The purpose of a Fake Object, a kind of Test Double, is to replace a collaborator with a functional copy. While Mocks prefer a specification of the behavior to check, Fake Objects are really a simplified version of the production object they substitute. A good Fake Object, however, will be very lightweight, easy to create and throw away to ensure test isolation. Usually it won't satisfy some other functional or non-functional requirements, otherwise we would use him instead of the real object. For example a UserRepository, which normally stores users in a database, may be substitued by a Fake that: does not send mails when an user is added. Doesn't really save anything in the database, but keeps a list in an array. When some complex method is called, just throws a NonImplementedException. Utility Of course performance and less brittleness of tests are the main selling points of a Fake Objects: think about testing with a real database and without one; however this is true for all Test Doubles and I don't need to repeat it. Often the Fake is used when the contract between the SUT and the collaborator is too complex to effectively build a Stub or Mock: many methods, or many calls to the same method are made. Return parameters are difficult to setup. A Fake avoids overspecifying a test for a very invasive contract, like the order of calls and precise parameters; a real implementation is more robust to changes in the contract. For example if the Fake is a collection, you can store and retrieve objects of another type, and still not change the Fake; expectations and configured return value instead would change (together). As an example, an embryonal Fake is PHPUnit $this->returnArgument() for the will() Mock expectation. It's a piece of functionality which substitutes an hard-coded expectation, in order to return one of the arguments of the call instead of adding the argument to the expectation itself. Note that many times we don't have to create new code to leverage Fakes: we can use the existing one with a different configuration (like a Cache maximumCachedItems=0, or with an adapter that use an in-memory cache instead of APC or Memcache) or something provided for us, like an sqlite in-memory database created by PDO. NullObjects are sometimes Fake, but are used in production more than in testing. Implementation Several steps are necessary for building an hand-rolled Fake: create the Fake class by hand, and define the simplest implementation that could do the job for this test. Remember, you're not testing the Fake, you're testing the SUT, so the Fake doesn't have to be perfect, but just passable for the test case at end. Instance an object from the Fake class: the constructor may be different from the real collaborator's one as it's not part of the contract. Install the Fake into the SUT, and proceed as normal. The steps for generating a Fake with PHPUnit are a little different; however, you will still have to write its businss logic: create the "mock" as always via getMock() or getMockBuilder(). Define expectations for its methods with will($this->returnCallback()) and some anonymous functions (or via self-shunting like we did for "Stubs"). Often objects passed in this closures (such as ArrayObject instances) can aid in making the Fake methods communicate with each other. Install the Fake and proceed with the test. The second approach is invaluable when you have a single method on the Fake: it's very fast. Variations Fake Database: an alternative implementation of the Database Facade that lets you test classes that depend on the database without actually using it. For exaple, a DAO which internally use an SplObjectStorage instead of the connection. In-Memory Database: sqlite3 in-memory database used with ORMs for tests that isolate you from the real database, but not from using PDO. High Return On Investment. Fake Web Service: mimics the interface of some web service like Google Analytics, when you have to interact also in write and not only in read. Fake Service Layer: simplified implementations of Service Layer classes, which avoid checking concurrency, authorization, authentication and so on in order to simplify functional testing. Example In this code sample, the test target a ForumManager class which needs to manipulate Posts. It's not a simple Facade: it moves Post around, merges threads and so on. So we inject in it a FakePostDao: ForumManager will call it instead of the database. When you I have many tests of this type, the time for writing the Fake implementation is well spent. array( new Post('Hello'), new Post('Hello!'), new Post('') ), 2 => array( new Post('Hi'), new Post('Hi!') ), 3 => array( new Post('Good morning.') ) )); $forumManager = new ForumManager($dao); $forumManager->mergeThreadsByIds(1, 2); $thread = $dao->getThread(1); $this->assertEquals(5, count($thread)); } } /** * The SUT. */ class ForumManager { private $dao; public function __construct(PostsDao $dao) { $this->dao = $dao; } public function mergeThreadsByIds($originalId, $toBeMergedId) { $original = $this->dao->getThread($originalId); $toBeMerged = $this->dao->getThread($toBeMergedId); $newOne = array_merge($original, $toBeMerged); $this->dao->removeThread($originalId); $this->dao->removeThread($toBeMergedId); $this->dao->addThread($originalId, $newOne); } } /** * Interface for the collaborator to substitute with the Test Double. */ interface PostsDao { public function getThread($id); public function removeThread($id); public function addThread($id, array $thread); } /** * Fake implementation. */ class FakePostDao implements PostsDao { private $threads; public function __construct(array $initialState) { $this->threads = $initialState; } public function getThread($id) { return $this->threads[$id]; } public function removeThread($id) { unset($this->threads[$id]); } /** * We model Thread as array of Posts for simplicity. */ public function addThread($id, array $thread) { $this->threads[$id] = $thread; } } /** * Again a Dummy object: minimal implementation, to make this test pass. */ class Post { }

There has been a lot of buzz lately on high-availability and clustering. Most developers don't care and why should they? These features should be transparent to the application architecture and not something of concern to the developers of that application. But knowledge never hurts, so I emerged myself into the world of load balancing, heartbeats and virtual IP addresses. And you know what? Next time we need a infrastructure like this, I can at least sit down with the guys from the infrastructure department and at least know what the hell they are talking about. So what exactly is a high-availability clustered infrastructure (HACI, as I'll call it from now on) ? In essence, it should be a zero-downtime infrastructure (or at least perceived as one by the end user, which means never ever returning a default browser 404 page), capable of horizontal scaling when the need for it arises and without a single point of failure. It's the SLA writer's dream. A basic HACI setup looks like this: The users enters through a virtual IP address, assigned to one of the two load balancers. Only one of the load-balancers is active (the active master, LB1), the other one is there in the event LB1 fails ((LB2, a passive slave). The two load balancers are redundant, ie. having the exact same configuration. The load balancers redirect all traffic to the real servers. This can be done through round-robin assignment or through other means like sticky sessions, where the same user is redirected to the same server each and every time within a session. Servers can be added at any moment and configured on the load balancers. Ideally, the load balancer configuration is aware of the hardware specification and balances the load accordingly, but that's beyond the scope of this article (it involves adding weights). If all servers balanced by the load balancer fail, a backup server should be used to redirect all traffic coming from the load balancer. This can be a very lightweight server, which purpose is only to provide a sensible error page to the user (something like 'Sorry, we are performing maintenance'). Again, perception and immediate feedback to the user is key. You don't want to show the user a plain 404 page. Off course, if the backup server goes down too, you're in trouble (off course, by that time, warning bells should have gone off on every level in the hierarchy). So how to achieve this with as little effort as possible? If you want to try this out, I suggest you start by installing a virtual machine like VirtualBox or VMWare. This way you can try out the configuration yourself. In this example, I'll be load-balancing 3 Tomcat servers using sticky sessions using 2 load balancers in active-passive mode. I'm assuming all 3 Tomcat servers share the same hardware configuration, so they are all able to handle the same amount of traffic each. I'm also throwing in a backup server, in case all 3 Tomcat servers go down (serving a custom 503 page kindly informing the user of a catastrophic failure, instead of dropping the standard 404 bomb). You want to start off by assigning IP addresses to the servers. This will make your life a bit easier. We'll need 7 addresses: 3 for the tomcat server, 1 for the backup server, 2 for the loadbalancers and 1 virtual IP address to be shared between the load balancers (and which will be the entry point for your users). So our assignment will be: Virtual IP 10.0.5.99 www.haci.local LB1 10.0.5.100 lb1.haci.local #MASTER LB2 10.0.5.101 lb2.haci.local #SLAVE WEB1 10.0.5.102 web1.haci.local WEB2 10.0.5.103 web2.haci.local WEB3 10.0.5.104 web3.haci.local BACKUP 10.0.5.105 backup.haci.local Setting up the web servers is easy. You just install Tomcat on each server and create a simple JSP file to be served to users (make a small change, like the background color, on each server to distinguish the servers). I won't be covering session replication between the Tomcat servers, as it'll take me too far. If you want, you can configure the appropriate session replication and storage (using multicast or JDBC for example). The backup server I'm using is a basic LAMP server that returns a simple 503 page on every request it gets. The 503 error code is important, because it reflects the current state of the system: currently unavailable. For the loadbalancers I'll be using 2 applications: HAProxy and keepalived. HAProxy is going to handle load balancing, while keepalived will handle the failover between the two load balancers. First, we're going to configure HAProxy for both LB1 and LB2. Installing HAProxy is quite easy on an ubuntu system. Just do a sudo apt-get install haproxy and you're off. After the install, backup the current HAProxy config and start editing away. cp /etc/haproxy.cfg /etc/haproxy.cfg_orig cat /dev/null > /etc/haproxy.cfg vi /etc/haproxy.cfg The content of the config to reflect our setup should become something like this (same config on LB1 and LB2): global log 127.0.0.1 local0 log 127.0.0.1 local1 notice #log loghost local0 info maxconn 4096 #debug #quiet user haproxy group haproxy defaults log global mode http option httplog option dontlognull retries 3 redispatch maxconn 2000 contimeout 5000 clitimeout 50000 srvtimeout 50000 frontend http-in bind 10.0.5.99:80 default_backend servers backend servers mode http stats enable stats auth someuser:somepassword balance roundrobin cookie JSESSIONID prefix option httpclose option forwardfor option httpchk HEAD /check.txt HTTP/1.0 server web1 10.0.5.102:80 cookie haci_web1 check server web2 10.0.5.103:80 cookie haci_web2 check server web3 10.0.5.104:80 cookie haci_web3 check server webbackup 10.0.5.105:80 backup After this, enable HAProxy on both LB1 and LB2 by editing /etc/defaults/haproxy # Set ENABLED to 1 if you want the init script to start haproxy. ENABLED=1 # Add extra flags here. #EXTRAOPTS="-de -m 16" So far for the HAProxy configuration. We can't start it up yet, as LB1 and LB2 aren't listening yet on the virtual IP address. Next we'll configure the failover of the loadbalancers using keepalived. Installing it on Ubuntu is as easy as it was for HAProxy: sudo apt-get install keepalived. But its configuration is slightly different on both load balancers. First, we need to configure the both servers to be able to listen to the shared IP address. Add the following line to /etc/sysctl.conf: net.ipv4.ip_nonlocal_bind=1 And run sysctl -p Now, we configure keepalived so that LB1 is configured as the main load balancer and binds to the shared IP address, while LB2 is on standby, ready to take over whenever LB1 goes down. The configuration for LB1 looks like this (edit /etc/keepalived/keepalived.conf): vrrp_script chk_haproxy { # Requires keepalived-1.1.13 script "killall -0 haproxy" # cheaper than pidof interval 2 # check every 2 seconds weight 2 # add 2 points of prio if OK } vrrp_instance VI_1 { interface eth0 state MASTER virtual_router_id 51 priority 101 # 101 on master, 100 on backup virtual_ipaddress { 10.0.5.99 } track_script { chk_haproxy } } Start up keepalived and check whether it is listening to the virtual IP address. /etc/init.d/keepalived start ip addr sh eth0 It should return something like this, indicating it is listening to the virtual IP address 2: eth0: mtu 1500 qdisc pfifo_fast qlen 1000 link/ether 00:0c:29:a5:5b:93 brd ff:ff:ff:ff:ff:ff inet 10.0.5.100/24 brd 10.0.5.255 scope global eth0 inet 10.0.5.99/32 scope global eth0 inet6 fe80::20c:29ff:fea5:5b93/64 scope link valid_lft forever preferred_lft forever Next, we configure LB2. The configuration is almost the same, exception for the priority. vrrp_script chk_haproxy { # Requires keepalived-1.1.13 script "killall -0 haproxy" # cheaper than pidof interval 2 # check every 2 seconds weight 2 # add 2 points of prio if OK } vrrp_instance VI_1 { interface eth0 state MASTER virtual_router_id 51 priority 100 # 101 on master, 100 on backup virtual_ipaddress { 10.0.5.99 } track_script { chk_haproxy } } Start up keepalived and check the network interface. /etc/init.d/keepalived start ip addr sh eth0 It should return something like this, indicating it is not listening to the virtual IP address 2: eth0: mtu 1500 qdisc pfifo_fast qlen 1000 link/ether 00:0c:29:a5:5b:93 brd ff:ff:ff:ff:ff:ff inet 10.0.5.101/24 brd 10.0.5.255 scope global eth0 inet6 fe80::20c:29ff:fea5:5b93/64 scope link valid_lft forever preferred_lft forever Now, start up HAProxy on both LB1 and LB2. /etc/init.d/haproxy start Now you can issue requests to 10.0.5.99 (or www.haci.local), which will go to LB1, which in turn will load-balance the request to either WEB1, WEB2 and WEB3. You can test the load balancing by turning off WEB1 (or the server you're currently on). You can also the backup server by turning all main webservers (WEB1, WEB2 and WEB3). And you can test the loadbalancer failover by turning off LB1. At that point LB2 will kick in and act as the master, loadbalancing all requests. When you turn LB1 back on, it'll take over the master role once again. HAProxy allows you to add extra servers very easily, reloading the configuration without breaking existing sessions. See the HAProxy documentation for more info or on ServerFault. (http://serverfault.com/questions/165883/is-there-a-way-to-add-more-backend-server-to-haproxy-without-restarting-haproxy). Cheap and effective. While most enterprise shops have hardware load balancers, which also have these possibilities and more, if you're on a tight budget or need to simulate a HACI environment for development purposes (a lesson here: always simulate your production environment when you're testing during development), this might be the sane option. To finish, I'll quickly explain how to set up the backup server (a simple LAMP server). Create a vhost configuration on the apache for www.haci.local or any other domain pointing to the virtual IP address and set up mod_rewrite for it: RewriteEngine On RewriteCond %{REQUEST_URI} !\.(css|gif|ico|jpg|js|png|swf|txt)$ [NC] RewriteConf %{REQUEST_URI} !/503.php RewriteRule .* /503.php [L] Then create the 503.php file and add this to the top of it: Sorry, our servers are currently undergoing maintenance. Please check back with us in a while. Thank you for your patience. You can decorate the 503.php file any way you like. You can even use CSS, JavaScript and image files in the php file. Now, back to my IDE. I'm getting withdrawal symptoms.

The concept of behavior verification consists in verifying not only the output of the System Under Test, but the calls to other components. These method calls are an output normally not visible to a caller like the test; unless he injects, instead of the real collaborator, a Test Double which can be accessed also by him. Today we will explore the first pattern for behavior verification: the Test Spy. It is a Test Double which records calls so that assertions can be made on them later in the test (after the exercise part). Spies gives us the ability to observe side-effects of the SUT, expressed as interactions with other objects. Usually behavior verification is taught with Mocks, which also verify the calls that are made on them but with specifications provided before the exercise phase. Use cases Why using a Spy instead a Mock? I bet you have heard already about Mocks at this point in the series. Here are some possible use cases on when to define Test Spies. When we cannot predict what the collaborator will be called with (if we can, we may use a Mock instead). When an assertion on calls is complex to define beforehand, or needs all (or more than one of) the calls to be completed before taking place. The matchers used for Mocks are not sufficient for verification. When a Mock that immediately threw an exception would only result in the SUT swallowing it. So a Spy is better in this case as it leaves verification for later. PHPUnit matchers in general do not always throw an exception or always wait for the end of test (it depends on the particular constraint.) The behavior involves more than one collaborator (it happens sometimes). In this case, you can make an assertion using the data recorded by more than one Spy. Implementation 1. Like with all Test Doubles, provide an alternate implementation or subclass. Probably the calls to the Spy won't return anything (or will return something canned to avoid the SUT's failure, like in a Stub). The methods implementations will just record all calls, or some property of the calls like the first parameter or their total number. 2. Inject the Test Double and exercise the SUT. 3. Make assertions on the Test Double data gathered by the Spy. Variations The variations of the pattern are mainly on how to retrieve the data from the Test Spy. Retrieval Interface: the Test Spy is a class with additional methods, or public properties (eek) that expose the recorded data. This variation cannot be coded with PHPUnit generation of Test Doubles, only by hand-rolling them. Self Shunt: the Test Spy and Test Case Object are a single object. This means we inject $this as the Test Double and we have the maximum freedom of defining new methods and access the recording. The caveat is that it can only be done with interfaces in PHPUnit, because Test Case Classes must extend PHPUnit_Framework_TestCase. That's a good reason to extract an interface, though. Inner Test Double: we inject a private class (not existent in PHP) or a closure which records everything. A closure for example can access $this public properties or methods, or some other local ArrayObject passed by handler (or variable passed by reference) where the data can be kept. Indirect Output Registry: same as Inner Test Double, but the target for recordings is a full-fledged object. Almost always an overkill. Examples The sample code shows you how to implement a Spy, in its different variations and in use cases where it actually make sense. Most of the times, Mock are used instead and the pattern is equivalent. Test Spies are a little more difficult to write, but they are invaluable in the case where your verification logic does not fit the framework of Mocks (predefined, single-object expectations). createUser(array('mail' => '[email protected]', 'nickname' => 'johndoe')); $this->assertEquals(array('executeQuery', 'mail'), $this->order); } private $order = array(); private $queries = array(); public function executeQuery($query, array $params) { $this->order[] = 'executeQuery'; $this->queries[] = $query; } private $mails = array(); public function mail($to, $subject, $object) { $this->order[] = 'mail'; $this->mails[] = array('to' => $to, 'subject' => $subject, 'object' => $object); } public function testInnerTestDoubleArrayObject() { $parts = new ArrayObject(); $receiver = $this->getMock('Receiver'); $receiver->expects($this->any()) ->method('definePart') ->will($this->returnCallback(function($amount) use ($parts) { $parts[] = $amount; })); $sut = new RandomDivider($receiver); $sut->divide(10); $this->assertEquals(10, array_sum($parts->getArrayCopy())); } public function testInnerTestDoubleArrayPassedByReference() { $parts = array(); // an array would not be passed by handler by default $receiver = $this->getMock('Receiver'); $receiver->expects($this->any()) ->method('definePart') ->will($this->returnCallback(function($amount) use (&$parts) { // but we can pass it by reference $parts[] = $amount; })); $sut = new RandomDivider($receiver); $sut->divide(10); $this->assertEquals(10, array_sum($parts)); } } interface Mailer { public function mail($to, $subject, $object); } interface Db { public function executeQuery($query, array $params); } class UserDao { private $db; private $mailer; public function __construct(DB $db, Mailer $mailer) { $this->db = $db; $this->mailer = $mailer; } public function createUser(array $userDetails) { // internally it would use PDO $this->db->executeQuery("INSERT INTO users ...", $userDetails); $this->mailer->mail($userDetails['mail'], 'You have been registered on example.com', '...'); } } interface Receiver { public function definePart($amount); } class RandomDivider { private $receiver; public function __construct(Receiver $receiver) { $this->receiver = $receiver; } public function divide($total) { $part = ceil(rand() * $total); $this->receiver->definePart($part); $this->receiver->definePart($total - $part); } }

Creating a statically linked version of a unix software saves you in all the circumstances where you can’t install the software as root, when you don’t have development tools on the target machine, when you cannot find a prepackaged version for the specific server, when the libraries available on a server are conflicting with the ones required by your software, etc. Examples: The unix server that hosts this blog, provides me a restricted shell to manage mysql and public web files with minimal access privileges. On this server I periodically update WordPress and other things, and I like to keep the changes under control and backed up with git. The problem is that on this server I haven’t git binaries, and I cannot install software and libraries required. Having the files of my WordPress installation under git version control saved me a lot of time in the past when somebody hacked into the server and changed the php files to do nasty stuff. I was able to easily check what was exactly changed since last legitimate update with “git status” command, and restore things as before. I own a little NAS (network attached storage) which runs a mini distribution of Linux which doesn’t have a prepackaged version of git to install, nor I cannot use any package management tool like apt rpm or yum etc. So I thought that I can compile by myself the git binaries and have them deployed on the target machine. It is possible, but there is an important note: when you build a software on a unix server, the resulting binaries are referencing the libraries installed there, so you cannot easily port the binaries between servers since they have dependencies. What you need is a “statically linked” version of the software, which fortunately it’s not so difficult to achieve. Binaries which are statically linked are usually bigger in size and will possibly require more memory to execute, but they won’t require the specific libraries to be present on the executing computer, since the libraries code is contained in the binaries themselves. Here is how I built a static version of git on an ubuntu virtual machine, that can be ported to other unix servers: # let's make sure we have all we need to proceed $ sudo apt-get install libexpat1-dev asciidoc libz-dev gettext curl # let's create the directory to host the built artifacts $ sudo mkdir /opt/git-1.7.4.1-static # we are ready to download and unpack latest version of git sources $ curl http://kernel.org/pub/software/scm/git/git-1.7.4.1.tar.bz2 | tar xvj $ cd git-1.7.4.1/ # then compile and install the files in the target directory we created $ ./configure --prefix=/opt/git-1.7.4.1-static CFLAGS="${CFLAGS} -static" NO_OPENSSL=1 NO_CURL=1 $ sudo make install $ sudo make install-doc On the above commands, the thing to notice is the CFLAGS=”${CFLAGS} -static” which is used to specify that the libraries must be statically linked with the binaries. The last thing to do is create a tarball of /opt/git-1.7.4.1-static folder and copy that on the target machine; adding the /opt/git-1.7.4.1-static/bin directory to the PATH variable and the /opt/git-1.7.4.1-static/share/man to the MANPATH variable. If possible, it’s a good idea to keep the same installation path on target machine (/opt/git-1.7.4.1-static) since this path gets hardcoded in some git scripts during the build process. But it shouldn’t give too many problems, anyway. From http://en.newinstance.it/2011/02/27/how-to-create-a-statically-linked-version-of-git-binaries/

While we have been evaluating in our blog posts the various features available on popular Cloud Computing platforms today, I thought it might be a good idea to understand when and how all this started and look back at where this began and trace some of the key events in the progress of cloud computing. Amazon like all other Internet companies in the period of the dot com bubble were left with large amounts of underutilized computing infrastructure, reports suggest less than 10% of the server infrastructure of many companies were being used. Amazon may have use cloud computing as a way to provide this unused resources as utility computing service when they launched S3 as the first true cloud computing service in March 2006. 1. Launch of Amazon Web Services in July 2002 The initial version of AWS in 2002 was focused more on making information available from Amazon to partners through a web services model with programmatic and developer support and was very focused on Amazon as a retailer. While this set the stage for the next steps the launch of S3 was the true step towards building a cloud platform. Amazon Press Release 2. S3 Launches in March 2006 Here are some interesting articles on the launch of S3 in 2006. The real breakthrough however was the pricing model for S3 which defined the model of 'pay-per-use' which has now become the defacto standard for cloud pricing. Also the launch of S3 really defined the shift of Amazon from being just a retailer to a strong player in the technology space. Techcrunch Post on S3 on March 14th, 2006 Read Write Web Post on S3 and EC2 on Nov 3rd, 2006 Business Week Article on Jeff Bezos vision on cloud computing on Nov 13th, 2006 3. EC2 Launches in August 2006 EC2 had a much quieter launch in August 2006 but i would think had the bigger impact by making core computing infrastructure available. This completed the loop on enabling a more complete cloud infrastructure being available. In fact at that time analysts had some difficulty in understanding what the big deal is, and thought it looks similar to other hosting services available online only with a different pricing model. Some interesting articles from that time on the launch: Technologyevangelist Blog Virtualization Info 4. Launch of Google App Engine in April 2008 The launch of Google App Engine in 2008 was the entry of the first pure play technology company into the Cloud Computing market. Google a dominant Internet company entering into this market was clearly a major step towards wide spread adoption of cloud computing. As with all their other products they introduced radical pricing models with a free entry level plan and extremely low cost computing and storage services which are currently among the lowest in the market. Techcrunch post on App Engine Launch Google App Engine Launch Post 5. Windows Azure launches Beta in Nov 2009 The entry of Microsoft into Cloud Computing is a clear indication of the growth of the space. Microsoft for long has not accepted the Internet and the web as a significant market and has continued to focus on the desktop market for all these years. I think this is a realization that a clear shift is taking place. The launch of Azure is a key event in the history of cloud computing with the largest software company making a small but significant shift to the web. Launch of Azure Beta Azure General Availability - Feb 2010 You might also like: Cloud Computing, Google App Engine: How big is the market Really ? Comparing Google App Engine with Amazon EC2 Comparing Amazon EC2 and Microsoft Azure Languages Supported by Google App Engine Cloud Computing: What is it really ?

This is ongoing blog on Getting Started with Android. In earlier blog, I provided an architecture overview of android application, followed by setting up the development environment for Android and creating and running a simple application. In this blog, I will describe how to invoke web services (soap based services) via Android. In my next blog, I will follow it up with how to invoke REST based services. For trying out the tutorial, you need to have the android development environment setup as mentioned in my previous blog. There are two ways in which invoke web services Raw APIs : Use the HttpClient and XML parser to manually create a soap request and parse the soap response. Using a soap client library : like KSOAP library which does the low level work for parsing and dealing with soap messages – For Android, there is library available at http://code.google.com/p/ksoap2-android/ . Its good to see some active development for KSOAP 2, I remembered I wrote the first article on KSOAP 2 way back in 2003 ( http://naveenbalani.com/index.php/2010/05/deliver-web-services-to-mobiles/)and good to see it back in development for android. I would start development with the later approach, but I plan to use RAW APIs in the follow up post - Download the KSOAP2 library , go to http://code.google.com/p/ksoap2-android/ , click on downloads link on the menu, and download the latest release artifact – http://code.google.com/p/ksoap2-android/source/browse/m2-repo/com/google/code/ksoap2-android/ksoap2-android-assembly/2.5.2/ksoap2-android-assembly-2.5.2-jar-with-dependencies.jar . In the release artifact page, click on “View raw file” and select “Save Link as” and download the jar file which has all the required dependencies. Next we would create a sample android project which would invoke a .NET web service. I decided to host a simple .NET web service in my website , so it would easier for you all to try out the sample . The web service is available at http://naveenbalani.com/WassupAndroid.asmx This is a simple .NET service, with one operation called todayMessage(), which display “Wassup Android from a .NET application “ as output. To create an andrioid project. Start eclipse. Select File > New > Project. Select Android > Android Project, Click Next. Enter the following information for the project - Project name – AndroidClientService Build Target – Android 2.3 Application name – WasuppTodaysMessage Package name – org.android.websevice.client.samples Create Activity – AndroidClientService Min SDK Version – 9 Click Finish This would create a Project called AndroidClientService in your workspace. Next , add the ksoap2-andriod dependency to the project. Select the AndroidClientService, click properties , click on Java build path , click on Libraries , select Add External Jars and add the ksoap2 library (ksoap2-android-assembly-2.5.2-jar-with-dependencies.jar) and click Ok. Next, open up the WasuppServiceClientAndroid class and replace the onCreate method with the following onCreate() method as shown in listing below. Following shows the complete code listing. This project would invoke the web service and display – “ “ on the device when the application is executed. To build the project, select Project -> Clean package android.websevice.client.samples; import org.ksoap2.SoapEnvelope;import org.ksoap2.serialization.SoapObject;import org.ksoap2.serialization.SoapSerializationEnvelope;import org.ksoap2.transport.HttpTransportSE; import android.app.Activity;import android.os.Bundle;import android.widget.TextView; public class AndroidClientService extends Activity { private static final String SOAP_ACTION = "http://www.naveenbalani.com/webservices/WassupAndroidService/todaysMessage"; private static final String OPERATION_NAME = "todaysMessage"; private static final String WSDL_TARGET_NAMESPACE = "http://www.naveenbalani.com/webservices/WassupAndroidService/"; private static final String SOAP_ADDRESS = "http://naveenbalani.com/WassupAndroid.asmx"; @Overridepublic void onCreate(Bundle savedInstanceState) {super.onCreate(savedInstanceState); TextView textView = new TextView(this); setContentView(textView); SoapObject request = new SoapObject(WSDL_TARGET_NAMESPACE,OPERATION_NAME); SoapSerializationEnvelope envelope = new SoapSerializationEnvelope(SoapEnvelope.VER11);envelope.dotNet = true; envelope.setOutputSoapObject(request); HttpTransportSE httpTransport = new HttpTransportSE(SOAP_ADDRESS); try { httpTransport.call(SOAP_ACTION, envelope); Object response = envelope.getResponse(); textView.setText(response.toString()); } catch (Exception exception) { textView.setText(exception.toString()); } } To run the AndroidClientService Android application, click on it and select Run As > Android Application. On the eclipse console, you would see the following similar message – [AndroidClientService] Performing android.websevice.client.samples.AndroidClientService activity launch [AndroidClientService] Automatic Target Mode: launching new emulator with compatible AVD ‘AVD’ [AndroidClientService] Launching a new emulator with Virtual Device ‘AVD’ [AndroidClientService] Waiting for HOME (‘android.process.acore’) to be launched… You should see the Android AVD being launched. After the above message, it takes a while (2-3 minutes) for the first time to get the Android home page on the emulator. After the device is started, you should see the following message on console.. [AndroidClientService] Uploading AndroidClientService.apk onto device ‘emulator-5554′ [AndroidClientService] Installing AndroidClientService.apk… [AndroidClientService] Success! [AndroidClientService] Starting activity android.websevice.client.samples.AndroidClientService on device emulator-5554 If the application doesn’t show up on the emulator, Click on Menu option on the emulator and you would see the WasuppTodayMessage android application and message being displayed. Issues encountered during invoking the web services application from Android Emulator Unknown host exception – If you get the following exception – “java.net.UnKnownHostException: naveenbalani.com’, than you need to add required domain name server which emulator would use to resolve domain. A list of network limitations on emulator is available at – http://developer.android.com/guide/developing/tools/emulator.html#networkinglimitations As per the documentation – “ “At startup, the emulator reads the list of DNS servers that your system is currently using. It then stores the IP addresses of up to four servers on this list and sets up aliases to them on the emulated addresses 10.0.2.3, 10.0.2.4, 10.0.2.5 and 10.0.2.6 as needed. On Linux and OS X, the emulator obtains the DNS server addresses by parsing the file /etc/resolv.conf. On Windows, the emulator obtains the addresses by calling the GetNetworkParams() API. Note that this usually means that the emulator ignores the content of your “hosts file” Now, to add the domain name server, click on Run configurations and select AndroidClientService and add the following -dns-server ns15.unitechost.in in the additional emulator command line options as shown below. Click Run to run the configuration Security If you get a permission issue while accessing internet, you need to add the following line in to allow application to access internet Here is the complete listing of AndroidManifest.xml From http://naveenbalani.com/index.php/2011/01/invoke-webservices-from-android/

If you create small per-test datasets, as DBUnit advises, you’ll get intermittent build failures due to foreign-key violations. This post explains (1) why this happens, (2) why small per-test datasets are still a good idea, and (3) one simple way to get around the problem. NB When I searched for solutions to this problem, I discovered that other kinds of foreign-key problem come up with DBUnit. Some people have circular dependencies in their relational database schemas, which stops DBUnit from loading the test data. If such is your case, I’m sorry to say that this post won’t help you with it, and your best option is probably to just take yourself outside and shoot yourself now. (Although some people seem to chosen instead to disable foreign key checking during test runs.) What causes the foreign-key violations The cause of the problem is simple, and illustrated by a trivial example. Suppose you have two entity classes, HitchHiker and SpaceShip. The HitchHiker table has a foreign key that references SpaceShip. The test data for HitchHikerDaoTest contains lines from both tables, whereas the test data for SpaceShipDaoTest contains only lines from SpaceShip. DBUnit’s default setup operation, CLEAN_INSERT, wipes data from every table occurring in the test dataset and then inserts the lines listed in that dataset. When SpaceShipDaoTest runs, DBUnit will start by deleting everything in the SpaceShip table. If any HitchHikers are currently riding in the SpaceShips that are about to be deleted, the database will object to their untimely eviction (I’m not sure whether the error message will read like Vogon poetry, though). If you start from an empty database, and execute SpaceShipDaoTest and then HitchHikerDaoTest, you’ll be fine; but if you do it in the other order, your build will fail. It’s that second-worst kind of bug, the unpredictable kind, since you don’t (usually) specify the order in which tests run. After all, they’re supposed to be independent! So you may well find that you have no problems for months on end, until one day you get an error running individual tests in a particular sequence, or Maven changes the order in which it runs your tests on the CI server, and BOOM! Why you should still use small independent datasets It’s tempting to circumvent the problem by using a single monolithic dataset for all your integration tests. I’ve tried this, and I advise against it. A big data file is hard to work with: you waste a lot of time scrolling around looking for the line you need, and it’s very hard to follow and understand foreign-key relations. Worse still: by modifying the data to make one test pass, you can easily accidentally break another one. The larger the dataset and the test suite become, the more fragile they get, and the more painstaking it becomes to modify them. How to avoid the foreign-key problem with small independent datasets One working but unsatisfactory solution would be to pad out every XML dataset with the list of all tables touched in the test suite. It’s unsatisfactory because the only way to add a table into a FlatXmlDataSet is to list a line of that table — a FlatXmlDataSet can’t contain empty tables — and there’s no justification for polluting the test data with lines from tables that are not part of the test. The solution I found was to use a DTD to clean tables before tests. Every XML file has different contents, but they all reference a single DTD which lists all the tables involved in the test suite. The DTD is easy to generate from the database schema, and useful for auto-complete and catching typos in column names, so you should probably already be using one. The code to exploit its contents is very simple: private IDataSet loadTestDataWithDtdTableList(String dtdFilename) throws IOException, DataSetException, SQLException { Reader dtdReader = new FileReader(new ClassPathResource(dtdFilename).getFile()); IDataSet dtdDataset = new FlatDtdDataSet(dtdReader); FlatXmlDataSetBuilder builder = new FlatXmlDataSetBuilder(); builder.setMetaDataSet(new DatabaseDataSet(dbUnitConnection, false)); IDataSet xmlDataset = builder.build(asFile(xmlFilename)); return new CompositeDataSet(dtdDataset, xmlDataset);} How it works: DBUnit provides a facility to load a dataset from a DTD. This dataset contains all the tables listed in the DTD, but of course empty of data. The DTD dataset is then combined with a FlatXmlDataSet representing your test data. The graphic below illustrates the composite dataset that would be produced for the SpaceShip example. If you have dictionary tables whose contents never change, you can and should leave them out of the DTD as well as out of the XML datasets, to improve test performance a little. One further detail: you should close the FileReader after test setup. I couldn’t find a hook into the end of the test setup operation (short of writing my own DatabaseOperation), so I saved the reference as a member variable and hooked the close() call into the tear-down phase of the test. NB For a more complete code example, see this Gist snippet of a base class for TestNG+Spring+DBUnit tests that adds the above-described DBUnit setup operation to Spring’s TestNG helper class. Happy database testing! From http://www.andrewspencer.net/2011/solve-foreign-key-problems-in-dbunit-test-data/

In order to perform tests, you create fixtures such as the System Under Test and input or output data. Usually this fixtures are just objects. Garbage collection, a mechanism present in many languages, deletes objects and variables (and thus fixtures) when there are no references for them anymore in the available scopes. The logic behind garbage collection is that objects which are not reachable anymore by any other object in the local or global scope are as good as garbage. Their code can't possibly be executed as you cannot call a method on them without a reference. Basically, if you do not put your fixtures in strange places, they will be automatically deleted with the test case. While a setUp() method is very common, a tearDown() method is usually not necessary. Implementation References like local variables or fields on $this (at the Testcase Object level) are deleted with the test case object. Currently, the PHPUnit manual say that the gabage collection of Testcase Objects is not predictable, so you may have to perform an unset() over $this->field to have the object removed. Static references are special: they are always reachable, as they are kept on classes (the current or other ones) instead of objects, and won't be deleted. Even if an object is pointed by another object in a static reference, it won't be garbage-collected for transitivity. Also singletons are a big problem (which derives from the static fields where their instances are kept): if you put something in Zend_Registry or in a singleton, it won't be deleted after a test. That's one reason singletons are dangerous: it's not only a matter of performance, but of them being capable of bringing objects from one test to another, spoiling our pursue of test isolation. PHP garbage collection Up to PHP 5.2, objects with mutual references won't usually be collected, even if neither of them was reachable from the other scopes. This is not a problem for a script that produces a web page and exits, but it can be for a test suite which runs for some minutes when at full capacity. In fact, an example of problematic behavior is the ezComponents test suite, which is cited on php.net. The suite would require several gigabytes of memory to run, due to the leaks of the long-running PHP process. From PHP 5.3, if zend.enable_gc is enabled (by default, it is) cycles will be collected when the table of roots is full. So this particular issue (circular references) won't worry you anymore. In case you encounter erratic behavior, see if you can change the php.ini directive memory_limit to detect the problem, and use the native function memory_get_usage() to gather statistics. Examples In this code sample, I show you tests that create fixtures which are automatically collected. How do we prove that these objects are really deleted? With a simple __destruct() method. Keep in mind that explicit teardown is usually not necessary: you should adopt it only if you have an heavy fixture, which you want to be gargabe-collected before the end of the PHP process. Running this test case: deletedFixture = new SomeFixture(1); $this->abandonedFixture = new SomeFixture(2); } public function testFirst() { } public function testSecond() { } public function tearDown() { // this fixture will be garbage-collected at the end of each test unset($this->deletedFixture); // since we do not touch $this->abandonedFixture, its collection // is not predictable. It can happen at any time after the tests execution. } } class SomeFixture { private $number; public function __construct($number) { $this->number = $number; } public function __destruct() { echo "Cleaning up fixture $this->number.\n"; } } gives this output: [13:11:15][giorgio@Desmond:~]$ phpunit code/practical-php-testing-patterns/GarbageCollectedTeardownTest.php PHPUnit 3.5.5 by Sebastian Bergmann. Cleaning up fixture 1. .Cleaning up fixture 1. . Time: 0 seconds, Memory: 4.75Mb OK (2 tests, 0 assertions) Cleaning up fixture 2. Cleaning up fixture 2.

DataFactory is a project I just released which allows you to easily generate test data. It was primarily written for populating database for dev or test environments by providing values for names, addresses, email addresses, phone numbers, text, and dates. To add DataFactory to your maven project, just add it as a dependency in your pom.xml file. org.fluttercode.datafactory datafactory 0.8 jar Generating Test Data Now you can create instances of the DataFactory class and create data : public class Main { public static void main(String[] args) { DataFactory df = new DataFactory(); for (int i = 0; i < 100; i++) { String name = df.getFirstName() + " "+ df.getLastName(); System.out.println(name); } } } The produced output is : Lindsey Craft Erica Larsen Ryan Levine Erika Smith Brooklyn Sloan Karen Mayer Eddie O'neill Nancy Stevens The DataFactory class can generate different types of values, from addresses to random text to random dates, to dates within a fixed time period. Addresses and business names can be created using the following code : DataFactory df = new DataFactory(); for (int i = 0; i < 100; i++) { String address = df.getAddress()+","+df.getCity()+","+df.getNumberText(5); String business = df.getBusinessName(); System.out.println(business + " located at " + address); } to produce Uvalda Signs located at 1383 Beam Way,Lyons,19316 Alma Accounting located at 1386 Countiss St,Nashville,14967 Fort Stewart Engineering located at 1753 Bethesda Rd,Springfield,26306 Sugar Hill Textiles located at 1141 Loudon Circle,Cordele,83937 Albany Engineering located at 1185 Grieves Avenue,Sugar Hill,36753 Poulan Insurance located at 816 Cohen Blvd,Lake City,74839 Crescent Services located at 1085 Cloveridge Boulevard,Bemiss,08769 Dates There are a number of features to create dates, the first being creating a random date which is usually in a given sensible date range. DataFactory df = new DataFactory(); Date minDate = df.getDate(2000, 1, 1); Date maxDate = new Date(); for (int i = 0; i < 10; i++) { Date start = df.getDateBetween(minDate, maxDate); System.out.println("Date = "+start); } This produces a list of random dates between 1/1/2000 and the current date. Typically, a random date might be constrained by some other date, for example you can’t have an end date that occurs before the start date. In this case, you would plug the start date in as the minimum date value : DataFactory df = new DataFactory(); Date minDate = df.getDate(2000, 1, 1); Date maxDate = new Date(); for (int i = 0; i < 10; i++) { Date start = df.getDateBetween(minDate, maxDate); Date end = df.getDateBetween(start, maxDate); System.out.println("Date range = " + dateToString(start) + " to " + dateToString(end)); } The result is a list of dates where the second date is always later than the first : Date range = 04/29/2005 to 07/16/2006 Date range = 08/07/2009 to 01/19/2010 Date range = 09/22/2000 to 12/15/2003 Date range = 07/31/2004 to 03/24/2009 Date range = 06/27/2003 to 01/10/2007 Date range = 07/10/2003 to 04/02/2008 Date range = 01/04/2003 to 01/12/2005 In many cases, you might want your end date to be only within a few days of the start date. For example, helpdesk support tickets or hotel stays don’t last for years. To do this, you can specify the number of days from the base date you want to generate a result. In this case, we make the end date within 10 days of the begin date : for (int i = 0; i < 10; i++) { Date start = df.getDateBetween(minDate, maxDate); Date end = df.getDate(start, 0, 10); //set end to within 10 days of the start System.out.println("Date range = " + dateToString(start) + " to " + dateToString(end)); } And the result : Date range = 04/29/2005 to 04/30/2005 Date range = 12/29/2003 to 12/30/2003 Date range = 06/25/2003 to 07/03/2003 Date range = 10/19/2009 to 10/19/2009 You can also specify a negative minimum days value that could return a date prior to the base date or a positive minimum date value to get a later date. Here’s a more complex example that uses different date rules to come up with some complex test data. for (int i = 0; i < 10; i++) { //generate an order date Date orderDate = df.getDateBetween(minDate, maxDate); //estimate delivery 4-10 days after ordering Date estimatedDeliveryDate = df.getDate(orderDate, 4, 10); //deliver between 2 days prior and 3 days after delivery estimate Date actualDeliveryDate = df.getDate(estimatedDeliveryDate, -2, 3); String msg = "Ordered on "+dateToString(orderDate) + " deliver by = "+dateToString(estimatedDeliveryDate)+ " delivered on " + dateToString(actualDeliveryDate); if (estimatedDeliveryDate.before(actualDeliveryDate)) { msg = msg + " - LATE"; } if (estimatedDeliveryDate.after(actualDeliveryDate)) { msg = msg + " - EARLY"; } System.out.println(msg); } Here we calculate an order date, and create a delivery date that is at least 4 days out but no more than 10, and then we created an actual delivery date that is between 2 days prior and 3 days after the expected delivery date. Notice how we cherry picked the dates, the estimated delivery date is at least 4 days out from the order date, and the actual delivery date will only be at most 2 days prior to the estimated date. This means the actual delivery date is always at least 2 days out from the order date and we won’t get a delivery date value that is before the item was ordered. This code produces the following values: Ordered on 04/29/2005 deliver by = 05/06/2005 delivered on 05/06/2005 Ordered on 08/07/2009 deliver by = 08/13/2009 delivered on 08/13/2009 Ordered on 09/22/2000 deliver by = 09/27/2000 delivered on 09/25/2000 - EARLY Ordered on 07/31/2004 deliver by = 08/07/2004 delivered on 08/09/2004 - LATE Ordered on 06/27/2003 deliver by = 07/04/2003 delivered on 07/04/2003 Ordered on 07/10/2003 deliver by = 07/19/2003 delivered on 07/18/2003 - EARLY Ordered on 01/04/2003 deliver by = 01/08/2003 delivered on 01/08/2003 Custom Random Values If there is a set of values that is very specific to your application that you might want to generate data from, you can use methods on the DataFactory class to return values with the option for it to be randomly be a default value. public static void main(String[] args) { DataFactory df = new DataFactory(); //favorite animal String[] values = {"Cat","Dog","Goat","Horse","Sheep"}; for (int i = 0; i < 100; i++) { System.out.println(df.getItem(values,80,"None")); } } This example uses the array of animals and returns a value with a 20% chance of being the default value of “None” to produce the following : Sheep None Dog Horse Textual Data Random text data comes in two forms, absolutely random data and text data made up of words. You can generate either using the following methods : DataFactory df = new DataFactory(); System.out.println(df.getRandomText(20, 25)); System.out.println(df.getRandomChars(20)); System.out.println(df.getRandomWord(4, 10)) which produces badly numbers good hot I ywyypgqorighfawpftjq demanded All three of these methods can be passed a single length which returns a fixed length string, or a min/max length which produces a random string with a length somewhere between the min/max. For the single word method, if there are no words in the dictionary of suitable length, then a word is generated using random characters. Changing the test data values produced The data used to generate the values come from classes that can be replaced with other versions. For example, the name values can be changed by providing the DataFactory instance with an object that implements the NameDataValues interface. Here is a simple class that does that to return Scandinavian first names and delegates to the the default implementation to return all the other values. public class ScandinavianNames implements NameDataValues { //first name values to use String[] firstNames = {"Anders","Freydís","Gerlach","Sigdis"}; //delegate to the default implementation for the other values NameDataValues defaults = new DefaultNameDataValues(); public String[] getFirstNames() { //return our custom list of names return firstNames; } //for the other values, just use the defaults public String[] getLastNames() { return defaults.getLastNames(); } public String[] getPrefixes() { return defaults.getPrefixes(); } public String[] getSuffixes() { return defaults.getSuffixes(); } } Obviously, to use all your own names you would add and return values for last name and the suffix/prefix values. To use this new implementation, just create an instance of the data provider and pass it to the instance of the data factory. public static void main(String[] args) { DataFactory df = new DataFactory(); df.setNameDataValues(new ScandinavianNames()); for (int i = 0; i < 10; i++) { System.out.println(df.getName()); } } Our results are Sigdis Craft Gerlach Larsen Sigdis Levine Sigdis Smith Freydís Sloan Gerlach Mayer You can always start working with the default implementation and use a more locale specific implementation if you need it later. The different pieces that can be replaced are as follows : NameDataValues – Generates names and suffix/prefixes ContentDataValues.java – Generates words, business types, email domain names and top level domain values AddressDataValues – Generates city names, street names and address suffixes Note that if you intend on replacing the component that generates words, you should have a good collection of words of various lengths from 2 up to say 8 or more characters. Hopefully this will give you a head start in generating data in development and test environments for new projects. Now I have DataFactory in the Central Maven Repository I plan on using this in the Knappsack archetypes rather than hard coding the data which was in fact generated from an earlier DataFactory implementation. From http://www.andygibson.net/blog/article/generate-test-data-with-datafactory/

If you have worked with JUnit Rule API which was introduced in version 4.8 then you might probably also think that they are very useful. For example, there is a Rule called TemporaryFolder which creates files and folder before test is executed and deletes them after the test method finishes(whether test passes or fails). For those who are not familiar with Rule API , a Rule is an alteration in how a test method, or set of methods, is run and reported. Before version 4.9 Rule can only be applied to a test method not to a Test class or JUnit test suite. But with JUnit 4.9 which is not yet released you can use @ClassRule annotation to apply a rule to test class or a test suite. If you want to use JUnit 4.9 download it from JUnit git repository. The @ClassRule annotation can be applied to any public static field of Type org.junit.rules.TestRule. The class level rule can be applied in scenarios where you use normally use @BeforeClass and @AfterClass annotation. Some of the scenarios can be like when you want to start a server or any other external resource before a test class or test suite, or when you want to make sure that your test suite or test class runs within a specified time, etc. The advantage of using class level rule is that they can be reused among different modules and classes. Let's take an example when we want to make sure that our test suite should run within x seconds otherwise test should timeout. As you can see below we TestSuite AllTests runs TestCase1 and TestCase2. I have defined a suite level rule which would make sure that test run in three seconds otherwise suite will fail. Test Suite @RunWith(Suite.class) @SuiteClasses({ TestCase1.class, TestCase2.class }) public class AllTests { @ClassRule public static Timeout timeout = new Timeout(3000); } TestCase 1 public class TestCase1 { @Test public void test1() throws Exception{ Thread.sleep(1000); Assert.assertTrue(true); } } TestCase2 public class TestCase2 { @Test public void test2() throws Exception{ Thread.sleep(1000); Assert.assertTrue(true); } } This is the most important feature which will be released in version 4.9 of JUnit.

I've blogged a few times now about Git (which I pronounce with a hard 'g' a la "get", as it's supposed to be named for Linus Torvalds, a self-described git, but which I've also heard called pronounced with a soft 'g' like "jet"). Either way, I'm finding it way more efficient and less painful than either CVS or SVN combined. So, to continue this series ([1], [2], [3]), here is how (and why) to pull an SVN repo down as a Git repo, but with the omission of old (irrelevant) revisions and branches. Using SVN for SVN repos In days of yore when working with the JBoss Tools and JBoss Developer Studio SVN repos, I would keep a copy of everything in trunk on disk, plus the current active branch (most recent milestone or stable branch maintenance). With all the SVN metadata, this would eat up substantial amounts of disk space but still require network access to pull any old history of files. The two repos were about 2G of space on disk, for each branch. Sure, there's tooling to be able to diff and merge between branches w/o having both branches physically checked out, but nothing beats the ability to place two folders side by side OFFLINE for deep comparisons. So, at times, I would burn as much as 6-8G of disk simply to have a few branches of source for comparison and merging. With my painfullly slow IDE drive, this would grind my machine to a halt, especially when doing any SVN operation or counting files / disk usage. Using Git for SVN repos naively Recently, I started using git-svn to pull the whole JBDS repo into a local Git repo, but it was slow to create and still unwieldy. And the JBoss Tools repo was too large to even create as a Git repo - the operation would run out of memory while processing old revisions of code to play forward. At this point, I was stuck having individual Git repos for each JBoss Tools component (major source folder) in SVN: archives, as, birt, bpel, build, etc. It worked, but replicating it when I needed to create a matching repo-collection for a branch was painful and time-consuming. As well, all the old revision information was eating even more disk than before: jbosstools' trunk as multiple git-svn clones: 6.1G devstudio's trunk as single git-svn clone: 1.3G So, now, instead of a couple Gb per branch, I was at nearly 4x as much disk usage. But at least I could work offline and not deal w/ network-intense activity just to check history or commit a change. Still, far from ideal. Cloning SVN with standard layout & partial history This past week, I discovered two ways to make the git-svn experience at least an order of magnitude better: Standard layout (-s) - this allows your generated Git repo to contain the usual trunk, branches/* and tags/* layout that's present in the source SVN repo. This is a win because it means your repo will contain the branch information so you can easily switch between branches within the same repo on disk. No more remote network access needed! Revision filter (-r) - this allows your generated Git repo to start from a known revision number instead of starting at its birth. Now instead of taking hours to generate, you can get a repo in minutes by excluding irrelevant (ancient) revisions. So, why is this cool? Because now, instead of having 2G of source+metadata to copy when I want to do a local comparison between branches, the size on disk is merely: jbosstools' trunk as single git-svn clone w/ trunk and single branch: 1.3G devstudio's trunk as single git-svn clone w/ trunk and single branch: 0.13G So, not only is the footprint smaller, but the performance is better and I need never do a full clone (or svn checkout) again - instead, I can just copy the existing Git repo, and rebase it to a different branch. Instead of hours, this operation takes seconds (or minutes) and happens without the need for a network connection. Okay, enough blather. Show me the code! Check out the repo, including only the trunk & most recent branch # Figure out the revision number based on when a branch was created, then # from r28571, returns -r28571:HEAD rev=$(svn log --stop-on-copy \ http://svn.jboss.org/repos/jbosstools/branches/jbosstools-3.2.x \ | egrep "r[0-9]+" | tail -1 | sed -e "s#\(r[0-9]\+\).\+#-\1:HEAD#") # now, fetch repo starting from the branch's initial commit git svn clone -s $rev http://svn.jboss.org/repos/jbosstools jbosstools_GIT Now you have a repo which contains trunk & a single branch git branch -a # list local (Git) and remote (SVN) branches * master remotes/jbosstools-3.2.x remotes/trunk Switch to the branch git checkout -b local/jbosstools-3.2.x jbosstools-3.2.x # connect a new local branch to remote one Checking out files: 100% (609/609), done. Switched to a new branch 'local/jbosstools-3.2.x' git svn info # verify now working in branch URL: http://svn.jboss.org/repos/jbosstools/branches/jbosstools-3.2.x Repository Root: http://svn.jboss.org/repos/jbosstools Switch back to trunk git checkout -b local/trunk trunk # connect a new local branch to remote trunk Switched to a new branch 'local/trunk' git svn info # verify now working in branch URL: http://svn.jboss.org/repos/jbosstools/trunk Repository Root: http://svn.jboss.org/repos/jbosstools Rewind your changes, pull updates from SVN repo, apply your changes; won't work if you have local uncommitted changes git svn rebase Fetch updates from SVN repo (ignoring local changes?) git svn fetch Create a new branch (remotely with SVN) svn copy \ http://svn.jboss.org/repos/jbosstools/branches/jbosstools-3.2.x \ http://svn.jboss.org/repos/jbosstools/branches/some-new-branch From http://divby0.blogspot.com/2011/01/howto-partially-clone-svn-repo-to-git.html

we've all seen them on various websites. crappy search utilities. they are a constant reminder that search is not something you should take lightly when building a website or application. search is not just google's game anymore. when a java library called lucene was introduced into the apache ecosystem, and then solr was built on top of that, open source developers began to wield some serious power when it came to customizing search features. in this article you'll be introduced to apache solr and a wealth of applications that have been built with it. the content is divided as follows: introduction setup solr applications summary 1. introduction apache solr is an open source search server. it is based on the full text search engine called apache lucene . so basically solr is an http wrapper around an inverted index provided by lucene. an inverted index could be seen as a list of words where each word-entry links to the documents it is contained in. that way getting all documents for the search query "dzone" is a simple 'get' operation. one advantage of solr in enterprise projects is that you don't need any java code, although java itself has to be installed. if you are unsure when to use solr and when lucene, these answers could help. if you need to build your solr index from websites, you should take a look into the open source crawler called apache nutch before creating your own solution. to be convinced that solr is actually used in a lot of enterprise projects, take a look at this amazing list of public projects powered by solr . if you encounter problems then the mailing list or stackoverflow will help you. to make the introduction complete i would like to mention my personal link list and the resources page which lists books, articles and more interesting material. 2. setup solr 2.1. installation as the very first step, you should follow the official tutorial which covers the basic aspects of any search use case: indexing - get the data of any form into solr. examples: json, xml, csv and sql-database. this step creates the inverted index - i.e. it links every term to its documents. querying - ask solr to return the most relevant documents for the users' query to follow the official tutorial you'll have to download java and the latest version of solr here . more information about installation is available at the official description . next you'll have to decide which web server you choose for solr. in the official tutorial, jetty is used, but you can also use tomcat. when you choose tomcat be sure you are setting the utf-8 encoding in the server.xml . i would also research the different versions of solr, which can be quite confusing for beginners: the current stable version is 1.4.1. use this if you need a stable search and don't need one of the latest features. the next stable version of solr will be 3.x the versions 1.5 and 2.x will be skipped in order to reach the same versioning as lucene. version 4.x is the latest development branch. solr 4.x handles advanced features like language detection via tika, spatial search , results grouping (group by field / collapsing), a new "user-facing" query parser ( edismax handler ), near real time indexing, huge fuzzy search performance improvements, sql join-a like feature and more. 2.2. indexing if you've followed the official tutorial you have pushed some xml files into the solr index. this process is called indexing or feeding. there are a lot more possibilities to get data into solr: using the data import handler (dih) is a really powerful language neutral option. it allows you to read from a sql database, from csv, xml files, rss feeds, emails, etc. without any java knowledge. dih handles full-imports and delta-imports. this is necessary when only a small amount of documents were added, updated or deleted. the http interface is used from the post tool, which you have already used in the official tutorial to index xml files. client libraries in different languages also exist. (e.g. for java (solrj) or python ). before indexing you'll have to decide which data fields should be searchable and how the fields should get indexed. for example, when you have a field with html in it, then you can strip irrelevant characters , tokenize the text into 'searchable terms', lower case the terms and finally stem the terms . in contrast, if you would have a field with text in it that should not be interpreted (e.g. urls) you shouldn't tokenize it and use the default field type string. please refer to the official documentation about field and field type definitions in the schema.xml file. when designing an index keep in mind the advice from mauricio : "the document is what you will search for. " for example, if you have tweets and you want to search for similar users, you'll need to setup a user index - created from the tweets. then every document is a user. if you want to search for tweets, then setup a tweet index; then every document is a tweet. of course, you can setup both indices with the multi index options of solr. please also note that there is a project called solr cell which lets you extract the relevant information out of several different document types with the help of tika. 2.3. querying for debugging it is very convenient to use the http interface with a browser to query solr and get back xml. use firefox and the xml will be displayed nicely: you can also use the velocity contribution , a cross-browser tool, which will be covered in more detail in the section about 'search application prototyping' . to query the index you can use the dismax handler or standard query handler . you can filter and sort the results: q=superman&fq=type:book&sort=price asc you can also do a lot more ; one other concept is boosting. in solr you can boost while indexing and while querying. to prefer the terms in the title write: q=title:superman^2 subject:superman when using the dismax request handler write: q=superman&qf=title^2 subject check out all the various query options like fuzzy search , spellcheck query input , facets , collapsing and suffix query support . 3. applications now i will list some interesting use cases for solr - in no particular order. to see how powerful and flexible this open source search server is. 3.1. drupal integration the drupal integration can be seen as generic use case to integrate solr into php projects. for the php integration you have the choice to either use the http interface for querying and retrieving xml or json. or to use the php solr client library . here is a screenshot of a typical faceted search in drupal : for more information about faceted search look into the wiki of solr . more php projects which integrates solr: open source typo3- solr module magento enterprise - solr module . the open source integration is out dated. oxid - solr module . no open source integration available. 3.2. hathi trust the hathi trust project is a nice example that proves solr's ability to search big digital libraries. to quote directly from the article : "... the index for our one million book index is over 200 gigabytes ... so we expect to end up with a two terabyte index for 10 million books" other examples for libraries: vufind - aims to replace opac internet archive national library of australia 3.3. auto suggestions mainly, there are two approaches to implement auto-suggestions (also called auto-completion) with solr: via facets or via ngramfilterfactory . to push it to the extreme you can use a lucene index entirely in ram. this approach is used in a large music shop in germany. live examples for auto suggestions: kaufda.de 3.4. spatial search applications when mentioning spatial search, people have geographical based applications in mind. with solr, this ordinary use case is attainable . some examples for this are : city search - city guides yellow pages kaufda.de spatial search can be useful in many different ways : for bioinformatics, fingerprints search, facial search, etc. (getting the fingerprint of a document is important for duplicate detection). the simplest approach is implemented in jetwick to reduce duplicate tweets, but this yields a performance of o(n) where n is the number of queried terms. this is okay for 10 or less terms, but it can get even better at o(1)! the idea is to use a special hash set to get all similar documents. this technique is called local sensitive hashing . read this nice paper about 'near similarity search and plagiarism analysis' for more information. 3.5. duckduckgo duckduckgo is made with open source and its "zero click" information is done with the help of solr using the dismax query handler: the index for that feature contains 18m documents and has a size of ~12gb. for this case had to tune solr: " i have two requirements that differ a bit from most sites with respect to solr: i generally only show one result, with sometimes a couple below if you click on them. therefore, it was really important that the first result is what people expected. false positives are really bad in 0-click, so i needed a way to not show anything if a match wasn't too relevant. i got around these by a) tweaking dismax and schema and b) adding my own relevancy filter on top that would re-order and not show anything in various situations. " all the rest is done with tuned open source products. to quote gabriel again: "the main results are a hybrid of a lot of things, including external apis, e.g. bing, wolframalpha, yahoo, my own indexes and negative indexes (spam removal), etc. there are a bunch of different types of data i'm working with. " check out the other cool features such as privacy or bang searches . 3.6. clustering support with carrot2 carrot2 is one of the "contributed plugins" of solr. with carrot2 you can support clustering : " clustering is the assignment of a set of observations into subsets (called clusters) so that observations in the same cluster are similar in some sense. " see some research papers regarding clustering here . here is one visual example when applying clustering on the search "pannous" - our company : 3.7. near real time search solr isn't real time yet, but you can tune solr to the point where it becomes near real time, which means that the time ('real time latency') that a document takes to be searchable after it gets indexed is less than 60 seconds even if you need to update frequently. to make this work, you can setup two indices. one write-only index "w" for the indexer and one read-only index "r" for your application. index r refers to the same data directory of w, which has to be defined in the solrconfig.xml of r via: /pathto/indexw/data/ to make sure your users and the r index see the indexed documents of w, you have to trigger an empty commit every 60 seconds: wget -q http://localhost:port/solr/update?stream.body=%3ccommit/%3e -o /dev/null everytime such a commit is triggered a new searcher without any cache entries is created. this can harm performance for visitors hitting the empty cache directly after this commit, but you can fill the cache with static searches with the help of the newsearcher entry in your solrconfig.xml. additionally, the autowarmcount property needs to be tuned, which fills the cache with a newsearcher from old entries. also, take a look at the article 'scaling lucene and solr' , where experts explain in detail what to do with large indices (=> 'sharding') and what to do for high query volume (=> 'replicating'). 3.8. loggly = full text search in logs feeding log files into solr and searching them at near real-time shows that solr can handle massive amounts of data and queries the data quickly. i've setup a simple project where i'm doing similar things , but loggly has done a lot more to make the same task real-time and distributed. you'll need to keep the write index as small as possible otherwise commit time will increase too great. loggly creates a new solr index every 5 minutes and includes this when searching using the distributed capabilities of solr ! they are merging the cores to keep the number of indices small, but this is not as simple as it sounds. watch this video to get some details about their work. 3.9. solandra = solr + cassandra solandra combines solr and the distributed database cassandra , which was created by facebook for its inbox search and then open sourced. at the moment solandra is not intended for production use. there are still some bugs and the distributed limitations of solr apply to solandra too. tthe developers are working very hard to make solandra better. jetwick can now run via solandra just by changing the solrconfig.xml. solandra also has the advantages of being real-time (no optimize, no commit!) and distributed without any major setup involved. the same is true for solr cloud. 3.10. category browsing via facets solr provides facets , which make it easy to show the user some useful filter options like those shown in the "drupal integration" example. like i described earlier , it is even possible to browse through a deep category tree. the main advantage here is that the categories depend on the query. this way the user can further filter the search results with this category tree provided by you. here is an example where this feature is implemented for one of the biggest second hand stores in germany. a click on 'schauspieler' shows its sub-items: other shops: game-change 3.11. jetwick - open twitter search you may have noticed that twitter is using lucene under the hood . twitter has a very extreme use case: over 1,000 tweets per second, over 12,000 queries per second, but the real-time latency is under 10 seconds! however, the relevancy at that volume is often not that good in my opinion. twitter search often contains a lot of duplicates and noise. reducing this was one reason i created jetwick in my spare time. i'm mentioning jetwick here because it makes extreme use of facets which provides all the filters to the user. facets are used for the rss-alike feature (saved searches), the various filters like language and retweet-count on the left, and to get trending terms and links on the right: to make jetwick more scalable i'll need to decide which of the following distribution options to choose: use solr cloud with zookeeper use solandra move from solr to elasticsearch which is also based on apache lucene other examples with a lot of facets: cnet reviews - product reviews. electronics reviews, computer reviews & more. shopper.com - compare prices and shop for computers, cell phones, digital cameras & more. zappos - shoes and clothing. manta.com - find companies. connect with customers. 3.12. plaxo - online address management plaxo.com , which is now owned by comcast, hosts web addresses for more than 40 million people and offers smart search through the addresses - with the help of solr. plaxo is trying to get the latest 'social' information of your contacts through blog posts, tweets, etc. plaxo also tries to reduce duplicates . 3.13. replace fast or google search several users report that they have migrated from a commercial search solution like fast or google search appliance (gsa) to solr (or lucene). the reasons for that migration are different: fast drops linux support and google can make integration problems. the main reason for me is that solr isn't a black box —you can tweak the source code, maintain old versions and fix your bugs more quickly! 3.14. search application prototyping with the help of the already integrated velocity plugin and the data import handler it is possible to create an application prototype for your search within a few hours. the next version of solr makes the use of velocity easier. the gui is available via http://localhost:port/solr/browse if you are a ruby on rails user, you can take a look into flare. to learn more about search application prototyping, check out this video introduction and take a look at these slides. 3.15. solr as a whitelist imagine you are the new google and you have a lot of different types of data to display e.g. 'news', 'video', 'music', 'maps', 'shopping' and much more. some of those types can only be retrieved from some legacy systems and you only want to show the most appropriated types based on your business logic . e.g. a query which contains 'new york' should result in the selection of results from 'maps', but 'new yorker' should prefer results from the 'shopping' type. with solr you can set up such a whitelist-index that will help to decide which type is more important for the search query. for example if you get more or more relevant results for the 'shopping' type then you should prefer results from this type. without the whitelist-index - i.e. having all data in separate indices or systems, would make it nearly impossible to compare the relevancy. the whitelist-index can be used as illustrated in the next steps. 1. query the whitelist-index, 2. decide which data types to display, 3. query the sub-systems and 4. display results from the selected types only. 3.16. future solr is also useful for scientific applications, such as a dna search systems. i believe solr can also be used for completely different alphabets so that you can query nucleotide sequences - instead of words - to get the matching genes and determine which organism the sequence occurs in, something similar to blast . another idea you could harness would be to build a very personalized search. every user can drag and drop their websites of choice and query them afterwards. for example, often i only need stackoverflow, some wikis and some mailing lists with the expected results, but normal web search engines (google, bing, etc.) give me results that are too cluttered. my final idea for a future solr-based app could be a lucene/solr implementation of desktop search. solr's facets would be especially handy to quickly filter different sources (files, folders, bookmarks, man pages, ...). it would be a great way to wade through those extra messy desktops. 4. summary the next time you think about a problem, think about solr! even if you don't know java and even if you know nothing about search: solr should be in your toolbox. solr doesn't only offer professional full text search, it could also add valuable features to your application. some of them i covered in this article, but i'm sure there are still some exciting possibilities waiting for you!

I am a Spring framework user for last three years and I have really enjoyed working with Spring. One thing that I am seeing these days is the heavy use of AspectJ in most of SpringSource products. Spring Roo is a RAD tool for Java developers which makes use of AspectJ ITD's for separate compilation units. Spring transaction management and exception translation is also done using Aspectj. There are also numerous other usages of Aspectj in Spring products. In this article, I am going to talk about another cool usage of AspectJ by Spring - Mocking Static Methods. These days most of the developers write unit test cases and it is very common to use any of the mocking libraries like EasyMock, Mockito etc. to mock the external dependencies of a class. Using any of these mocking libraries it is very easy to mock calls to other class instance method. But most of these mocking framework does not provide the facility to mock the calls to static methods. Spring provides you the capability to mock static methods by using Spring Aspects library. In order to use this feature you need to add spring-aspects.jar dependency in your pom.xml. org.springframework spring-aspects 3.0.4.RELEASE Next thing you need to do is to convert your project in to a AspectJ project. If you are using Eclipse or STS(SpringSource Tool Suite) you can do that by right-click your project -> configure -> convert to AspectJ project. STS by default has AspectJ plugin, for eclipse users you need to install AspectJ plugin for above to work. I would recommend using STS for developing Spring based applications. The two aspects and one annotation that is of interest in spring-aspects.jar are : AbstractMethodMockingControl : This is an abstract aspect to enable mocking of methods picked out by a pointcut. All the child aspects need to define mockStaticsTestMethod() and methodToMock() pointcuts. The mockStaticTestMethod() pointcut is used to indicate when mocking should be triggered and methodToMock() pointcut is used to define which method invocations to mock. AnnotationDrivenStaticEntityMockingControl : This is the single implementation of AbstractMethodMockingControl aspect which exists in spring-aspects.jar. This is an annotation-based aspect to use in a test build to enable mocking static methods on Entity classes. In this aspect mockStaticTestMethod() pointcut defines that for classes marked with @MockStaticEntityMethods annotation mocking should be triggered and methodToMock() pointcut defines that all the public static methods in the classes marked with @Entity annotation should be mocked. MockStaticEntityMethods : Annotation to indicate a test class for whose @Test methods static methods on Entity classes should be mocked. The AnnotationDrivenStaticEntityMockingControl provide the facility to mock static methods of any class which is marked with @Entity annotation. But usually we would need to mock static method of classes other than marked with @Entity annotation. The only thing we need to do to make it work is to extend AbstractMethodMockingControl aspect and provide definitions for mockStaticsTestMethod() and methodToMock() pointcuts. For example, lets write an aspect which should mock all the public static methods of classes marked with @Component annotation. package com.shekhar.javalobby; import org.springframework.mock.staticmock.AbstractMethodMockingControl; import org.springframework.stereotype.Component; import org.springframework.mock.staticmock.MockStaticEntityMethods;; public aspect AnnotationDrivenStaticComponentMockingControl extends AbstractMethodMockingControl { public static void playback() { AnnotationDrivenStaticComponentMockingControl.aspectOf().playbackInternal(); } public static void expectReturn(Object retVal) { AnnotationDrivenStaticComponentMockingControl.aspectOf().expectReturnInternal(retVal); } public static void expectThrow(Throwable throwable) { AnnotationDrivenStaticComponentMockingControl.aspectOf().expectThrowInternal(throwable); } protected pointcut mockStaticsTestMethod() : execution(public * (@MockStaticEntityMethods *).*(..)); protected pointcut methodToMock() : execution(public static * (@Component *).*(..)); } The only difference between AnnotationDrivenStaticEntityMockingControl(comes with spring-aspects.jar) and AnnotationDrivenStaticComponentMockingControl(custom that we have written above) is in methodToMock() pointcut. In methodToMock() pointcut we have specified that it should mock all the static methods in any class marked with @Component annotation. Now that we have written the custom aspect lets test it. I have created a simple ExampleService with one static method. This is the method which we want to mock. @Component public class ExampleService implements Service { /** * Reads next record from input */ public String getMessage() { return myName(); } public static String myName() { return "shekhar"; } } This class will return "shekhar" when getMessage() method will be called. Lets test this without mocking package com.shekhar.javalobby; import org.junit.Assert; import org.junit.Test; public class ExampleConfigurationTests { private ExampleService service = new ExampleService(); @Test public void testSimpleProperties() throws Exception { String myName = service.getMessage(); Assert.assertEquals("shekhar", myName); } } This test will work fine. Now let's add mocking to this test class. There are two things that we need to do in our test We need to annotate our test with @MockStaticEntityMethods to indicate that static methods of @Component classes will be mocked. Please note that it is not required to use @MockStaticEntityMethods annotation you can create your own annotation and use that in mockStaticsTestMethod() pointcut. So, I could have created an annotation called @MockStaticComponentMethods and used that in mockStaticsTestMethod() pointcut. But I just reused the @MockStaticEntityMethods annotation. In our test methods we need to first invoke the static method which we want to mock so that it gets recorded. Next we need to set our expectation i.e. what should be returned from the mock and finally we need to call the playback method to stop recording mock calls and enter playback state. To make it more concrete lets apply mocking to the above test import org.junit.Assert; import org.junit.Test; import org.springframework.mock.staticmock.MockStaticEntityMethods; @MockStaticEntityMethods public class ExampleConfigurationTests { private ExampleService service = new ExampleService(); @Test public void testSimpleProperties() throws Exception { ExampleService.myName(); AnnotationDrivenStaticComponentMockingControl.expectReturn("shekhargulati"); AnnotationDrivenStaticComponentMockingControl.playback(); String myName = service.getMessage(); Assert.assertEquals("shekhargulati", myName); } } As you can see we annotated the test class with @MockStaticEntityMethods annotation and in the test method we first recorded the call (ExampleService.myName()), then we set the expectations, then we did the playback and finally called the actual code. In this way you can mock the static method of class.

Sometimes you have tests where the fixtures can be recyled, instead of being recreated; but that's not all: sometimes, even the state reset we preach for Shared Fixture gets in the way instead of being mandatory. For example, think of testing addition and removal of values over a collection object: to test the removal, you need a collection containing something. But this collection has to be populated somewhere, involving other methods which would have their own independent tests. If you look closer, however, you notice that the collection you need for testing values removal is just like the one that testAddsElements() produces when it succeeds, and simply throws away when the Test Methods returns... The pattern In these cases, the Chained Tests pattern can help you design isolated tests, by expressing the dependencies they implicitly have. The application of this pattern results in a form of Api encapsulation: a test that tests Collection::remove() does not need to use Collection::add(), and won't be influenced by a change in the name or the signature of add(), which it does not focus on. The test just needs just a collection which has been already filled with values. Before "spolverare" unserialization and other strange techniques, just consider using the leftovers from a previous test. Let's suppose we have two tests A and B, where B uses a variable produced in A. Of course, if A fails, the dependent Test Method B cannot (and shouldn't) be run; it will only add noise with an obvious failure which derives not from the behavior really under test there, but just from the fixture setup, which happens to be the previous test. If A instead runs correctly, B can be run. Implementation In PHPUnit, Chained Tests are supported when they are in the same Testcase Class. In this case, the Testcase Objects are not totally isolated anymore: you can return variables and objects from one test, in order for PHPUnit to pass them to other tests, that simply asks for them as dependencies. You can do so by adding @depends annotation to dependent tests. Note that this annotation can also be used without passing fixtures around, just for the sake of not executing complex tests where the base ones have already failed. @depends working by stopping the execution of tests down in the chain when the dependencies fail: they will be marked as skipped and signaled by an S in PHPUnit's ordinary output. The tests which produce a fixture as a leftover must return that fixture at the end of the test. The tests accepting fixtures, tagged with @depends, must accept parameters in the Test Method signatures. In case of multiple parameters to return, you can return an array() and use list() in the dependent method to extract variables from it. There is a known issue in running dependent tests with --filter: dependent tests would not run alone, so if you plan to use filter you may want to pick similar test names for couples of dependent/dependency in order to execute both of them when needed. Example Here we have a sample Testcase Class where the second and third tests are dependent on the first. Long chains of tests are not encouraged as it becomes increasingly difficult to tell what is the state of the fixture in the tests at the end of chain. assertEquals(15, $array[2], 'The third element was not 15.'); return $array; } /** * @depends testArrayAcceptsNewValues */ public function testArrayCanDeleteValues($array) { unset($array[2]); $this->assertEquals(2, count($array), 'There are more than 2 elements in the array.'); } /** * @depends testArrayAcceptsNewValues */ public function testArrayAcceptsDuplicateValues($array) { $array[] = 15; $this->assertEquals(4, count($array), 'There are less than 4 elements in the array.'); } } If we execute the whole test case, we obtain: [10:40:55][giorgio@Desmond:~]$ phpunit txt/articles/chainedtest.php PHPUnit 3.5.5 by Sebastian Bergmann. ... Time: 0 seconds, Memory: 5.00Mb OK (3 tests, 3 assertions) If we put a $this->fail() in the first test, instead, we get: [10:41:40][giorgio@Desmond:~]$ phpunit txt/articles/chainedtest.php PHPUnit 3.5.5 by Sebastian Bergmann. FSS Time: 0 seconds, Memory: 5.00Mb There was 1 failure: 1) ArrayTest::testArrayAcceptsNewValues /home/giorgio/Dropbox/txt/articles/chainedtest.php:13 FAILURES! Tests: 1, Assertions: 1, Failures: 1, Skipped: 2. If we filter the second test, we cannot run it due to the missing dependency: [10:49:25][giorgio@Desmond:~]$ phpunit --filter testArrayCan txt/articles/chainedtest.php PHPUnit 3.5.5 by Sebastian Bergmann. S Time: 0 seconds, Memory: 4.75Mb OK, but incomplete or skipped tests! Tests: 0, Assertions: 0, Skipped: 1. But if we filter the third test, thanks to the well-chosen name, we get: [10:49:16][giorgio@Desmond:~]$ phpunit --filter testArrayAccepts txt/articles/chainedtest.php PHPUnit 3.5.5 by Sebastian Bergmann. .. Time: 0 seconds, Memory: 5.00Mb OK (2 tests, 2 assertions)

In the previous article of this series, we saw how assertions are the key to self-validating tests. Even if an assertion's result is only a pass/fail value, programmers and testers find very handy to associate to the fail value some kind of additional information about the failure, to allow easier debugging without insertion of var_dump() and similar statements into the code. As I showed you in the previous article's code sample, assertions are useful the most when they fail, and in case they do, they should explain what happened so that a correction can be made quickly. An assertion that passes does not do anything; an assertion that fails shows an error. Whata happens when tests are failing The typical scenario is that of a large test suite, with hundreds or thousands of tests. At the time of a commit to the source code repository (or of a push if we're talking about git), the whole test suite must be green. Even if only one tests is red, the test suite is regarded as red. Sometimes, a regression is introduced by a programming error or an unforeseen scenario, and the output of the test suite becomes something like this: [12:30:02][giorgio@Desmond:~/txt/articles]$ phpunit assertionmessage.php PHPUnit 3.5.5 by Sebastian Bergmann. FFFF Time: 0 seconds, Memory: 3.75Mb There were 4 failures: 1) AssertionMessagesTest::testAsserTrue The object is not an Iterator. Failed asserting that is true. /home/giorgio/Dropbox/txt/articles/assertionmessage.php:9 2) AssertionMessagesTest::testAssertEquals The square root of 17 is not 4 but 4.1231056256177. Failed asserting that matches expected . /home/giorgio/Dropbox/txt/articles/assertionmessage.php:17 3) AssertionMessagesTest::testAssertContainsFailsWithCustomMessage The array does not contain 4. Failed asserting that an array contains . /home/giorgio/Dropbox/txt/articles/assertionmessage.php:24 4) AssertionMessagesTest::testAssertContains Failed asserting that an array contains . /home/giorgio/Dropbox/txt/articles/assertionmessage.php:36 FAILURES! Tests: 4, Assertions: 4, Failures: 4. The green tests are not noisy (which would be a smell), since they are not interesting at the moment. The red tests, which our focus is on, are shown with all the associated information. Along with the test name, the line of interest, and the kind of problem (failure/error), the assertion message is shown. How to write a meaningful message An assertion message can be passed as an argument of the assertion method, to be used in case of failure. Writing good assertion messages is an art, but before diving into some code to show you some messages I wrote, we can see how assertion messages are categorized in literature: Assertion-Identifying Message: includes information on which of the assertions in the same method caused the failure. This information can be for example the name of the checked variables. It's good to understand at a glance which assertion failed, but PHPUnit shows you the line of the test method that caused the failure, so it's not strictly necessary to include this kind of messages. Expectation-Describing Message: tells us what should have happened, but did not according to the data passed to the assertion method. Argument-Describing Message: when using assertions which are not very smart, like assertFalse() or assertTrue(), a failure would tell us by default only something like False was passed to assertTrue(). Adding an Argument-Describing Message will tell us instead The predicate on the set containing all red cars was false instead of true. Watch the test fail I can never say it enoug: when you write your tests, watch them fail (this mean also that each test should be written in advance with regard to the code it exercises, following the discipline of Test-Driven Development). This practice is not only useful for avoiding false positives, which are rare anyway, but to show at least for once the assertion message. You can check that when the test fails, the error message is meaningful and the test does not cause a Fatal Error. Watch your test fail to ensure that when they'll fail due to a regression in the future, you will be pleased by how fast you'll learn what's wrong. Example The code sample shows how to use PHPUnit's assertion methods with the $message optional argument. Any time you use basic methods like assertTrue(), a message is mandatory. The need for $message is inversely proportional to the specificity of the assertion: for example when assertContains() fails, PHPUnit has already valuable information in the the method itself to produce a readable error message. Note that the usage of "" (double quotes) enables fast composition of error messages via variable substitution. assertTrue($object instanceof Iterator, 'The object is not an Iterator.'); } public function testAssertEquals() { $expected = 4; $square = 17; $result = sqrt($square); $this->assertEquals($expected, $result, "The square root of $square is not $expected but $result."); } public function testAssertContainsFailsWithCustomMessage() { $array = array(1, 2, 3); $testElement = 4; $this->assertContains($testElement, $array, "The array does not contain $testElement."); } /** * This Assertion Message would be enough. It will even be better in some cases, * since PHPUnit would display nicely $testElement even if it was not "castable" * to a string (an object or array). */ public function testAssertContains() { $array = array(1, 2, 3); $testElement = 4; $this->assertContains($testElement, $array); } }

JUnit 4.8 introduced Categories: a mechanism to label and group tests, giving developers the option to include or exclude groups (or categories.) This post presents a brief overview of JUnit categories and some unexpected behavior I have found while using them. 1. Quick Introduction The following example shows how to use categories: (adapted from JUnit’s release notes) public interface FastTests { /* category marker */ } public interface SlowTests { /* category marker */ } public class A { @Category(SlowTests.class) @Test public void a() {} } @Category(FastTests.class}) public class B { @Test public void b() {} } @RunWith(Categories.class) @IncludeCategory(SlowTests.class) @ExcludeCategory(FastTests.class) @SuiteClasses({ A.class, B.class }) public class SlowTestSuite {} Lines 1, 2: we define two categories, FastTests and SlowTests. JUnit categories can be defined as classes or interfaces. Since a category acts like a label or marker, my intuition tells me to use interfaces. Line 5: we use the annotation @org.junit.experimental.categories.Category to label test classes and test methods with one or more categories. Lines 6, 9: test methods and test classes can be marked as belonging to one or more categories of tests. Labeling a test class with a category automatically includes all its test methods in such category. Lines 14 to 18: currently programmatic test suites (line 17) are the only way to specify which test categories (line 14) should be included (line 15) or excluded (line 16) when the suite is executed. I find this approach (especially the way test classes need to included in the suite) too verbose and not so flexible. Hopefully Ant, Maven and IDEs will provide support for categories (with a simpler configuration) in the very near future. Note: I recently discovered ClasspathSuite, a project that simplifies the creation of programmatic JUnit test suites. For example, we can specify we want to include in a test suite all tests whose names end with “UnitTest.” 2. Category Subtyping Categories also support subtyping. Let’s say we have the category IntegrationTests that extends SlowTests: public interface IntegrationTests extends SlowTests {} Any test class or test method labeled with the category IntegrationTests is also part of the category SlowTests. To be honest, I don’t know how handy category subtyping could be. I’ll need to experiment with it more to have an opinion. 3. Categories and Test Inheritance 3a. Method-level Categories JUnit behaves as expected when test inheritance is combined with method-level categories. For example: public class D { @Category(GuiTest.class) @Test public void d() {} } public class E extends D { @Category(GuiTest.class) @Test public void e() {} } @RunWith(Categories.class) @IncludeCategory(GuiTest.class) @SuiteClasses(E.class) public class TestSuite {} As I expected, when running TestSuite, test methods d and e are executed (both methods belong to the GuiTest category and E inherits method d from superclass D.) Nice! 3b. Class-level Categories On the other hand, unless I’m missing something, I think I found some strange behavior in JUnit in this scenario. Consider the following classes: @Category(GuiTest.class) public class A { @Test public void a() {} } public class B extends A { @Test public void b() {} } @RunWith(Categories.class) @IncludeCategory(GuiTest.class) @SuiteClasses(B.class) public class TestSuite {} As we can see, TestSuite should execute the tests in B that belong to the category GuiTest. I was expecting TestSuite to execute test method a, even though B is not marked as a GuiTest. Here is my reasoning: test method a belongs to the category GuiTest because test class A is labeled with such category test class B is an A and it inherits test method a Therefore, TestSuite should execute test method a. But it doesn’t! Here is a screenshot of the results I get (click to see full size.) There are two ways to fix this issue, depending on what test methods we want to actually run: Label class B with GuiTest. In this case, both methods, a and b, will be executed. Label method a with GuiTest. In this case, only method a will be executed. (I’ll be posting a question regarding this issue in the JUnit mailing list shortly.) 4. Categories vs. TestNG Groups (You saw this one coming, didn’t you?) Categories (or groups) have been part of TestNG for long time. Unlike JUnit’s, TestNG’s groups are defined as simple strings, not as classes or interfaces. As a static typing lover, I was pretty happy with JUnit categories. By using an IDE, we could safely rename a category or look for usages of a category within a project. Even though my observation was correct, I was missing one important point: all this works great as long as your test suite is written in Java. In the real world, I’d like to define a test suite in either Ant or Maven (or Gradle, or Rake.) In this scenario, having categories as Java types does not bring any benefit. In fact, I suspect it would be very verbose and error-prone to specify the fully-qualified name of a category in a build script. Renaming a category now would be limited to a text-based “search and replace.” Ant and Maven really need to provide a way to specify JUnit categories, clever enough to be fool-proof. As you may expect, I prefer the simplicity and pragmatism of TestNG’s groups. Update: my good friend (and creator of the TestNG framework,) Cédric, reminded me that we can use regular expressions to include or exclude groups in a test suite (details here.) This is really powerful! 5. My Usage of Categories I’m not using JUnit categories in my test suites yet. I started to look into JUnit categories because I wasn’t completely happy with the way we recognized GUI tests in FEST. We recognize test methods or test classes as “GUI tests” if they have been annotated with the @GUITest (provided by FEST.) When a “GUI test” fails, FEST automatically takes a screenshot of the desktop and includes it in the JUnit or TestNG HTML report. The problem is, our @GUITest annotation is duplicating the functionality of JUnit categories. To solve this issue, I created a JUnit extension that recognizes test methods or test classes as “GUI tests” if they belong to the GuiTest category. At this moment GuiTest is an interface provided by FEST, but I’m thinking about letting users specify their own GuiTest category as well. I also refactored this functionality out of the Swing testing module, expecting to reuse it once I implement a JavaFX testing module :) You can find the FEST code that deals with JUnit categories at github. 6. Conclusion Having the ability to label and group tests via categories is really a great feature. I still have some reservations about the practicality of defining categories as Java types, the lack of support for this feature from Ant and Maven (not JUnit’s fault,) and the unexpected behavior I noticed when combining class-level categories and test inheritance. On the brighter side, categories are still an experimental, non-final feature. I’m sure will see many improvements in future JUnit releases :) Feedback is always welcome. From http://alexruiz.developerblogs.com/?p=1711

Maven profiles, like chainsaws, are a valuable tool, with whose power you can easily get carried away, wielding them upon problems to which they are unsuited. Whilst you're unlikely to sever a leg misusing Maven profiles, I thought it worthwhile to share some suggestions about when and when not to use them. These three best practices are all born from real-world mishaps: The build must pass when no profile has been activated Never use Use profiles to manage build-time variables, not run-time variables and not (with rare exceptions) alternative versions of your artifact I'll expand upon these recommendations in a moment. First, though, let's have a brief round-up of what Maven profiles are and do. Maven Profiles 101 A Maven profile is a sub-set of POM declarations that you can activate or disactivate according to some condition. When activated, they override the definitions in the corresponding standard tags of the POM. One way to activate a profile is to simply launch Maven with a -P flag followed by the desired profile name(s), but they can also be activated automatically according to a range of contextual conditions: JDK version, OS name and version, presence or absence of a specific file or property. The standard example is when you want certain declarations to take effect automatically under Windows and others under Linux. Almost all the tags that can be placed directly in a POM can also be enclosed within a tag. The easiest place to read up further about the basics is the Build Profiles chapter of Sonatype's Maven book. It's freely available, readable, and explains the motivation behind profiles: making the build portable across different environments. The build must pass when no profile has been activated (Thanks to for this observation.) Why? Good practice is to minimise the effort required to make a successful build. This isn't hard to achieve with Maven, and there's no excuse for a simple mvn clean package not to work. A maintainer coming to the project will not immediately know that profile wibblewibble has to be activated for the build to succeed. Don't make her waste time finding it out. How to achieve it It can be achieved simply by providing sensible defaults in the main POM sections, which will be overridden if a profile is activated. Never use Why not? This flag activates the profile if no other profile is activated. Consequently, it will fail to activate the profile if any other profile is activated. This seems like a simple rule which would be hard to misunderstand, but in fact it's surprisingly easy to be fooled by its behaviour. When you run a multimodule build, the activeByDefault flag will fail to operate when any profile is activated, even if the profile is not defined in the module where the activeByDefault flag occurs. (So if you've got a default profile in your persistence module, and a skinny war profile in your web module... when you build the whole project, activating the skinny war profile because you don't want JARs duplicated between WAR and EAR, you'll find your persistence layer is missing something.) activeByDefault automates profile activation, which is a good thing; activates implicitly, which is less good; and has unexpected behaviour, which is thoroughly bad. By all means activate your profiles automatically, but do it explicitly and automatically, with a clearly defined rule. How to avoid it There's another, less documented way to achieve what aims to achieve. You can activate a profile in the absence of some property: !foo.bar This will activate the profile "nofoobar" whenever the property foo.bar is not defined. Define that same property in some other profile: nofoobar will automatically become active whenever the other is not. This is admittedly more verbose than , but it's more powerful and, most importantly, surprise-free. Use profiles to adapt to build-time context, not run-time context, and not (with rare exceptions) to produce alternative versions of your artifact Profiles, in a nutshell, allow you to have multiple builds with a single POM. You can use this ability in two ways: Adapt the build to variable circumstances (developer's machine or CI server; with or without integration tests) whilst still producing the same final artifact, or Produce variant artifacts. We can further divide the second option into: structural variants, where the executable code in the variants is different, and variants which vary only in the value taken by some variable (such as a database connection parameter). If you need to vary the value of some variable at run-time, profiles are typically not the best way to achieve this. Producing structural variants is a rarer requirement -- it can happen if you need to target multiple platforms, such as JDK 1.4 and JDK 1.5 -- but it, too, is not recommended by the Maven people, and profiles are not the best way of achieving it. The most common case where profiles seem like a good solution is when you need different database connection parameters for development, test and production environments. It is tempting to meet this requirement by combining profiles with Maven's resource filtering capability to set variables in the deliverable artifact's configuration files (e.g. Spring context). This is a bad idea. Why? It's indirect: the point at which a variable's value is determined is far upstream from the point at which it takes effect. It makes work for the software's maintainers, who will need to retrace the chain of events in reverse It's error prone: when there are multiple variants of the same artifact floating around, it's easy to generate or use the wrong one by accident. You can only generate one of the variants per build, since the profiles are mutually exclusive. Therefore you will not be able to use the Maven release plugin if you need release versions of each variant (which you typically will). It's against Maven convention, which is to produce a single artifact per project (plus secondary artifacts such as documentation). It slows down feedback: changing the variable's value requires a rebuild. If you configured at run-time you would only need to restart the application (and perhaps not even that). One should always aim for rapid feedback. Profiles are there to help you ensure your project will build in a variety of environments: a Windows developer's machine and a CI server, for instance. They weren't intended to help you build variant artifacts from the same project, nor to inject run-time configuration into your project. How to achieve it If you need to get variable runtime configuration into your project, there are alternatives: Use JNDI for your database connections. Your project only contains the resource name of the datasource, which never changes. You configure the appropriate database parameters in the JNDI resource on the server. Use system properties: Spring, for example, will pick these up when attempting to resolve variables in its configuration. Define a standard mechanism for reading values from a configuration file that resides outside the project. For example, you could specify the path to a properties file in a system property. Structural variants are harder to achieve, and I confess I have no first-hand experience with them. I recommend you read this explanation of how to do them and why they're a bad idea, and if you still want to do them, take the option of multiple JAR plugin or assembly plugin executions, rather than profiles. At least that way, you'll be able to use the release plugin to generate all your artifacts in one build, rather than a single one at a time. Further reading Profiles chapter from the Sonatype Maven book. Deploying to multiple environments (prod, test, dev): Stackoverflow.com discussion; see the first and top-rated answer. Short of creating a specific project for the run-time configuration, you could simply use run-time parameters such as system properties. Creating multiple artifacts from one project: How to Create Two JARs from One Project (…and why you shouldn’t) by Tim O'Brien of Sonatype (the Maven people) Blog post explaining the same technique Maven best practices (not specifically about profiles): http://mindthegab.com/2010/10/21/boost-your-maven-build-with-best-practices/ http://blog.tallan.com/2010/09/16/maven-best-practices/ This article is a completely reworked version of a post from my blog.

Both approaches have some advantages and disadvantages and so it’s a question of the likelihood of particular problems and business requirements and priorities.

Spock is an open source testing framework for Java and Groovy that has been attracting a growing following, especially in the Groovy community. It lets you write concise, expressive tests, using a quite readable BDD-style notation. It even comes with its own mocking library built in. Oh. I thought he was a sci-fi character. Can I see an example? Sure. Here's a simple one from a coding kata I did recently: import spock.lang.Specification; class RomanCalculatorSpec extends Specification { def "I plus I should equal II"() { given: def calculator = new RomanCalculator() when: def result = calculator.add("I", "I") then: result == "II" } } In Spock, you don't have tests, you have specifications. These are normal Groovy classes that extend the Specifications class, which is actually a JUnit class. Your class contains a set of specifications, represented by methods with funny-method-names-in-quotes™. The funny-method-names-in-quotes™ take advantage of some Groovy magic to let you express your requirements in a very readable form. And since these classes are derived from JUnit, you can run them from within Eclipse like a normal Groovy unit test, and they produce standard JUnit reports, which is nice for CI servers. Another thing: notice the structure of this test? We are using given:, when: and then: to express actions and expected outcomes. This structure is common in Behaviour-Driven Development, or BDD, frameworks like Cucumber and easyb. Though Spock-style tests are generally more concise more technically-focused than tools like Cucumber and easyb, which are often used for automating acceptance tests. But I digress... Actually, the example I gave earlier was a bit terse. We could make our intent clearer by adding text descriptions after the when: and then: labels, as I've done here: def "I plus I should equal II"() { when: "I add two roman numbers together" def result = calculator.add("I", "I") then: "the result should be the roman number equivalent of their sum" result == "II" } This is an excellent of clarifying your ideas and documenting your API. But where are the AssertEquals statements? Aha! I'm glad you asked! Spock uses a feature called Power Asserts. The statement after the then: is your assert. If this test fails, Spock will display a detailed analysis of what went wrong, along the following lines: I plus I should equal II(com.wakaleo.training.spocktutorial.RomanCalculatorSpec) Time elapsed: 0.33 sec <<< FAILURE! Condition not satisfied: result == "II" | | I false 1 difference (50% similarity) I(-) I(I) at com.wakaleo.training.spocktutorial .RomanCalculatorSpec.I plus I should equal II(RomanCalculatorSpec.groovy:17) Nice! But in JUnit, I have @Before and @After for fixtures. Can I do that in Spock? Sure, but you don't use annotations. Instead you implement setup() and cleanup() methods (which are run before and after each specification). I've added one here to show you what they look like: import spock.lang.Specification; class RomanCalculatorSpec extends Specification { def calculator def setup() { calculator = new RomanCalculator() } def "I plus I should equal II"() { when: def result = calculator.add("I", "I") then: result == "II" } } You can also define a setupSpec() and cleanupSpec(), which are run just before the first test and just after the last one. I'm a big fan of parameterized tests in JUnit 4. Can I do that in Spock! You sure can! In fact it's one of Spock's killer features! def "The lowest number should go at the end"() { when: def result = calculator.add(a, b) then: result == sum where: a | b | sum "X" | "I" | "XI" "I" | "X" | "XI" "XX" | "I" | "XXI" "XX" | "II" | "XXII" "II" | "XX" | "XXII" } This code will run the test 5 times. The variables a, b, and sum are initialized from the rows in the table in the where: clause. And if any of the tests fail, you get That's pretty cool too. What about mocking? Can I use Mockito? Sure, if you want. but Spock actually comes with it's own mocking framework, which is pretty neat. You set up a mock or a stub using the Mock() method. I've shown two possible ways to use this method here: given: Subscriber subscriber1 = Mock() def subscriber2 = Mock(Subscriber) ... You can set these mocks up to behave in certain ways. Here are a few examples. You can say a method should return a certain value using the >> operator: subscriber1.isActive() >> true subscriber2.isActive() >> false Or you could get a method to throw an exception when it is called: subscriber.activate() >> { throw new BlacklistedSubscriberException() } Then you can test outcomes in a few different ways. Here is a more complicated example to show you some of your options: def "Messages published by the publisher should only be received by active subscribers"() { given: "a publisher" def publisher = new Publisher() and: "some active subscribers" Subscriber activeSubscriber1 = Mock() Subscriber activeSubscriber2 = Mock() activeSubscriber1.isActive() >> true activeSubscriber2.isActive() >> true publisher.add activeSubscriber1 publisher.add activeSubscriber2 and: "a deactivated subscriber" Subscriber deactivatedSubscriber = Mock() deactivatedSubscriber.isActive() >> false publisher.add deactivatedSubscriber when: "a message is published" publisher.publishMessage("Hi there") then: "the active subscribers should get the message" 1 * activeSubscriber1.receive("Hi there") 1 * activeSubscriber2.receive({ it.contains "Hi" }) and: "the deactivated subscriber didn't receive anything" 0 * deactivatedSubscriber.receive(_) } That does look neat. So what is the best place to use Spock? Spock is great for unit or integration testing of Groovy or Grails projects. On the other hand, tools like easyb amd cucumber are probably better for automated acceptance tests - the format is less technical and the reporting is more appropriate for non-developers. From http://www.wakaleo.com/blog/303-an-introduction-to-spock

Have you ever had to create a mock object in which most methods do nothing and are not called, but in others something useful needs to be done? EasyMock has some newish functionality to let you stub individual methods. But before I had heard about that, I had built a little framework (one base class) for creating mock objects which stubs those methods you want to stub, as well as logging every call made to the classes being mocked. It works like this: you choose a class which you need to mock, for example a service class called FooService, and you create a new class called FooServiceMock. You make it extend from AbstractMock, where T is the class you are mocking. As an example: public class FooServiceMock extends AbstractMock { public FooServiceMock() { super(FooService.class); } It needs to have a constructor to call the super constructor passing the class being mocked too. Perhaps that could be optimised, I don't have too much time right now. Next, you implement only those methods you expect to be called. For example: public class FooServiceMock extends AbstractMock { public FooServiceMock() { super(FooService.class); } /** * this is a method which exists in FooService, * but I want it to do something else. */ public String sayHello(String name){ return "Hello " + name + ", Foo here! This is a stub method!"; } To use the mock, you'll notice that it doesn't extend the class which it mocks, which might be problematic... Well, there are good reasons. To do the mocking, the abstract base class is actually going to create a dynamic proxy which wraps itself behind the interface of the class being mocked. To the caller, it looks like the FooService, but it's not actually anything related to it. Anytime a call to the FooService is made, the first thing which the proxy does is log that call, using XStream to create an XML representation of the parameters being passed into the method. Then, the proxy goes and looks in the instance of the mock class to see if it can find the method being called (well at least a method which takes the same parameters and has the same name and return type). If it finds such a method, it calls it. In our example, the sayHello(String) method would get called. It returns the result if there is one, to the caller. In the case where it cannot find the method, it throws an exception, because it assumes that if it was not implemented, you didn't expect it to be called. You could of course change this to suit your needs, maybe even calling the actual FooService. So, how to you use the FooServiceMock to create a FooService instance which you can use to mock your service? In the test, where you setup the class under test, you do this: FooServiceMock fooService = new FooServiceMock(); //perhaps tell it about objects you would //like it to return... instanceOfClassUnderTest.setFooService( fooService.getMock()); The setFooService(FooService) method on the instance of the class you are testing is in my case present, but you might not have it and may need to use reflection to do it. It's a question of how testable you write your classes, and is a design choice. The getMock() method on the AbstractMock class is the method which creates the dynamic proxy which wraps the instance of the mock. You can now test the class. There is however still something useful you can do after testing, i.e. assert that the right calls were made in the correct order with the right parameters. You do this in the test class to: assertEquals(1, fooService.getCalls().size()); assertEquals("[sayHello: Ant]", fooService.getCalls().toString()); The above tests that the sayHello(String) method was called just once, and passed the name "Ant". There are times when you might want to clear the call log, between parts of the test. For that, call the clearCalls() method on the mock object: fooService.clearCalls(); Have fun! From http://blog.maxant.co.uk/pebble/2010/11/03/1288813500000.html