of I/O events for a particular I/O
session by setting its event mask.
</para>
<programlisting><![CDATA[
IOSession iosession = <...>
iosession.setEventMask(SelectionKey.OP_READ | SelectionKey.OP_WRITE);
]]></programlisting>
<para>
One can also toggle <literal>OP_READ</literal> and <literal>OP_WRITE</literal> flags
individually.
</para>
<programlisting><![CDATA[
IOSession iosession = <...>
iosession.setEvent(SelectionKey.OP_READ);
iosession.clearEvent(SelectionKey.OP_READ);
]]></programlisting>
<para>
Event notifications will not take place if the corresponding interest flag is not set.
</para>
</section>
<section>
<title>I/O session buffers</title>
<para>
Quite often I/O sessions need to maintain internal I/O buffers in order to transform
input / output data prior to returning it to the consumer or writing it to the
underlying channel. Memory management in HttpCore NIO is based on the fundamental
principle that the data a consumer can read, is only as much input data as it can process
without having to allocate more memory. That means, quite often some input data may
remain unread in one of the internal or external session buffers. The I/O reactor can
query the status of these session buffers, and make sure the consumer gets notified
correctly as more data gets stored in one of the session buffers, thus allowing the
consumer to read the remaining data once it is able to process it. I/O sessions can be
made aware of the status of external session buffers using the <interfacename>
SessionBufferStatus</interfacename> interface.
</para>
<programlisting><![CDATA[
IOSession iosession = <...>
SessionBufferStatus myBufferStatus = <...>
iosession.setBufferStatus(myBufferStatus);
iosession.hasBufferedInput();
iosession.hasBufferedOutput();
]]></programlisting>
</section>
<section>
<title>I/O session shutdown</title>
<para>
One can close an I/O session gracefully by calling <methodname>IOSession#close()
</methodname> allowing the session to be closed in an orderly manner or by calling
<methodname>IOSession#shutdown()</methodname> to forcibly close the underlying channel.
The distinction between two methods is of primary importance for those types of I/O
sessions that involve some sort of a session termination handshake such as SSL/TLS
connections.
</para>
</section>
<section>
<title>Listening I/O reactors</title>
<para>
<interfacename>ListeningIOReactor</interfacename> represents an I/O reactor capable of
listening for incoming connections on one or several ports.
</para>
<programlisting><![CDATA[
ListeningIOReactor ioreactor = <...>
ListenerEndpoint ep1 = ioreactor.listen(new InetSocketAddress(8081) );
ListenerEndpoint ep2 = ioreactor.listen(new InetSocketAddress(8082));
ListenerEndpoint ep3 = ioreactor.listen(new InetSocketAddress(8083));
// Wait until all endpoints are up
ep1.waitFor();
ep2.waitFor();
ep3.waitFor();
]]></programlisting>
<para>
Once an endpoint is fully initialized it starts accepting incoming connections and
propagates I/O activity notifications to the <interfacename>IOEventDispatch
</interfacename> instance.
</para>
<para>
One can obtain a set of registered endpoints at runtime, query the status of an
endpoint at runtime, and close it if desired.
</para>
<programlisting><![CDATA[
ListeningIOReactor ioreactor = <...>
Set<ListenerEndpoint> eps = ioreactor.getEndpoints();
for (ListenerEndpoint ep: eps) {
// Still active?
System.out.println(ep.getAddress());
if (ep.isClosed()) {
// If not, has it terminated due to an exception?
if (ep.getException() != null) {
ep.getException().printStackTrace();
}
} else {
ep.close();
}
}
]]></programlisting>
</section>
<section>
<title>Connecting I/O reactors</title>
<para>
<interfacename>ConnectingIOReactor</interfacename> represents an I/O reactor capable of
establishing connections with remote hosts.
</para>
<programlisting><![CDATA[
ConnectingIOReactor ioreactor = <...>
SessionRequest sessionRequest = ioreactor.connect(
new InetSocketAddress("www.google.com", 80),
null, null, null);
]]></programlisting>
<para>
Opening a connection to a remote host usually tends to be a time consuming process and
may take a while to complete. One can monitor and control the process of session
initialization by means of the <interfacename>SessionRequest</interfacename>interface.
</para>
<programlisting><![CDATA[
// Make sure the request times out if connection
// has not been established after 1 sec
sessionRequest.setConnectTimeout(1000);
// Wait for the request to complete
sessionRequest.waitFor();
// Has request terminated due to an exception?
if (sessionRequest.getException() != null) {
sessionRequest.getException().printStackTrace();
}
// Get hold of the new I/O session
IOSession iosession = sessionRequest.getSession();
]]></programlisting>
<para>
<interfacename>SessionRequest</interfacename> implementations are expected to be
thread-safe. Session request can be aborted at any time by calling <methodname>
IOSession#cancel()</methodname> from another thread of execution.
</para>
<programlisting><![CDATA[
if (!sessionRequest.isCompleted()) {
sessionRequest.cancel();
}
]]></programlisting>
<para>
One can pass several optional parameters to the <methodname>
ConnectingIOReactor#connect()</methodname> method to exert a greater control over the
process of session initialization.
</para>
<para>
A non-null local socket address parameter can be used to bind the socket to a specific
local address.
</para>
<programlisting><![CDATA[
ConnectingIOReactor ioreactor = <...>
SessionRequest sessionRequest = ioreactor.connect(
new InetSocketAddress("www.google.com", 80),
new InetSocketAddress("192.168.0.10", 1234),
null, null);
]]></programlisting>
<para>
One can provide an attachment object, which will be added to the new session's context
upon initialization. This object can be used to pass an initial processing state to
the protocol handler.
</para>
<programlisting><![CDATA[
SessionRequest sessionRequest = ioreactor.connect(
new InetSocketAddress("www.google.com", 80),
null, new HttpHost("www.google.ru"), null);
IOSession iosession = sessionRequest.getSession();
HttpHost virtualHost = (HttpHost) iosession.getAttribute(
IOSession.ATTACHMENT_KEY);
]]></programlisting>
<para>
It is often desirable to be able to react to the completion of a session request
asynchronously without having to wait for it, blocking the current thread of execution.
One can optionally provide an implementation <interfacename>SessionRequestCallback
</interfacename> interface to get notified of events related to session requests, such
as request completion, cancellation, failure or timeout.
</para>
<programlisting><![CDATA[
ConnectingIOReactor ioreactor = <...>
SessionRequest sessionRequest = ioreactor.connect(
new InetSocketAddress("www.google.com", 80), null, null,
new SessionRequestCallback() {
public void cancelled(SessionRequest request) {
}
public void completed(SessionRequest request) {
System.out.println("new connection to " +
request.getRemoteAddress());
}
public void failed(SessionRequest request) {
if (request.getException() != null) {
request.getException().printStackTrace();
}
}
public void timeout(SessionRequest request) {
}
});
]]></programlisting>
</section>
</section>
<section>
<title>I/O reactor configuration</title>
<para>
I/O reactors by default use system dependent configuration which in most cases should be
sensible enough.
</para>
<programlisting><![CDATA[
IOReactorConfig config = IOReactorConfig.DEFAULT;
IOReactor ioreactor = new DefaultListeningIOReactor(config);
]]></programlisting>
<para>
However in some cases custom settings may be necessary, for instance, in order to alter
default socket properties and timeout values. One should rarely need to change other
parameters.
</para>
<programlisting><![CDATA[
IOReactorConfig config = IOReactorConfig.custom()
.setTcpNoDelay(true)
.setSoTimeout(5000)
.setSoReuseAddress(true)
.setConnectTimeout(5000)
.build();
IOReactor ioreactor = new DefaultListeningIOReactor(config);
]]></programlisting>
<section>
<title>Queuing of I/O interest set operations</title>
<para>
Several older JRE implementations (primarily from IBM) include what Java API
documentation refers to as a naive implementation of the <classname>
java.nio.channels.SelectionKey</classname> class. The problem with <classname>
java.nio.channels.SelectionKey</classname> in such JREs is that reading or writing
of the I/O interest set may block indefinitely if the I/O selector is in the process
of executing a select operation. HttpCore NIO can be configured to operate in a special
mode wherein I/O interest set operations are queued and executed by on the dispatch
thread only when the I/O selector is not engaged in a select operation.
</para>
<programlisting><![CDATA[
IOReactorConfig config = IOReactorConfig.custom()
.setInterestOpQueued(true)
.build();
]]></programlisting>
</section>
</section>
<section>
<title>I/O reactor exception handling</title>
<para>
Protocol specific exceptions as well as those I/O exceptions thrown in the course of
interaction with the session's channel are to be expected and are to be dealt with by specific
protocol handlers. These exceptions may result in termination of an individual session but
should not affect the I/O reactor and all other active sessions. There are situations,
however, when the I/O reactor itself encounters an internal problem such as an I/O
exception in the underlying NIO classes or an unhandled runtime exception. Those types of
exceptions are usually fatal and will cause the I/O reactor to shut down automatically.
</para>
<para>
There is a possibility to override this behavior and prevent I/O reactors from shutting
down automatically in case of a runtime exception or an I/O exception in internal classes.
This can be accomplished by providing a custom implementation of the <interfacename>
IOReactorExceptionHandler</interfacename> interface.
</para>
<programlisting><![CDATA[
DefaultConnectingIOReactor ioreactor = <...>
ioreactor.setExceptionHandler(new IOReactorExceptionHandler() {
public boolean handle(IOException ex) {
if (ex instanceof BindException) {
// bind failures considered OK to ignore
return true;
}
return false;
}
public boolean handle(RuntimeException ex) {
if (ex instanceof UnsupportedOperationException) {
// Unsupported operations considered OK to ignore
return true;
}
return false;
}
});
]]></programlisting>
<para>
One needs to be very careful about discarding exceptions indiscriminately. It is often much
better to let the I/O reactor shut down itself cleanly and restart it rather than leaving
it in an inconsistent or unstable state.
</para>
<section>
<title>I/O reactor audit log</title>
<para>
If an I/O reactor is unable to automatically recover from an I/O or a runtime exception
it will enter the shutdown mode. First off, it will close all active listeners and
cancel all pending new session requests. Then it will attempt to close all active I/O
sessions gracefully giving them some time to flush pending output data and terminate
cleanly. Lastly, it will forcibly shut down those I/O sessions that still remain active
after the grace period. This is a fairly complex process, where many things can fail at
the same time and many different exceptions can be thrown in the course of the shutdown
process. The I/O reactor will record all exceptions thrown during the shutdown process,
including the original one that actually caused the shutdown in the first place, in an
audit log. One can examine the audit log and decide whether it is safe to restart the
I/O reactor.
</para>
<programlisting><![CDATA[
DefaultConnectingIOReactor ioreactor = <...>
// Give it 5 sec grace period
ioreactor.shutdown(5000);
List<ExceptionEvent> events = ioreactor.getAuditLog();
for (ExceptionEvent event: events) {
System.err.println("Time: " + event.getTimestamp());
event.getCause().printStackTrace();
}
]]></programlisting>
</section>
</section>
<section>
<title>Non-blocking HTTP connections</title>
<para>
Effectively non-blocking HTTP connections are wrappers around <interfacename>IOSession
</interfacename> with HTTP specific functionality. Non-blocking HTTP connections are
stateful and not thread-safe. Input / output operations on non-blocking HTTP connections
should be restricted to the dispatch events triggered by the I/O event dispatch thread.
</para>
<section>
<title>Execution context of non-blocking HTTP connections</title>
<para>
Non-blocking HTTP connections are not bound to a particular thread of execution and
therefore they need to maintain their own execution context. Each non-blocking HTTP
connection has an <interfacename>HttpContext</interfacename> instance associated with
it, which can be used to maintain a processing state. The <interfacename>HttpContext
</interfacename> instance is thread-safe and can be manipulated from multiple threads.
</para>
<programlisting><![CDATA[
DefaultNHttpClientConnection conn = <...>
Object myStateObject = <...>
HttpContext context = conn.getContext();
context.setAttribute("state", myStateObject);
]]></programlisting>
</section>
<section>
<title>Working with non-blocking HTTP connections</title>
<para>
At any point of time one can obtain the request and response objects currently being
transferred over the non-blocking HTTP connection. Any of these objects, or both, can
be null if there is no incoming or outgoing message currently being transferred.
</para>
<programlisting><![CDATA[
NHttpConnection conn = <...>
HttpRequest request = conn.getHttpRequest();
if (request != null) {
System.out.println("Transferring request: " +
request.getRequestLine());
}
HttpResponse response = conn.getHttpResponse();
if (response != null) {
System.out.println("Transferring response: " +
response.getStatusLine());
}
]]></programlisting>
<para>
However, please note that the current request and the current response may not
necessarily represent the same message exchange! Non-blocking HTTP connections can
operate in a full duplex mode. One can process incoming and outgoing messages
completely independently from one another. This makes non-blocking HTTP connections
fully pipelining capable, but at same time implies that this is the job of the protocol
handler to match logically related request and the response messages.
</para>
<para>
Over-simplified process of submitting a request on the client side may look like this:
</para>
<programlisting><![CDATA[
NHttpClientConnection conn = <...>
// Obtain execution context
HttpContext context = conn.getContext();
// Obtain processing state
Object state = context.getAttribute("state");
// Generate a request based on the state information
HttpRequest request = new BasicHttpRequest("GET", "/");
conn.submitRequest(request);
System.out.println(conn.isRequestSubmitted());
]]></programlisting>
<para>
Over-simplified process of submitting a response on the server side may look like this:
</para>
<programlisting><![CDATA[
NHttpServerConnection conn = <...>
// Obtain execution context
HttpContext context = conn.getContext();
// Obtain processing state
Object state = context.getAttribute("state");
// Generate a response based on the state information
HttpResponse response = new BasicHttpResponse(HttpVersion.HTTP_1_1,
HttpStatus.SC_OK, "OK");
BasicHttpEntity entity = new BasicHttpEntity();
entity.setContentType("text/plain");
entity.setChunked(true);
response.setEntity(entity);
conn.submitResponse(response);
System.out.println(conn.isResponseSubmitted());
]]></programlisting>
<para>
Please note that one should rarely need to transmit messages using these low level
methods and should use appropriate higher level HTTP service implementations instead.
</para>
</section>
<section>
<title>HTTP I/O control</title>
<para>
All non-blocking HTTP connections classes implement <interfacename>IOControl
</interfacename> interface, which represents a subset of connection functionality for
controlling interest in I/O even notifications. <interfacename>IOControl
</interfacename> instances are expected to be fully thread-safe. Therefore
<interfacename>IOControl</interfacename> can be used to request / suspend I/O event
notifications from any thread.
</para>
<para>
One must take special precautions when interacting with non-blocking connections.
<interfacename>HttpRequest</interfacename> and <interfacename>HttpResponse
</interfacename>are not thread-safe. It is generally advisable that all input / output
operations on a non-blocking connection are executed from the I/O event dispatch
thread.
</para>
<para>
The following pattern is recommended:
</para>
<itemizedlist>
<listitem>
<para>
Use <interfacename>IOControl</interfacename> interface to pass control over
connection's I/O events to another thread / session.
</para>
</listitem>
<listitem>
<para>
If input / output operations need be executed on that particular connection,
store all the required information (state) in the connection context and
request the appropriate I/O operation by calling <methodname>
IOControl#requestInput()</methodname> or <methodname>IOControl#requestOutput()
</methodname> method.
</para>
</listitem>
<listitem>
<para>
Execute the required operations from the event method on the dispatch thread
using information stored in connection context.
</para>
</listitem>
</itemizedlist>
<para>
Please note all operations that take place in the event methods should not block for
too long, because while the dispatch thread remains blocked in one session, it is
unable to process events for all other sessions. I/O operations with the underlying
channel of the session are not a problem as they are guaranteed to be non-blocking.
</para>
</section>
<section>
<title>Non-blocking content transfer</title>
<para>
The process of content transfer for non-blocking connections works completely
differently compared to that of blocking connections, as non-blocking connections need
to accommodate to the asynchronous nature of the NIO model. The main distinction
between two types of connections is inability to use the usual, but inherently blocking
<classname>java.io.InputStream</classname> and <classname>java.io.OutputStream
</classname> classes to represent streams of inbound and outbound content. HttpCore NIO
provides <interfacename>ContentEncoder</interfacename> and <interfacename>
ContentDecoder</interfacename> interfaces to handle the process of asynchronous content
transfer. Non-blocking HTTP connections will instantiate the appropriate implementation
of a content codec based on properties of the entity enclosed with the message.
</para>
<para>
Non-blocking HTTP connections will fire input events until the content entity is fully
transferred.
</para>
<programlisting><![CDATA[
ContentDecoder decoder = <...>
//Read data in
ByteBuffer dst = ByteBuffer.allocate(2048);
decoder.read(dst);
// Decode will be marked as complete when
// the content entity is fully transferred
if (decoder.isCompleted()) {
// Done
}
]]></programlisting>
<para>
Non-blocking HTTP connections will fire output events until the content entity is
marked as fully transferred.
</para>
<programlisting><![CDATA[
ContentEncoder encoder = <...>
// Prepare output data
ByteBuffer src = ByteBuffer.allocate(2048);
// Write data out
encoder.write(src);
// Mark content entity as fully transferred when done
encoder.complete();
]]></programlisting>
<para>
Please note, one still has to provide an HttpEntity instance when submitting an entity
enclosing message to the non-blocking HTTP connection. Properties of that entity will
be used to initialize an <interfacename>ContentEncoder</interfacename> instance to be
used for transferring entity content. Non-blocking HTTP connections, however, ignore
inherently blocking <methodname>HttpEntity#getContent()</methodname> and <methodname>
HttpEntity#writeTo()</methodname> methods of the enclosed entities.
</para>
<programlisting><![CDATA[
NHttpServerConnection conn = <...>
HttpResponse response = new BasicHttpResponse(HttpVersion.HTTP_1_1,
HttpStatus.SC_OK, "OK");
BasicHttpEntity entity = new BasicHttpEntity();
entity.setContentType("text/plain");
entity.setChunked(true);
entity.setContent(null);
response.setEntity(entity);
conn.submitResponse(response);
]]></programlisting>
<para>
Likewise, incoming entity enclosing message will have an <interfacename>HttpEntity
</interfacename> instance associated with them, but an attempt to call <methodname>
HttpEntity#getContent()</methodname> or <methodname>HttpEntity#writeTo()</methodname>
methods will cause an <classname>java.lang.IllegalStateException</classname>. The
<interfacename>HttpEntity</interfacename> instance can be used to determine properties
of the incoming entity such as content length.
</para>
<programlisting><![CDATA[
NHttpClientConnection conn = <...>
HttpResponse response = conn.getHttpResponse();
HttpEntity entity = response.getEntity();
if (entity != null) {
System.out.println(entity.getContentType());
System.out.println(entity.getContentLength());
System.out.println(entity.isChunked());
}
]]></programlisting>
</section>
<section>
<title>Supported non-blocking content transfer mechanisms</title>
<para>
Default implementations of the non-blocking HTTP connection interfaces support three
content transfer mechanisms defined by the HTTP/1.1 specification:
</para>
<itemizedlist>
<listitem>
<formalpara>
<title><literal>Content-Length</literal> delimited:</title>
<para>
The end of the content entity is determined by the value of the
<literal>Content-Length</literal> header. Maximum entity length:
<methodname>Long#MAX_VALUE</methodname>.
</para>
</formalpara>
</listitem>
<listitem>
<formalpara>
<title>Identity coding:</title>
<para>
The end of the content entity is demarcated by closing the underlying
connection (end of stream condition). For obvious reasons the identity encoding
can only be used on the server side. Max entity length: unlimited.
</para>
</formalpara>
</listitem>
<listitem>
<formalpara>
<title>Chunk coding:</title>
<para>
The content is sent in small chunks. Max entity length: unlimited.
</para>
</formalpara>
</listitem>
</itemizedlist>
<para>
The appropriate content codec will be created automatically depending on properties of
the entity enclosed with the message.
</para>
</section>
<section>
<title>Direct channel I/O</title>
<para>
Content codes are optimized to read data directly from or write data directly to the
underlying I/O session's channel, whenever possible avoiding intermediate buffering in
a session buffer. Moreover, those codecs that do not perform any content transformation
(<literal>Content-Length</literal> delimited and identity codecs, for example) can
leverage NIO <classname>java.nio.FileChannel</classname> methods for significantly
improved performance of file transfer operations both inbound and outbound.
</para>
<para>
If the actual content decoder implements <interfacename>FileContentDecoder
</interfacename> one can make use of its methods to read incoming content directly to a
file bypassing an intermediate <classname>java.nio.ByteBuffer</classname>.
</para>
<programlisting><![CDATA[
ContentDecoder decoder = <...>
//Prepare file channel
FileChannel dst;
//Make use of direct file I/O if possible
if (decoder instanceof FileContentDecoder) {
long Bytesread = ((FileContentDecoder) decoder)
.transfer(dst, 0, 2048);
// Decode will be marked as complete when
// the content entity is fully transmitted
if (decoder.isCompleted()) {
// Done
}
}
]]></programlisting>
<para>
If the actual content encoder implements <interfacename>FileContentEncoder
</interfacename> one can make use of its methods to write outgoing content directly
from a file bypassing an intermediate <classname>java.nio.ByteBuffer</classname>.
</para>
<programlisting><![CDATA[
ContentEncoder encoder = <...>
// Prepare file channel
FileChannel src;
// Make use of direct file I/O if possible
if (encoder instanceof FileContentEncoder) {
// Write data out
long bytesWritten = ((FileContentEncoder) encoder)
.transfer(src, 0, 2048);
// Mark content entity as fully transferred when done
encoder.complete();
}
]]></programlisting>
</section>
</section>
<section>
<title>HTTP I/O event dispatchers</title>
<para>
HTTP I/O event dispatchers serve to convert generic I/O events triggered by an I/O reactor
to HTTP protocol specific events. They rely on <interfacename>NHttpClientEventHandler
</interfacename> and <interfacename>NHttpServerEventHandler</interfacename> interfaces to
propagate HTTP protocol events to a HTTP protocol handler.
</para>
<para>
Server side HTTP I/O events as defined by the <interfacename>NHttpServerEventHandler
</interfacename> interface:
</para>
<itemizedlist>
<listitem>
<formalpara>
<title><methodname>connected</methodname>:</title>
<para>
Triggered when a new incoming connection has been created.
</para>
</formalpara>
</listitem>
<listitem>
<formalpara>
<title><methodname>requestReceived</methodname>:</title>
<para>
Triggered when a new HTTP request is received. The connection passed as a parameter
to this method is guaranteed to return a valid HTTP request object. If the request
received encloses a request entity this method will be followed a series of
<methodname>inputReady</methodname> events to transfer the request content.
</para>
</formalpara>
</listitem>
<listitem>
<formalpara>
<title><methodname>inputReady</methodname>:</title>
<para>
Triggered when the underlying channel is ready for reading a new portion of
the request entity through the corresponding content decoder. If the content
consumer is unable to process the incoming content, input event notifications can
temporarily suspended using <interfacename>IOControl</interfacename> interface
(super interface of <interfacename>NHttpServerConnection</interfacename>).
</para>
<para>
Please note that the <interfacename>NHttpServerConnection</interfacename> and
<interfacename>ContentDecoder</interfacename> objects are not thread-safe and
should only be used within the context of this method call. The <interfacename>
IOControl</interfacename> object can be shared and used on other thread to resume
input event notifications when the handler is capable of processing more content.
</para>
</formalpara>
</listitem>
<listitem>
<formalpara>
<title><methodname>responseReady</methodname>:</title>
<para>
Triggered when the connection is ready to accept new HTTP response. The protocol
handler does not have to submit a response if it is not ready.
</para>
</formalpara>
</listitem>
<listitem>
<formalpara>
<title><methodname>outputReady</methodname>:</title>
<para>
Triggered when the underlying channel is ready for writing a next portion of the
response entity through the corresponding content encoder. If the content producer
is unable to generate the outgoing content, output event notifications can be
temporarily suspended using <interfacename>IOControl</interfacename> interface
(super interface of <interfacename>NHttpServerConnection</interfacename>).
</para>
<para>
Please note that the <interfacename>NHttpServerConnection</interfacename> and
<interfacename>ContentEncoder</interfacename> objects are not thread-safe and
should only be used within the context of this method call. The <interfacename>
IOControl</interfacename> object can be shared and used on other thread to resume
output event notifications when more content is made available.
</para>
</formalpara>
</listitem>
<listitem>
<formalpara>
<title><methodname>exception</methodname>:</title>
<para>
Triggered when an I/O error occurrs while reading from or writing to the underlying
channel or when an HTTP protocol violation occurs while receiving an HTTP request.
</para>
</formalpara>
</listitem>
<listitem>
<formalpara>
<title><methodname>timeout</methodname>:</title>
<para>
Triggered when no input is detected on this connection over the maximum period of
inactivity.
</para>
</formalpara>
</listitem>
<listitem>
<formalpara>
<title><methodname>closed</methodname>:</title>
<para>
Triggered when the connection has been closed.
</para>
</formalpara>
</listitem>
</itemizedlist>
<para>
Client side HTTP I/O events as defined by the <interfacename>NHttpClientEventHandler
</interfacename> interface:
</para>
<itemizedlist>
<listitem>
<formalpara>
<title><methodname>connected</methodname>:</title>
<para>
Triggered when a new outgoing connection has been created. The attachment object
passed as a parameter to this event is an arbitrary object that was attached to
the session request.
</para>
</formalpara>
</listitem>
<listitem>
<formalpara>
<title><methodname>requestReady</methodname>:</title>
<para>
Triggered when the connection is ready to accept new HTTP request. The protocol
handler does not have to submit a request if it is not ready.
</para>
</formalpara>
</listitem>
<listitem>
<formalpara>
<title><methodname>outputReady</methodname>:</title>
<para>
Triggered when the underlying channel is ready for writing a next portion of the
request entity through the corresponding content encoder. If the content producer
is unable to generate the outgoing content, output event notifications can be
temporarily suspended using <interfacename>IOControl</interfacename> interface
(super interface of <interfacename>NHttpClientConnection</interfacename>).
</para>
<para>
Please note that the <interfacename>NHttpClientConnection</interfacename> and
<interfacename>ContentEncoder</interfacename> objects are not thread-safe and
should only be used within the context of this method call. The <interfacename>
IOControl</interfacename> object can be shared and used on other thread to resume
output event notifications when more content is made available.
</para>
</formalpara>
</listitem>
<listitem>
<formalpara>
<title><methodname>responseReceived</methodname>:</title>
<para>
Triggered when an HTTP response is received. The connection passed as a parameter to
this method is guaranteed to return a valid HTTP response object. If the response
received encloses a response entity this method will be followed a series of
<methodname>inputReady</methodname> events to transfer the response content.
</para>
</formalpara>
</listitem>
<listitem>
<formalpara>
<title><methodname>inputReady</methodname>:</title>
<para>
Triggered when the underlying channel is ready for reading a new portion of the
response entity through the corresponding content decoder. If the content consumer
is unable to process the incoming content, input event notifications can be
temporarily suspended using <interfacename>IOControl</interfacename> interface
(super interface of <interfacename>NHttpClientConnection</interfacename>).
</para>
<para>
Please note that the <interfacename>NHttpClientConnection</interfacename> and
<interfacename>ContentDecoder</interfacename> objects are not thread-safe and
should only be used within the context of this method call. The <interfacename>
IOControl</interfacename> object can be shared and used on other thread to resume
input event notifications when the handler is capable of processing more content.
</para>
</formalpara>
</listitem>
<listitem>
<formalpara>
<title><methodname>exception</methodname>:</title>
<para>
Triggered when an I/O error occurs while reading from or writing to the underlying
channel or when an HTTP protocol violation occurs while receiving an HTTP response.
</para>
</formalpara>
</listitem>
<listitem>
<formalpara>
<title><methodname>timeout</methodname>:</title>
<para>
Triggered when no input is detected on this connection over the maximum period of
inactivity.
</para>
</formalpara>
</listitem>
<listitem>
<formalpara>
<title><methodname>closed</methodname>:</title>
<para>
Triggered when the connection has been closed.
</para>
</formalpara>
</listitem>
</itemizedlist>
</section>
<section>
<title>Non-blocking HTTP content producers</title>
<para>
As discussed previously the process of content transfer for non-blocking connections works
completely differently compared to that for blocking connections. For obvious reasons
classic I/O abstraction based on inherently blocking <classname>java.io.InputStream
</classname> and <classname>java.io.OutputStream</classname> classes is not well suited
for asynchronous data transfer. In order to avoid inefficient and potentially blocking
I/O operation redirection through <methodname>java.nio.channels.Channles#newChannel
</methodname> non-blocking HTTP entities are expected to implement NIO specific extension
interface <interfacename>HttpAsyncContentProducer</interfacename>.
</para>
<para>
The <interfacename>HttpAsyncContentProducer</interfacename> interface defines several
additional method for efficient streaming of content to a non-blocking HTTP connection:
</para>
<itemizedlist>
<listitem>
<formalpara>
<title><methodname>produceContent</methodname>:</title>
<para>
Invoked to write out a chunk of content to the <interfacename>ContentEncoder
</interfacename>. The <interfacename>IOControl</interfacename> interface can be
used to suspend output events if the entity is temporarily unable to produce more
content. When all content is finished, the producer MUST call
<methodname>ContentEncoder#complete()</methodname>. Failure to do so may cause
the entity to be incorrectly delimited.
</para>
<para>
Please note that the <interfacename>ContentEncoder</interfacename> object is
not thread-safe and should only be used within the context of this method call.
The <interfacename>IOControl</interfacename> object can be shared and used on other
thread resume output event notifications when more content is made available.
</para>
</formalpara>
</listitem>
<listitem>
<formalpara>
<title><methodname>isRepeatable</methodname>:</title>
<para>
Determines whether or not this producer is capable of producing its content more
than once. Repeatable content producers are expected to be able to recreate
their content even after having been closed.
</para>
</formalpara>
</listitem>
<listitem>
<formalpara>
<title><methodname>close</methodname>:</title>
<para>
Closes the producer and releases all resources currently allocated by it.
</para>
</formalpara>
</listitem>
</itemizedlist>
<section>
<title>Creating non-blocking entities</title>
<para>
Several HTTP entity implementations included in HttpCore NIO support
<interfacename>HttpAsyncContentProducer</interfacename> interface:
</para>
<itemizedlist>
<listitem>
<para>
<link linkend="bytearray-n-entity">
<classname>NByteArrayEntity</classname>
</link>
</para>
</listitem>
<listitem>
<para>
<link linkend="string-n-entity">
<classname>NStringEntity</classname>
</link>
</para>
</listitem>
<listitem>
<para>
<link linkend="file-n-entity">
<classname>NFileEntity</classname>
</link>
</para>
</listitem>
</itemizedlist>
<section id="bytearray-n-entity">
<title><classname>NByteArrayEntity</classname></title>
<para>
This is a simple self-contained repeatable entity, which receives its content from
a given byte array. This byte array is supplied to the constructor.
</para>
<programlisting><![CDATA[
NByteArrayEntity entity = new NByteArrayEntity(new byte[] {1, 2, 3});
]]></programlisting>
</section>
<section id="string-n-entity">
<title><classname>NStringEntity</classname></title>
<para>
This is a simple, self-contained, repeatable entity that retrieves its data from a
<classname>java.lang.String</classname> object. It has 2 constructors, one simply
constructs with a given string where the other also takes a character encoding for
the data in the <classname>java.lang.String</classname>.
</para>
<programlisting><![CDATA[
NStringEntity myEntity = new NStringEntity("important message",
Consts.UTF_8);
]]></programlisting>
</section>
<section id="file-n-entity">
<title><classname>NFileEntity</classname></title>
<para>
This entity reads its content body from a file. This class is mostly used to stream
large files of different types, so one needs to supply the content type of the file
to make sure the content can be correctly recognized and processed by the
recipient.
</para>
<programlisting><![CDATA[
File staticFile = new File("/path/to/myapp.jar");
NFileEntity entity = new NFileEntity(staticFile,
ContentType.create("application/java-archive", null));
]]></programlisting>
<para>
The <classname>NHttpEntity</classname> will make use of the direct channel I/O
whenever possible, provided the content encoder is capable of transferring data
directly from a file to the socket of the underlying connection.
</para>
</section>
</section>
</section>
<section>
<title>Non-blocking HTTP protocol handlers</title>
<section>
<title>Asynchronous HTTP service</title>
<para>
<classname>HttpAsyncService</classname> is a fully asynchronous HTTP server side
protocol handler based on the non-blocking (NIO) I/O model. <classname>
HttpAsyncService</classname> translates individual events fired through the
<interfacename>NHttpServerEventHandler</interfacename> interface into logically
related HTTP message exchanges.
</para>
<para>
Upon receiving an incoming request the <classname>HttpAsyncService</classname>
verifies the message for compliance with the server expectations using <interfacename>
HttpAsyncExpectationVerifier</interfacename>, if provided, and then <interfacename>
HttpAsyncRequestHandlerResolver</interfacename> is used to resolve the request URI to
a particular <interfacename>HttpAsyncRequestHandler</interfacename> intended to handle
the request with the given URI. The protocol handler uses the selected <interfacename>
HttpAsyncRequestHandler</interfacename> instance to process the incoming request and
to generate an outgoing response.
</para>
<para>
<classname>HttpAsyncService</classname> relies on <interfacename>HttpProcessor
</interfacename> to generate mandatory protocol headers for all outgoing messages
and apply common, cross-cutting message transformations to all incoming and outgoing
messages, whereas individual HTTP request handlers are expected to implement
application specific content generation and processing.
</para>
<programlisting><![CDATA[
HttpProcessor httpproc = HttpProcessorBuilder.create()
.add(new ResponseDate())
.add(new ResponseServer("MyServer-HTTP/1.1"))
.add(new ResponseContent())
.add(new ResponseConnControl())
.build();
HttpAsyncService protocolHandler = new HttpAsyncService(httpproc, null);
IOEventDispatch ioEventDispatch = new DefaultHttpServerIODispatch(
protocolHandler,
new DefaultNHttpServerConnectionFactory(ConnectionConfig.DEFAULT));
ListeningIOReactor ioreactor = new DefaultListeningIOReactor();
ioreactor.execute(ioEventDispatch);
]]></programlisting>
<section>
<title>Non-blocking HTTP request handlers</title>
<para>
<interfacename>HttpAsyncRequestHandler</interfacename> represents a routine for
asynchronous processing of a specific group of non-blocking HTTP requests.
Protocol handlers are designed to take care of protocol specific aspects, whereas
individual request handlers are expected to take care of application specific HTTP
processing. The main purpose of a request handler is to generate a response object
with a content entity to be sent back to the client in response to the given
request.
</para>
<programlisting><![CDATA[
HttpAsyncRequestHandler<HttpRequest> rh = new HttpAsyncRequestHandler<HttpRequest>() {
public HttpAsyncRequestConsumer<HttpRequest> processRequest(
final HttpRequest request,
final HttpContext context) {
// Buffer request content in memory for simplicity
return new BasicAsyncRequestConsumer();
}
public void handle(
final HttpRequest request,
final HttpAsyncExchange httpexchange,
final HttpContext context) throws HttpException, IOException {
HttpResponse response = httpexchange.getResponse();
response.setStatusCode(HttpStatus.SC_OK);
NFileEntity body = new NFileEntity(new File("static.html"),
ContentType.create("text/html", Consts.UTF_8));
response.setEntity(body);
httpexchange.submitResponse(new BasicAsyncResponseProducer(response));
}
};
]]></programlisting>
<para>
Request handlers must be implemented in a thread-safe manner. Similarly to
servlets, request handlers should not use instance variables unless access to those
variables are synchronized.
</para>
</section>
<section>
<title>Asynchronous HTTP exchange</title>
<para>
The most fundamental difference of the non-blocking request handlers compared to
their blocking counterparts is ability to defer transmission of the HTTP response
back to the client without blocking the I/O thread by delegating the process of
handling the HTTP request to a worker thread or another service. The instance of
<interfacename>HttpAsyncExchange</interfacename> passed as a parameter to the
<methodname>HttpAsyncRequestHandler#handle</methodname> method to submit
a response as at a later point once response content becomes available.
</para>
<para>
The <interfacename>HttpAsyncExchange</interfacename> interface can be interacted
with using the following methods:
</para>
<itemizedlist>
<listitem>
<formalpara>
<title><methodname>getRequest</methodname>:</title>
<para>
Returns the received HTTP request message.
</para>
</formalpara>
</listitem>
<listitem>
<formalpara>
<title><methodname>getResponse</methodname>:</title>
<para>
Returns the default HTTP response message that can submitted once ready.
</para>
</formalpara>
</listitem>
<listitem>
<formalpara>
<title><methodname>submitResponse</methodname>:</title>
<para>
Submits an HTTP response and completed the message exchange.
</para>
</formalpara>
</listitem>
<listitem>
<formalpara>
<title><methodname>isCompleted</methodname>:</title>
<para>
Determines whether or not the message exchange has been completed.
</para>
</formalpara>
</listitem>
<listitem>
<formalpara>
<title><methodname>setCallback</methodname>:</title>
<para>
Sets <interfacename>Cancellable</interfacename> callback to be invoked
in case the underlying connection times out or gets terminated prematurely
by the client. This callback can be used to cancel a long running response
generating process if a response is no longer needed.
</para>
</formalpara>
</listitem>
<listitem>
<formalpara>
<title><methodname>setTimeout</methodname>:</title>
<para>
Sets timeout for this message exchange.
</para>
</formalpara>
</listitem>
<listitem>
<formalpara>
<title><methodname>getTimeout</methodname>:</title>
<para>
Returns timeout for this message exchange.
</para>
</formalpara>
</listitem>
</itemizedlist>
<programlisting><![CDATA[
HttpAsyncRequestHandler<HttpRequest> rh = new HttpAsyncRequestHandler<HttpRequest>() {
public HttpAsyncRequestConsumer<HttpRequest> processRequest(
final HttpRequest request,
final HttpContext context) {
// Buffer request content in memory for simplicity
return new BasicAsyncRequestConsumer();
}
public void handle(
final HttpRequest request,
final HttpAsyncExchange httpexchange,
final HttpContext context) throws HttpException, IOException {
new Thread() {
@Override
public void run() {
try {
Thread.sleep(10);
}
catch(InterruptedException ie) {}
HttpResponse response = httpexchange.getResponse();
response.setStatusCode(HttpStatus.SC_OK);
NFileEntity body = new NFileEntity(new File("static.html"),
ContentType.create("text/html", Consts.UTF_8));
response.setEntity(body);
httpexchange.submitResponse(new BasicAsyncResponseProducer(response));
}
}.start();
}
};
]]></programlisting>
<para>
Please note <interfacename>HttpResponse</interfacename> instances are not
thread-safe and may not be modified concurrently. Non-blocking request handlers
must ensure HTTP response cannot be accessed by more than one thread at a time.
</para>
</section>
<section>
<title>Asynchronous HTTP request consumer</title>
<para>
<interfacename>HttpAsyncRequestConsumer</interfacename> facilitates the process of
asynchronous processing of HTTP requests. It is a callback interface used by
<interfacename>HttpAsyncRequestHandler</interfacename>s to process an incoming
HTTP request message and to stream its content from a non-blocking server side
HTTP connection.
</para>
<para>
HTTP I/O events and methods as defined by the <interfacename>
HttpAsyncRequestConsumer</interfacename> interface:
</para>
<itemizedlist>
<listitem>
<formalpara>
<title><methodname>requestReceived</methodname>:</title>
<para>
Invoked when a HTTP request message is received.
</para>
</formalpara>
</listitem>
<listitem>
<formalpara>
<title><methodname>consumeContent</methodname>:</title>
<para>
Invoked to process a chunk of content from the <interfacename>
ContentDecoder</interfacename>. The <interfacename>IOControl
</interfacename> interface can be used to suspend input events if
the consumer is temporarily unable to consume more content.
</para>
<para>
The consumer can use the <methodname>ContentDecoder#isCompleted()
</methodname> method to find out whether or not the message content
has been fully consumed.
</para>
<para>
Please note that the <interfacename>ContentDecoder</interfacename> object
is not thread-safe and should only be used within the context of this
method call. The <interfacename>IOControl</interfacename> object can be
shared and used on other thread to resume input event notifications
when the consumer is capable of processing more content.
</para>
<para>
This event is invoked only if the incoming request message has
a content entity enclosed in it.
</para>
</formalpara>
</listitem>
<listitem>
<formalpara>
<title><methodname>requestCompleted</methodname>:</title>
<para>
Invoked to signal that the request has been fully processed.
</para>
</formalpara>
</listitem>
<listitem>
<formalpara>
<title><methodname>failed</methodname>:</title>
<para>
Invoked to signal that the request processing terminated abnormally.
</para>
</formalpara>
</listitem>
<listitem>
<formalpara>
<title><methodname>getException</methodname>:</title>
<para>
Returns an exception in case of an abnormal termination. This method
returns <code>null</code> if the request execution is still ongoing or if
it completed successfully.
</para>
</formalpara>
</listitem>
<listitem>
<formalpara>
<title><methodname>getResult</methodname>:</title>
<para>
Returns a result of the request execution, when available. This method
returns <code>null</code> if the request execution is still ongoing.
</para>
</formalpara>
</listitem>
<listitem>
<formalpara>
<title><methodname>isDone</methodname>:</title>
<para>
Determines whether or not the request execution completed. If the
request processing terminated normally <methodname>getResult()</methodname>
can be used to obtain the result. If the request processing terminated
abnormally <methodname>getException()</methodname> can be used to obtain
the cause.
</para>
</formalpara>
</listitem>
<listitem>
<formalpara>
<title><methodname>close</methodname>:</title>
<para>
Closes the consumer and releases all resources currently allocated by it.
</para>
</formalpara>
</listitem>
</itemizedlist>
<para>
<interfacename>HttpAsyncRequestConsumer</interfacename> implementations are
expected to be thread-safe.
</para>
<para>
<classname>BasicAsyncRequestConsumer</classname> is a very basic implementation
of the <interfacename>HttpAsyncRequestConsumer</interfacename> interface shipped
with the library. Please note that this consumer buffers request content in memory and
therefore should be used for relatively small request messages.
</para>
</section>
<section>
<title>Asynchronous HTTP response producer</title>
<para>
<interfacename>HttpAsyncResponseProducer</interfacename> facilitates the process of
asynchronous generation of HTTP responses. It is a callback interface used by
<interfacename>HttpAsyncRequestHandler</interfacename>s to generate an HTTP response
message and to stream its content to a non-blocking server side HTTP connection.
</para>
<para>
HTTP I/O events and methods as defined by the
<interfacename>HttpAsyncResponseProducer</interfacename> interface:
</para>
<itemizedlist>
<listitem>
<formalpara>
<title><methodname>generateResponse</methodname>:</title>
<para>
Invoked to generate a HTTP response message header.
</para>
</formalpara>
</listitem>
<listitem>
<formalpara>
<title><methodname>produceContent</methodname>:</title>
<para>
Invoked to write out a chunk of content to the <interfacename>
ContentEncoder</interfacename>. The <interfacename>IOControl
</interfacename> interface can be used to suspend output events if
the producer is temporarily unable to produce more content.
</para>
<para>
When all content is finished, the producer MUST call <methodname>
ContentEncoder#complete()</methodname>. Failure to do so may cause
the entity to be incorrectly delimited.
</para>
<para>
Please note that the <interfacename>ContentEncoder</interfacename> object
is not thread-safe and should only be used within the context of this
method call. The <interfacename>IOControl</interfacename> object can be
shared and used on other thread resume output event notifications when
more content is made available.
</para>
<para>
This event is invoked only for if the outgoing response message has
a content entity enclosed in it, that is <methodname>
HttpResponse#getEntity()</methodname> returns <code>null</code>.
</para>
</formalpara>
</listitem>
<listitem>
<formalpara>
<title><methodname>responseCompleted</methodname>:</title>
<para>
Invoked to signal that the response has been fully written out.
</para>
</formalpara>
</listitem>
<listitem>
<formalpara>
<title><methodname>failed</methodname>:</title>
<para>
Invoked to signal that the response processing terminated abnormally.
</para>
</formalpara>
</listitem>
<listitem>
<formalpara>
<title><methodname>close</methodname>:</title>
<para>
Closes the producer and releases all resources currently allocated by it.
</para>
</formalpara>
</listitem>
</itemizedlist>
<para>
<interfacename>HttpAsyncResponseProducer</interfacename> implementations are
expected to be thread-safe.
</para>
<para>
<classname>BasicAsyncResponseProducer</classname> is a basic implementation
of the <interfacename>HttpAsyncResponseProducer</interfacename> interface shipped
with the library. The producer can make use of the <interfacename>
HttpAsyncContentProducer</interfacename> interface to efficiently stream out
message content to a non-blocking HTTP connection, if it is implemented by the
<interfacename>HttpEntity</interfacename> enclosed in the response.
</para>
</section>
<section>
<title>Non-blocking request handler resolver</title>
<para>
The management of non-blocking HTTP request handlers is quite similar to that of
blocking HTTP request handlers. Usually an instance of <interfacename>
HttpAsyncRequestHandlerResolver</interfacename> is used to maintain a registry of
request handlers and to matches a request URI to a particular request handler.
HttpCore includes only a very simple implementation of the request handler resolver
based on a trivial pattern matching algorithm: <classname>
HttpAsyncRequestHandlerRegistry</classname> supports only three formats:
<literal>*</literal>, <literal>&lt;uri&gt;*</literal> and
<literal>*&lt;uri&gt;</literal>.
</para>
<programlisting><![CDATA[
HttpAsyncRequestHandler<?> myRequestHandler1 = <...>
HttpAsyncRequestHandler<?> myRequestHandler2 = <...>
HttpAsyncRequestHandler<?> myRequestHandler3 = <...>
UriHttpAsyncRequestHandlerMapper handlerReqistry =
new UriHttpAsyncRequestHandlerMapper();
handlerReqistry.register("/service/*", myRequestHandler1);
handlerReqistry.register("*.do", myRequestHandler2);
handlerReqistry.register("*", myRequestHandler3);
]]></programlisting>
<para>
Users are encouraged to provide more sophisticated implementations of
<interfacename>HttpAsyncRequestHandlerResolver</interfacename>, for instance, based
on regular expressions.
</para>
</section>
</section>
<section>
<title>Asynchronous HTTP request executor</title>
<para>
<classname>HttpAsyncRequestExecutor</classname> is a fully asynchronous client side
HTTP protocol handler based on the NIO (non-blocking) I/O model. <classname>
HttpAsyncRequestExecutor</classname> translates individual events fired through the
<interfacename>NHttpClientEventHandler</interfacename> interface into logically
related HTTP message exchanges.
</para>
<para>
<classname>HttpAsyncRequestExecutor</classname> relies on <interfacename>
HttpAsyncRequestExecutionHandler</interfacename> to implement application specific
content generation and processing and to handle logically related series of HTTP
request / response exchanges, which may also span across multiple connections.
<interfacename>HttpProcessor</interfacename> provided by the <interfacename>
HttpAsyncRequestExecutionHandler</interfacename> instance will be used to generate
mandatory protocol headers for all outgoing messages and apply common, cross-cutting
message transformations to all incoming and outgoing messages. The caller is expected
to pass an instance of <interfacename>HttpAsyncRequestExecutionHandler</interfacename>
to be used for the next series of HTTP message exchanges through the connection
context using <methodname>HttpAsyncRequestExecutor#HTTP_HANDLER</methodname> attribute.
HTTP exchange sequence is considered complete when the <methodname>
HttpAsyncRequestExecutionHandler#isDone()</methodname> method returns <code>true</code>.
</para>
<programlisting><![CDATA[
HttpAsyncRequestExecutor ph = new HttpAsyncRequestExecutor();
IOEventDispatch ioEventDispatch = new DefaultHttpClientIODispatch(ph,
new DefaultNHttpClientConnectionFactory(ConnectionConfig.DEFAULT));
ConnectingIOReactor ioreactor = new DefaultConnectingIOReactor();
ioreactor.execute(ioEventDispatch);
]]></programlisting>
<para>
The <classname>HttpAsyncRequester</classname> utility class can be used to abstract
away low level details of <interfacename>HttpAsyncRequestExecutionHandler
</interfacename> management. Please note <classname>HttpAsyncRequester</classname>
supports single HTTP request / response exchanges only. It does not support HTTP
authentication and does not handle redirects automatically.
</para>
<programlisting><![CDATA[
HttpProcessor httpproc = HttpProcessorBuilder.create()
.add(new RequestContent())
.add(new RequestTargetHost())
.add(new RequestConnControl())
.add(new RequestUserAgent("MyAgent-HTTP/1.1"))
.add(new RequestExpectContinue(true))
.build();
HttpAsyncRequester requester = new HttpAsyncRequester(httpproc);
NHttpClientConnection conn = <...>
Future<HttpResponse> future = requester.execute(
new BasicAsyncRequestProducer(
new HttpHost("localhost"),
new BasicHttpRequest("GET", "/")),
new BasicAsyncResponseConsumer(),
conn);
HttpResponse response = future.get();
]]></programlisting>
<section>
<title>Asynchronous HTTP request producer</title>
<para>
<interfacename>HttpAsyncRequestProducer</interfacename> facilitates the process of
asynchronous generation of HTTP requests. It is a callback interface whose methods
get invoked to generate an HTTP request message and to stream message content to
a non-blocking client side HTTP connection.
</para>
<para>
Repeatable request producers capable of generating the same request message more
than once can be reset to their initial state by calling the <methodname>
resetRequest()</methodname> method, at which point request producers are expected
to release currently allocated resources that are no longer needed or re-acquire
resources needed to repeat the process.
</para>
<para>
HTTP I/O events and methods as defined by the
<interfacename>HttpAsyncRequestProducer</interfacename> interface:
</para>
<itemizedlist>
<listitem>
<formalpara>
<title><methodname>getTarget</methodname>:</title>
<para>
Invoked to obtain the request target host.
</para>
</formalpara>
</listitem>
<listitem>
<formalpara>
<title><methodname>generateRequest</methodname>:</title>
<para>
Invoked to generate a HTTP request message header. The message is expected
to implement the <interfacename>HttpEntityEnclosingRequest</interfacename>
interface if it is to enclose a content entity.
</para>
</formalpara>
</listitem>
<listitem>
<formalpara>
<title><methodname>produceContent</methodname>:</title>
<para>
Invoked to write out a chunk of content to the <interfacename>
ContentEncoder</interfacename>. The <interfacename>IOControl
</interfacename> interface can be used to suspend output events if
the producer is temporarily unable to produce more content.
</para>
<para>
When all content is finished, the producer MUST call <methodname>
ContentEncoder#complete()</methodname>. Failure to do so may cause
the entity to be incorrectly delimited
</para>
<para>
Please note that the <interfacename>ContentEncoder</interfacename> object
is not thread-safe and should only be used within the context of this
method call. The <interfacename>IOControl</interfacename> object can be
shared and used on other thread resume output event notifications when
more content is made available.
</para>
<para>
This event is invoked only for if the outgoing request message has
a content entity enclosed in it, that is <methodname>
HttpEntityEnclosingRequest#getEntity()</methodname> returns <code>null
</code>.
</para>
</formalpara>
</listitem>
<listitem>
<formalpara>
<title><methodname>requestCompleted</methodname>:</title>
<para>
Invoked to signal that the request has been fully written out.
</para>
</formalpara>
</listitem>
<listitem>
<formalpara>
<title><methodname>failed</methodname>:</title>
<para>
Invoked to signal that the request processing terminated abnormally.
</para>
</formalpara>
</listitem>
<listitem>
<formalpara>
<title><methodname>resetRequest</methodname>:</title>
<para>
Invoked to reset the producer to its initial state. Repeatable request
producers are expected to release currently allocated resources that are
no longer needed or re-acquire resources needed to repeat the process.
</para>
</formalpara>
</listitem>
<listitem>
<formalpara>
<title><methodname>close</methodname>:</title>
<para>
Closes the producer and releases all resources currently allocated by it.
</para>
</formalpara>
</listitem>
</itemizedlist>
<para>
<interfacename>HttpAsyncRequestProducer</interfacename> implementations are
expected to be thread-safe.
</para>
<para>
<classname>BasicAsyncRequestProducer</classname> is a basic implementation
of the <interfacename>HttpAsyncRequestProducer</interfacename> interface shipped
with the library. The producer can make use of the <interfacename>
HttpAsyncContentProducer</interfacename> interface to efficiently stream out
message content to a non-blocking HTTP connection, if it is implemented by the
<interfacename>HttpEntity</interfacename> enclosed in the request.
</para>
</section>
<section>
<title>Asynchronous HTTP response consumer</title>
<para>
<interfacename>HttpAsyncResponseConsumer</interfacename> facilitates the process of
asynchronous processing of HTTP responses. It is a callback interface whose methods
get invoked to process an HTTP response message and to stream message content from
a non-blocking client side HTTP connection.
</para>
<para>
HTTP I/O events and methods as defined by the <interfacename>
HttpAsyncResponseConsumer</interfacename> interface:
</para>
<itemizedlist>
<listitem>
<formalpara>
<title><methodname>responseReceived</methodname>:</title>
<para>
Invoked when a HTTP response message is received.
</para>
</formalpara>
</listitem>
<listitem>
<formalpara>
<title><methodname>consumeContent</methodname>:</title>
<para>
Invoked to process a chunk of content from the <interfacename>
ContentDecoder</interfacename>. The <interfacename>IOControl
</interfacename> interface can be used to suspend input events if
the consumer is temporarily unable to consume more content.
</para>
<para>
The consumer can use the <methodname>ContentDecoder#isCompleted()
</methodname> method to find out whether or not the message content
has been fully consumed.
</para>
<para>
Please note that the <interfacename>ContentDecoder</interfacename> object
is not thread-safe and should only be used within the context of this
method call. The <interfacename>IOControl</interfacename> object can be
shared and used on other thread to resume input event notifications
when the consumer is capable of processing more content.
</para>
<para>
This event is invoked only for if the incoming response message has
a content entity enclosed in it.
</para>
</formalpara>
</listitem>
<listitem>
<formalpara>
<title><methodname>responseCompleted</methodname>:</title>
<para>
Invoked to signal that the response has been fully processed.
</para>
</formalpara>
</listitem>
<listitem>
<formalpara>
<title><methodname>failed</methodname>:</title>
<para>
Invoked to signal that the response processing terminated abnormally.
</para>
</formalpara>
</listitem>
<listitem>
<formalpara>
<title><methodname>getException</methodname>:</title>
<para>
Returns an exception in case of an abnormal termination. This method
returns <code>null</code> if the response processing is still ongoing or
if it completed successfully.
</para>
</formalpara>
</listitem>
<listitem>
<formalpara>
<title><methodname>getResult</methodname>:</title>
<para>
Returns a result of the response processing, when available. This method
returns <code>null</code> if the response processing is still ongoing.
</para>
</formalpara>
</listitem>
<listitem>
<formalpara>
<title><methodname>isDone</methodname>:</title>
<para>
Determines whether or not the response processing completed. If the
response processing terminated normally <methodname>getResult()</methodname>
can be used to obtain the result. If the response processing terminated
abnormally <methodname>getException()</methodname> can be used to obtain
the cause.
</para>
</formalpara>
</listitem>
<listitem>
<formalpara>
<title><methodname>close</methodname>:</title>
<para>
Closes the consumer and releases all resources currently allocated by it.
</para>
</formalpara>
</listitem>
</itemizedlist>
<para>
<interfacename>HttpAsyncResponseConsumer</interfacename> implementations are
expected to be thread-safe.
</para>
<para>
<classname>BasicAsyncResponseConsumer</classname> is a very basic implementation
of the <interfacename>HttpAsyncResponseConsumer</interfacename> interface shipped
with the library. Please note that this consumer buffers response content in memory
and therefore should be used for relatively small response messages.
</para>
</section>
</section>
</section>
<section>
<title>Non-blocking connection pools</title>
<para>
Non-blocking connection pools are quite similar to blocking one with one significant
distinction that they have to reply an I/O reactor to establish new connections.
As a result connections leased from a non-blocking pool are returned fully initialized and
already bound to a particular I/O session. Non-blocking connections managed by a connection
pool cannot be bound to an arbitrary I/O session.
</para>
<programlisting><![CDATA[
HttpHost target = new HttpHost("localhost");
ConnectingIOReactor ioreactor = <...>
BasicNIOConnPool connpool = new BasicNIOConnPool(ioreactor);
connpool.lease(target, null,
10, TimeUnit.SECONDS,
new FutureCallback<BasicNIOPoolEntry>() {
@Override
public void completed(BasicNIOPoolEntry entry) {
NHttpClientConnection conn = entry.getConnection();
System.out.println("Connection successfully leased");
// Update connection context and request output
conn.requestOutput();
}
@Override
public void failed(Exception ex) {
System.out.println("Connection request failed");
ex.printStackTrace();
}
@Override
public void cancelled() {
}
});
]]></programlisting>
<para>
Please note due to event-driven nature of asynchronous communication model it is quite
difficult to ensure proper release of persistent connections back to the pool. One can make
use of <classname>HttpAsyncRequester</classname> to handle connection lease and release
behind the scene.
</para>
<programlisting><![CDATA[
ConnectingIOReactor ioreactor = <...>
HttpProcessor httpproc = <...>
BasicNIOConnPool connpool = new BasicNIOConnPool(ioreactor);
HttpAsyncRequester requester = new HttpAsyncRequester(httpproc);
HttpHost target = new HttpHost("localhost");
Future<HttpResponse> future = requester.execute(
new BasicAsyncRequestProducer(
new HttpHost("localhost"),
new BasicHttpRequest("GET", "/")),
new BasicAsyncResponseConsumer(),
connpool);
]]></programlisting>
</section>
<section>
<title>Pipelined request execution</title>
<para>
In addition to the normal request / response execution mode <classname>HttpAsyncRequester
</classname> is also capable of executing requests in the so called pipelined mode whereby
several requests are immediately written out to the underlying connection. Please note that
entity enclosing requests can be executed in the pipelined mode but the 'expect: continue'
handshake should be disabled (request messages should contains no 'Expect: 100-continue'
header).
</para>
<programlisting><![CDATA[
HttpProcessor httpproc = <...>
HttpAsyncRequester requester = new HttpAsyncRequester(httpproc);
HttpHost target = new HttpHost("www.apache.org");
List<BasicAsyncRequestProducer> requestProducers = Arrays.asList(
new BasicAsyncRequestProducer(target, new BasicHttpRequest("GET", "/index.html")),
new BasicAsyncRequestProducer(target, new BasicHttpRequest("GET", "/foundation/index.html")),
new BasicAsyncRequestProducer(target, new BasicHttpRequest("GET", "/foundation/how-it-works.html"))
);
List<BasicAsyncResponseConsumer> responseConsumers = Arrays.asList(
new BasicAsyncResponseConsumer(),
new BasicAsyncResponseConsumer(),
new BasicAsyncResponseConsumer()
);
HttpCoreContext context = HttpCoreContext.create();
Future<List<HttpResponse>> future = requester.executePipelined(
target, requestProducers, responseConsumers, pool, context, null);
]]></programlisting>
<para>
Please note that older web servers and especially older HTTP proxies may be unable to handle
pipelined requests correctly. Use the pipelined execution mode with caution.
</para>
</section>
<section>
<title>Non-blocking TLS/SSL</title>
<section>
<title>SSL I/O session</title>
<para>
<classname>SSLIOSession</classname> is a decorator class intended to transparently
extend any arbitrary <interfacename>IOSession</interfacename> with transport layer
security capabilities based on the SSL/TLS protocol. Default HTTP connection
implementations and protocol handlers should be able to work with SSL sessions without
special preconditions or modifications.
</para>
<programlisting><![CDATA[
SSLContext sslcontext = SSLContext.getInstance("Default");
sslcontext.init(null, null, null);
// Plain I/O session
IOSession iosession = <...>
SSLIOSession sslsession = new SSLIOSession(
iosession, SSLMode.CLIENT, sslcontext, null);
iosession.setAttribute(SSLIOSession.SESSION_KEY, sslsession);
NHttpClientConnection conn = new DefaultNHttpClientConnection(
sslsession, 8 * 1024);
]]></programlisting>
<para>
One can also use <classname>SSLNHttpClientConnectionFactory</classname> or <classname>
SSLNHttpServerConnectionFactory</classname> classes to conveniently create SSL
encrypterd HTTP connections.
</para>
<programlisting><![CDATA[
SSLContext sslcontext = SSLContext.getInstance("Default");
sslcontext.init(null, null, null);
// Plain I/O session
IOSession iosession = <...>
SSLNHttpClientConnectionFactory connfactory = new SSLNHttpClientConnectionFactory(
sslcontext, null, ConnectionConfig.DEFAULT);
NHttpClientConnection conn = connfactory.createConnection(iosession);
]]></programlisting>
<section>
<title>SSL setup handler</title>
<para>
Applications can customize various aspects of the TLS/SSl protocol by passing a
custom implementation of the <interfacename>SSLSetupHandler</interfacename>
interface.
</para>
<para>
SSL events as defined by the <interfacename>SSLSetupHandler</interfacename>
interface:
</para>
<itemizedlist>
<listitem>
<formalpara>
<title><methodname>initalize</methodname>:</title>
<para>
Triggered when the SSL connection is being initialized. The handler can use
this callback to customize properties of the <classname>
javax.net.ssl.SSLEngine</classname> used to establish the SSL session.
</para>
</formalpara>
</listitem>
<listitem>
<formalpara>
<title><methodname>verify</methodname>:</title>
<para>
Triggered when the SSL connection has been established and initial SSL
handshake has been successfully completed. The handler can use this
callback to verify properties of the SSLSession. For instance this would
be the right place to enforce SSL cipher strength, validate certificate
chain and do hostname checks.
</para>
</formalpara>
</listitem>
</itemizedlist>
<programlisting><![CDATA[
SSLContext sslcontext = SSLContexts.createDefault();
// Plain I/O session
IOSession iosession = <...>
SSLIOSession sslsession = new SSLIOSession(
iosession, SSLMode.CLIENT, sslcontext, new SSLSetupHandler() {
public void initalize(final SSLEngine sslengine) throws SSLException {
// Enforce TLS and disable SSL
sslengine.setEnabledProtocols(new String[] {
"TLSv1",
"TLSv1.1",
"TLSv1.2" });
// Enforce strong ciphers
sslengine.setEnabledCipherSuites(new String[] {
"TLS_RSA_WITH_AES_256_CBC_SHA",
"TLS_DHE_RSA_WITH_AES_256_CBC_SHA",
"TLS_DHE_DSS_WITH_AES_256_CBC_SHA" });
}
public void verify(
final IOSession iosession,
final SSLSession sslsession) throws SSLException {
X509Certificate[] certs = sslsession.getPeerCertificateChain();
// Examine peer certificate chain
for (X509Certificate cert: certs) {
System.out.println(cert.toString());
}
}
});
]]></programlisting>
<para>
<interfacename>SSLSetupHandler</interfacename> impelemntations can also be used with
the <classname>SSLNHttpClientConnectionFactory</classname> or <classname>
SSLNHttpServerConnectionFactory</classname> classes.
</para>
<programlisting><![CDATA[
SSLContext sslcontext = SSLContexts.createDefault();
// Plain I/O session
IOSession iosession = <...>
SSLSetupHandler mysslhandler = new SSLSetupHandler() {
public void initalize(final SSLEngine sslengine) throws SSLException {
// Enforce TLS and disable SSL
sslengine.setEnabledProtocols(new String[] {
"TLSv1",
"TLSv1.1",
"TLSv1.2" });
}
public void verify(
final IOSession iosession, final SSLSession sslsession) throws SSLException {
}
};
SSLNHttpClientConnectionFactory connfactory = new SSLNHttpClientConnectionFactory(
sslcontext, mysslhandler, ConnectionConfig.DEFAULT);
NHttpClientConnection conn = connfactory.createConnection(iosession);
]]></programlisting>
</section>
</section>
<section>
<title>TLS/SSL aware I/O event dispatches</title>
<para>
Default <interfacename>IOEventDispatch</interfacename> implementations shipped with
the library such as <classname>DefaultHttpServerIODispatch</classname> and <classname>
DefaultHttpClientIODispatch</classname> automatically detect SSL encrypted sessions
and handle SSL transport aspects transparently. However, custom I/O event dispatchers
that do not extend <classname>AbstractIODispatch</classname> are required to take some
additional actions to ensure correct functioning of the transport layer encryption.
</para>
<itemizedlist>
<listitem>
<para>
The I/O dispatch may need to call <methodname>SSLIOSession#initalize()
</methodname> method in order to put the SSL session either into a client or
a server mode, if the SSL session has not been yet initialized.
</para>
</listitem>
<listitem>
<para>
When the underlying I/O session is input ready, the I/O dispatcher should check
whether the SSL I/O session is ready to produce input data by calling
<methodname>SSLIOSession#isAppInputReady()</methodname>, pass control to the
protocol handler if it is, and finally call <methodname>
SSLIOSession#inboundTransport()</methodname> method in order to do the
necessary SSL handshaking and decrypt input data.
</para>
</listitem>
<listitem>
<para>
When the underlying I/O session is output ready, the I/O dispatcher should
check whether the SSL I/O session is ready to accept output data by calling
<methodname>SSLIOSession#isAppOutputReady()</methodname>, pass control to the
protocol handler if it is, and finally call <methodname>
SSLIOSession#outboundTransport()</methodname> method in order to do the necessary
SSL handshaking and encrypt application data.
</para>
</listitem>
</itemizedlist>
</section>
</section>
<section>
<title>Embedded non-blocking HTTP server</title>
<para>
As of version 4.4 HttpCore ships with an embedded non-blocking HTTP server based on
non-blocking I/O components described above.
</para>
<programlisting><![CDATA[
HttpAsyncRequestHandler<?> requestHandler = <...>
HttpProcessor httpProcessor = <...>
SocketConfig socketConfig = SocketConfig.custom()
.setSoTimeout(15000)
.setTcpNoDelay(true)
.build();
final HttpServer server = ServerBootstrap.bootstrap()
.setListenerPort(8080)
.setHttpProcessor(httpProcessor)
.setSocketConfig(socketConfig)
.setExceptionLogger(new StdErrorExceptionLogger())
.registerHandler("*", requestHandler)
.create();
server.start();
server.awaitTermination(Long.MAX_VALUE, TimeUnit.DAYS);
Runtime.getRuntime().addShutdownHook(new Thread() {
@Override
public void run() {
server.shutdown(5, TimeUnit.SECONDS);
}
});
]]></programlisting>
</section>
</chapter>