<?xml version="1.0"?><!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"><html xmlns="http://www.w3.org/1999/xhtml" lang="en"><head><title>Twisted Documentation: Generating Deferreds</title><link href="../howto/stylesheet.css" type="text/css" rel="stylesheet" /></head><body bgcolor="white"><h1 class="title">Generating Deferreds</h1><div class="toc"><ol><li><a href="#auto0">Class overview</a></li><ul><li><a href="#auto1">Basic Callback Functions</a></li></ul><li><a href="#auto2">What Deferreds don't do: make your code asynchronous</a></li><li><a href="#auto3">Advanced Processing Chain Control</a></li><li><a href="#auto4">Returning Deferreds from synchronous functions</a></li><li><a href="#auto5">Integrating blocking code with Twisted</a></li><li><a href="#auto6">Possible sources of error</a></li><ul><li><a href="#auto7">Firing Deferreds more than once is impossible</a></li><li><a href="#auto8">Synchronous callback execution</a></li></ul></ol></div><div class="content"><span></span><p><code base="twisted.internet.defer" class="API">Deferred</code> objects are signals that a function you have called does not yet have the data you want available. When a function returns a Deferred object, your calling function attaches callbacks to it to handle the data when available.</p><p>This document addresses the other half of the question: writing functions that return Deferreds, that is, constructing Deferred objects, arranging for them to be returned immediately without blocking until data is available, and firing their callbacks when the data is available.</p><p>This document assumes that you are familiar with the <a href="async.html">asynchronous model</a> used by Twisted, and with <a href="defer.html">using deferreds returned by functions</a>.</p><a name="class"></a><h2>Class overview<a name="auto0"></a></h2><p>This is an overview API reference for Deferred from the point of creating a Deferred and firing its callbacks and errbacks. It is not meant to be a substitute for the docstrings in the Deferred class, but can provide guidelines for its use.</p><p>There is a parallel overview of functions used by calling function which the Deferred is returned to at <a href="defer.html#class">Using Deferreds</a>.</p><h3>Basic Callback Functions<a name="auto1"></a></h3><ul><li><code class="py-prototype">callback(result)</code><p>Run success callbacks with the given result. <em>This can only be run once.</em> Later calls to this or <code>errback</code> will raise <code class="API">twisted.internet.defer.AlreadyCalledError</code>. If further callbacks or errbacks are added after this point, addCallbacks will run the callbacks immediately.</p></li><li><code class="py-prototype">errback(failure)</code><p>Run error callbacks with the given failure. <em>This can only be run once.</em> Later calls to this or <code>callback</code> will raise <code class="API">twisted.internet.defer.AlreadyCalledError</code>. If further callbacks or errbacks are added after this point, addCallbacks will run the callbacks immediately.</p></li></ul><h2>What Deferreds don't do: make your code asynchronous<a name="auto2"></a></h2><p><em>Deferreds do not make the code magically not block.</em></p><p>Let's take this function as an example:</p><pre class="python"> <span class="py-src-keyword">from</span> <span class="py-src-variable">twisted</span>.<span class="py-src-variable">internet</span> <span class="py-src-keyword">import</span> <span class="py-src-variable">defer</span> <span class="py-src-variable">TARGET</span> = <span class="py-src-number">10000</span> <span class="py-src-keyword">def</span> <span class="py-src-identifier">largeFibonnaciNumber</span>(): <span class="py-src-comment"># create a Deferred object to return: </span> <span class="py-src-variable">d</span> = <span class="py-src-variable">defer</span>.<span class="py-src-variable">Deferred</span>() <span class="py-src-comment"># calculate the ten thousandth Fibonnaci number </span> <span class="py-src-variable">first</span> = <span class="py-src-number">0</span> <span class="py-src-variable">second</span> = <span class="py-src-number">1</span> <span class="py-src-keyword">for</span> <span class="py-src-variable">i</span> <span class="py-src-keyword">in</span> <span class="py-src-variable">xrange</span>(<span class="py-src-variable">TARGET</span> - <span class="py-src-number">1</span>): <span class="py-src-variable">new</span> = <span class="py-src-variable">first</span> + <span class="py-src-variable">second</span> <span class="py-src-variable">first</span> = <span class="py-src-variable">second</span> <span class="py-src-variable">second</span> = <span class="py-src-variable">new</span> <span class="py-src-keyword">if</span> <span class="py-src-variable">i</span> % <span class="py-src-number">100</span> == <span class="py-src-number">0</span>: <span class="py-src-keyword">print</span> <span class="py-src-string">"Progress: calculating the %dth Fibonnaci number"</span> % <span class="py-src-variable">i</span> <span class="py-src-comment"># give the Deferred the answer to pass to the callbacks: </span> <span class="py-src-variable">d</span>.<span class="py-src-variable">callback</span>(<span class="py-src-variable">second</span>) <span class="py-src-comment"># return the Deferred with the answer: </span> <span class="py-src-keyword">return</span> <span class="py-src-variable">d</span> <span class="py-src-keyword">import</span> <span class="py-src-variable">time</span> <span class="py-src-variable">timeBefore</span> = <span class="py-src-variable">time</span>.<span class="py-src-variable">time</span>() <span class="py-src-comment"># call the function and get our Deferred </span><span class="py-src-variable">d</span> = <span class="py-src-variable">largeFibonnaciNumber</span>() <span class="py-src-variable">timeAfter</span> = <span class="py-src-variable">time</span>.<span class="py-src-variable">time</span>() <span class="py-src-keyword">print</span> <span class="py-src-string">"Total time taken for largeFibonnaciNumber call: %0.3f seconds"</span> % (<span class="py-src-variable">timeAfter</span> - <span class="py-src-variable">timeBefore</span>) <span class="py-src-comment"># add a callback to it to print the number </span> <span class="py-src-keyword">def</span> <span class="py-src-identifier">printNumber</span>(<span class="py-src-parameter">number</span>): <span class="py-src-keyword">print</span> <span class="py-src-string">"The %dth Fibonacci number is %d"</span> % (<span class="py-src-variable">TARGET</span>, <span class="py-src-variable">number</span>) <span class="py-src-keyword">print</span> <span class="py-src-string">"Adding the callback now."</span> <span class="py-src-variable">d</span>.<span class="py-src-variable">addCallback</span>(<span class="py-src-variable">printNumber</span>) </pre><p>You will notice that despite creating a Deferred in the <code>largeFibonnaciNumber</code> function, these things happened:</p><ul><li>the "Total time taken for largeFibonnaciNumber call" output shows that the function did not return immediately as asynchronous functions are expected to do; and</li><li>rather than the callback being added before the result was available and called after the result is available, it isn't even added until after the calculation has been completed.</li></ul><p> The function completed its calculation before returning, blocking the process until it had finished, which is exactly what asynchronous functions are not meant to do. Deferreds are not a non-blocking talisman: they are a signal for asynchronous functions to <em>use</em> to pass results onto callbacks, but using them does not guarantee that you have an asynchronous function.</p><h2>Advanced Processing Chain Control<a name="auto3"></a></h2><ul><li><code class="py-prototype">pause()</code><p>Cease calling any methods as they are added, and do not respond to <code>callback</code>, until <code>self.unpause()</code> is called.</p></li><li><code class="py-prototype">unpause()</code><p>If <code>callback</code> has been called on this Deferred already, call all the callbacks that have been added to this Deferred since <code>pause</code> was called.</p><p>Whether it was called or not, this will put this Deferred in a state where further calls to <code>addCallbacks</code> or <code>callback</code> will work as normal.</p></li></ul><h2>Returning Deferreds from synchronous functions<a name="auto4"></a></h2><p>Sometimes you might wish to return a Deferred from a synchronous function. There are several reasons why, the major two are maintaining API compatibility with another version of your function which returns a Deferred, or allowing for the possiblity that in the future your function might need to be asynchronous.</p><p>In the <a href="defer.html">Using Deferreds</a> reference, we gave the following example of a synchronous function:</p><div class="py-listing"><pre> <span class="py-src-keyword">def</span> <span class="py-src-identifier">synchronousIsValidUser</span>(<span class="py-src-parameter">user</span>): <span class="py-src-string">''' Return true if user is a valid user, false otherwise '''</span> <span class="py-src-keyword">return</span> <span class="py-src-variable">user</span> <span class="py-src-keyword">in</span> [<span class="py-src-string">"Alice"</span>, <span class="py-src-string">"Angus"</span>, <span class="py-src-string">"Agnes"</span>] </pre><div class="caption">Source listing - <a href="listings/deferred/synch-validation.py"><span class="filename">listings/deferred/synch-validation.py</span></a></div></div><p>While we can require that callers of our function wrap our synchronous result in a Deferred using <code base="twisted.internet.defer" class="API">maybeDeferred</code>, for the sake of API compatibility it is better to return a Deferred ourself using <code base="twisted.internet" class="API">defer.succeed</code>:</p><pre class="python"> <span class="py-src-keyword">from</span> <span class="py-src-variable">twisted</span>.<span class="py-src-variable">internet</span> <span class="py-src-keyword">import</span> <span class="py-src-variable">defer</span> <span class="py-src-keyword">def</span> <span class="py-src-identifier">immediateIsValidUser</span>(<span class="py-src-parameter">user</span>): <span class="py-src-string">''' Returns a Deferred resulting in true if user is a valid user, false otherwise '''</span> <span class="py-src-variable">result</span> = <span class="py-src-variable">user</span> <span class="py-src-keyword">in</span> [<span class="py-src-string">"Alice"</span>, <span class="py-src-string">"Angus"</span>, <span class="py-src-string">"Agnes"</span>] <span class="py-src-comment"># return a Deferred object already called back with the value of result </span> <span class="py-src-keyword">return</span> <span class="py-src-variable">defer</span>.<span class="py-src-variable">succeed</span>(<span class="py-src-variable">result</span>) </pre><p>There is an equivalent <code base="twisted.internet" class="API">defer.fail</code> method to return a Deferred with the errback chain already fired.</p><h2>Integrating blocking code with Twisted<a name="auto5"></a></h2><p>At some point, you are likely to need to call a blocking function: many functions in third party libraries will have long running blocking functions. There is no way to 'force' a function to be asynchronous: it must be written that way specifically. When using Twisted, your own code should be asynchronous, but there is no way to make third party functions asynchronous other than rewriting them.</p><p>In this case, Twisted provides the ability to run the blocking code in a separate thread rather than letting it block your application. The <code class="API">twisted.internet.threads.deferToThread</code> function will set up a thread to run your blocking function, return a Deferred and later fire that Deferred when the thread completes.</p><p>Let's assume our <code class="python">largeFibonnaciNumber</code> function from above is in a third party library (returning the result of the calculation, not a Deferred) and is not easily modifiable to be finished in discrete blocks. This example shows it being called in a thread, unlike in the earlier section we'll see that the operation does not block our entire program:</p><pre class="python"> <span class="py-src-keyword">def</span> <span class="py-src-identifier">largeFibonnaciNumber</span>(): <span class="py-src-string">""" Represent a long running blocking function by calculating the TARGETth Fibonnaci number """</span> <span class="py-src-variable">TARGET</span> = <span class="py-src-number">10000</span> <span class="py-src-variable">first</span> = <span class="py-src-number">0</span> <span class="py-src-variable">second</span> = <span class="py-src-number">1</span> <span class="py-src-keyword">for</span> <span class="py-src-variable">i</span> <span class="py-src-keyword">in</span> <span class="py-src-variable">xrange</span>(<span class="py-src-variable">TARGET</span> - <span class="py-src-number">1</span>): <span class="py-src-variable">new</span> = <span class="py-src-variable">first</span> + <span class="py-src-variable">second</span> <span class="py-src-variable">first</span> = <span class="py-src-variable">second</span> <span class="py-src-variable">second</span> = <span class="py-src-variable">new</span> <span class="py-src-keyword">return</span> <span class="py-src-variable">second</span> <span class="py-src-keyword">from</span> <span class="py-src-variable">twisted</span>.<span class="py-src-variable">internet</span> <span class="py-src-keyword">import</span> <span class="py-src-variable">threads</span>, <span class="py-src-variable">reactor</span> <span class="py-src-keyword">def</span> <span class="py-src-identifier">fibonacciCallback</span>(<span class="py-src-parameter">result</span>): <span class="py-src-string">""" Callback which manages the largeFibonnaciNumber result by printing it out """</span> <span class="py-src-keyword">print</span> <span class="py-src-string">"largeFibonnaciNumber result ="</span>, <span class="py-src-variable">result</span> <span class="py-src-comment"># make sure the reactor stops after the callback chain finishes, </span> <span class="py-src-comment"># just so that this example terminates </span> <span class="py-src-variable">reactor</span>.<span class="py-src-variable">stop</span>() <span class="py-src-keyword">def</span> <span class="py-src-identifier">run</span>(): <span class="py-src-string">""" Run a series of operations, deferring the largeFibonnaciNumber operation to a thread and performing some other operations after adding the callback """</span> <span class="py-src-comment"># get our Deferred which will be called with the largeFibonnaciNumber result </span> <span class="py-src-variable">d</span> = <span class="py-src-variable">threads</span>.<span class="py-src-variable">deferToThread</span>(<span class="py-src-variable">largeFibonnaciNumber</span>) <span class="py-src-comment"># add our callback to print it out </span> <span class="py-src-variable">d</span>.<span class="py-src-variable">addCallback</span>(<span class="py-src-variable">fibonacciCallback</span>) <span class="py-src-comment"># unless the largeFibonnaciNumber thread returns very fast, these print </span> <span class="py-src-comment">#lines should happen first </span> <span class="py-src-keyword">print</span> <span class="py-src-string">"1st line after the addition of the callback"</span> <span class="py-src-keyword">print</span> <span class="py-src-string">"2nd line after the addition of the callback"</span> <span class="py-src-keyword">if</span> <span class="py-src-variable">__name__</span> == <span class="py-src-string">'__main__'</span>: <span class="py-src-variable">run</span>() <span class="py-src-variable">reactor</span>.<span class="py-src-variable">run</span>() </pre><h2>Possible sources of error<a name="auto6"></a></h2><p>Deferreds greatly simplify the process of writing asynchronous code by providing a standard for registering callbacks, but there are some subtle and sometimes confusing rules that you need to follow if you are going to use them. This mostly applies to people who are writing new systems that use Deferreds internally, and not writers of applications that just add callbacks to Deferreds produced and processed by other systems. Nevertheless, it is good to know.</p><h3>Firing Deferreds more than once is impossible<a name="auto7"></a></h3><p>Deferreds are one-shot. You can only call <code>Deferred.callback</code> or <code>Deferred.errback</code> once. The processing chain continues each time you add new callbacks to an already-called-back-to Deferred.</p><h3>Synchronous callback execution<a name="auto8"></a></h3><p>If a Deferred already has a result available, addCallback <strong>may</strong> call the callback synchronously: that is, immediately after it's been added. In situations where callbacks modify state, it is might be desirable for the chain of processing to halt until all callbacks are added. For this, it is possible to <code>pause</code> and <code>unpause</code> a Deferred's processing chain while you are adding lots of callbacks.</p><p>Be careful when you use these methods! If you <code>pause</code> a Deferred, it is <em>your</em> responsibility to make sure that you unpause it. The function adding the callbacks must unpause a paused Deferred, it should <em>never</em> be the responsibility of the code that actually fires the callback chain by calling <code>callback</code> or <code>errback</code> as this would negate its usefulness!</p></div><p><a href="../howto/index.html">Index</a></p><span class="version">Version: 2.5.0</span></body></html>