I have a web application which currently exists as a single large class containing a front controller, and a bootstrap file which runs the class and passes in the settings.
Over time, the class has become over-large, with multiple concerns, that I have long wanted to refactor into what will be rather obvious subclasses. I'm very familiar with all kinds of refactorings.
However, for this application, at present there is no test coverage. Because all interactions are done by reading GET and POST parameters, there is only one public method into the system, i.e. such URL interactions through the front controller entry point. So the class is hard to unit test (I will be using PHPUnit) as it stands.
Obviously it is far safer to refactor with tests already in place. So I would welcome views on which strategy is best:
1) Create a pile of tests that implement GET and POST interactions as per a user using the web application; or
2) Create a pile of tests against the private functions by using the ReflectionClass workaround, and convert these to standard PHPUnit tests concurrently with the refactoring; or
3) Add unit testing after doing the subclass refactoring, testing the public entry API points to the new subclasses.
Either way how you approach it, I would do it one-functionality at a time if you can, it makes the task less daunting and therefore more achievable...
When confronted with such situations in the past (quite a few times in fact), I usually do this:
First of all, does the application use external dependencies like databases? The first thing is that we need to make these predictable. I usually try to find where the creation of these external dependencies occur and then delegate it to a good Dependency Injection framework (like Pimple).
Once that is done you can add rules to the DI factory methods that return test databases depending on some conditions set: I usually use a test domain for this, so if the site is foo.local then i also have a domain test.foo.local that sets (in the Apache config for example) an environment variable APPLICATION_ENV set to 'tests'. The DI container can then inspect this environment variable to serve the right dependency, in our case a connection pointing to a test database.
(It don't like that production code is aware of that it might be in a test scenario, but it's a lesser evil, and it can be alleviated by using config files that have the environment variable in the filename for example. I digress.)
When this is done, in my tests I then have a reference to the same test database which I then use to set up some predictable test data with some thing like DbUnit or test-db-acle (disclaimer, I wrote the latter one, so I tend towards that).
Once the test data is in place, I issue a request to the url in question with something like Guzzle (in various important variations) and record the output(s), which I then use to make a test that sets the behaviour 'in stone'.
NOTE: this is a very imprecise method, the best way for tests is in a test-driven way, but this is not a solution in this case obviously.
However, if you are reasonable sure that some manual testing has been done on this, then this is a good way to at least get a few smoke tests into place. In my experience the value of having even one rudimentary smoke test in place is infinitely more valuable than not having any.
Now that you have a smoke test in place, you can refactor this part of the application and move to a new one.
I hope this help....
Related
I have some doubts regarding the best approach to perform unit tests in Quarkus.
One option is using mocks, but I have a feeling that using mocks I just "make happy" the test coverage plugins, but on real I don't test anything with this approach.
Another option is using a real database, like embedded H2 database, but for this, I need to put the unit test in order (Insert, Get, Update, Delete), sometimes I'll need the inserted ID from another test to perform a delete operation, for example. There are some complex entities that create some difficulty to insert or delete. So, with this approach, I'll lose the concept of a unit test, because I'll lose the tests interdependency.
Does someone have a suggestion for this scenario?
Extra information: I'm using liquidbase, panache entity, junity.
Looks like you're looking for integration tests. I would probably go with the next options:
Depending on database type, you can either go with in-memory h2, or use testcontainers for external services.
Have test startup scripts for sql for general data, insert data directly for small test-cases.
Use tools like http://rest-assured.io/ (usually included in initial setup) to do real api calls
In case you really want unit tests, in this case 90% of the time you don't need to have the database to test functionality. Because of decoupling you probably have controllers(resources) separate from services. So in case of unit tests I would probably go with:
As stated above split logic from the networking layer, so that any Request handling and wrapping the output to satisfy clients is out of the logic layer. Split the logic itself into smaller chunks if possible, this will eliminate your main concern about complex data
For any dependent dependencies, either override them with test beans, or mock them using Mockito. This is especially important for persistency, you need to mock or fake any connection to database, you don't need to test this part because it's handled by libraries usually.
You can read JSON objects from files for really complex entities
If you really need to have persistency, look at the very first point in this answer.
So it feels like you want to check if your service is doing the job correctly on a higher level, which might not be a unit test, hope the answer will help
I am new to unit testing. Suppose I am building a web application. How do I know what to test? All the examples that you see are some sort of basic sum function that really has no real value, or at least I've never written a function to add to inputs and then return!
So, my question...on a web application, what are the sort of things that need tested?
I know that this is a broad question but anything will be helpful. I would be interested in links or anything that gives real life examples as opposed to concept examples that don't have any real life usage.
Take a look at your code, especially the bits where you have complex logic with loops, conditionals, etc, and ask yourself: How do I know if this works?
If you need to change the complex logic to take into account other corner cases then how do you know that the changes you introduce don't break the existing cases? This is precisely what unit testing is intended to address.
So, to answer your question about how it applies to web applications: suppose you have some code that lays out the page differently depending on the browser. One of your customers refuses to upgrade from IE6 and insists that you support that. So you unit test your layout code by simulating the connection string from IE6 and checking that the layout is what you expect.
A customer tells you they've found a security hole where using a particular cookie will give you administrator access. How do you know that you've fixed the bug and it doesn't happen again? Create a unit test for it, and run the unit tests on a daily basis so that you get an early warning if it fails.
You discover a bug where users with accents in their names get corrupted in the database. Abstract out the webform input from the database layer and add unit tests to ensure that (eg) UTF8 encoded data is stored in the database correctly and can be retrieved.
You get the idea. Anywhere where part of the process has a well-defined input and output is ideal for unit testing. Anything that doesn't is ideal for refactoring until it is well defined. Take a look at projects such as WebUnit, HTMLUnit, XMLUnit, CSSUnit.
The first part of testing is to write testable applications. Separate out as much functionality as possible from the UI. Refactor into smaller methods. Learn about dependency injection, and try using that to create methods that can take simple, throw-away input that produces known (and therefor testable) results. Look at mocking tools.
Infrastructure and data layer code is easiest to test.
Look at behavior-driven testing as well as test-driven design. For my money, behavior testing is better than pure unit testing; you can follow use-cases, so that tests match against expected usage patterns.
Unit testing means testing any unit of work, the smallest units of work are methods and functions., The art of unit testing is to define tests for a function that cannot just be checked by inspection, what unit test aims at is to test every possible functional requirement of a method.
Consider for example you have a login function, then there could be following tests that you could write for failures:
1. Does the function fail on empty username and password
2. Does the function fail on the correct username but the wrong password
3. Does the function fail on the correct password but the wrong username
The you would also write tests that the function would pass:
1. Does the function pass on correct username and password
This is just a basic example but this is what unit testing attempts to achieve, testing out things that may have been overlooked during development.
Then there is a purist approach too where a developer is first supposed to write tests and then the code to pass those tests (aka test driven development).
Resources:
http://devzone.zend.com/article/2772
http://www.ibm.com/developerworks/library/j-mocktest.html
If you're new to TDD, may I suggest a quick trip into the world of BDD? My experience is that the language really helps people pick up TDD more quickly. Particularly, I point you at this article, in which Dan North suggests "what to test":
http://blog.dannorth.net/introducing-bdd/
Note for transparency: I may be heavily involved in the BDD movement.
Regarding the classes to unit test in a web-app, I'd consider starting with controllers, domain objects if they have complex behaviour, and anything called "service", "manager", "helper" or "util". Please also try renaming any classes like this so that they are less generic and actually say what they do. Classes called "calculator" or "converter" are also good candidates, and you'll probably find more in the same package / folder.
There are a couple of good books which could help you too:
Martin Fowler, "Refactoring"
Michael Feathers, "Working Effectively with Legacy Code"
Good luck!
If you start out saying, "How do I test my web app?" that is biting off a lot at once, and it's going to be hard to see unit testing as providing any kind of benefit. I got into unit testing by starting with small pieces that were isolated, then writing libraries test-first, and only then building whole applications that were testable.
Generally a web app has a domain model, it has data access objects that do queries on a database and return domain objects, it has services that call the data access objects, and it has controllers that accept http requests and call the services.
Tests for the controllers will check that they call the right service method with the right parameters. Service objects can be mocks injected during test setup.
Tests for the services will check that they call the right data access objects and perform whatever logic they need to be performing. Data access objects can be mocks injected during test setup.
Tests for the data access objects will check that they perform the right database operation (query or update or whatever) by checking the contents of the database before and after. For dao tests you'll need a database, and a tool like DBUnit to pre-populate it before the test. Also your domain objects' getters and setters will get exercised with this test so you won't need a separate test for them.
Tests for the domain model will check that whatever domain logic you have encoded in them works (Sometimes you may not have any). If you design your domain model so it is not coupled to the database then the more logic you put in the domain model the better because it's easy to test. You shouldn't need any mocks for these tests.
For a web app the kind of tests you need to do are slightly different. Unit tests are tests which test a particular component of your program. For a web app, you would need to test that forms accept/reject the right inputs, that all links point to the right place, that it can cope with unexpected inputs etc. I'd have a look at Selenium if I were you, I've used it extensively in testing a number of sites: Selenium HQ
I don't have experience of testing web apps, but speaking generally: you unit test the smallest 'chunks' of your program possible. That means you test each function on an individual basis. Anything on a larger scale becomes an integration test.
Of course, there are going to be methods so simple that its not worth your time to write a test for them, but on the whole aim to test as great a proportion of your code as possible.
A rule of thumb is that if it is not worth testing it is not worth writing.
However, some things are very difficult to test, so you have the do some cost benefit analysis on what you test. If you initially aim for 70% code coverage, you will be on the right track.
If I'm having Data Access Layer (nHibernate) for example a class called UserProvider
and a Business Logic class UserBl, should I both test their methods SaveUser or GetUserById, or any other public method in DA layer which is called from BL layer. Is this a redundancy or a common practice to do?
Is it common to unit test DA layer, or that belongs to Integration test domain?
Is it better to have test database, or create database data during test?
Any help is appreciated.
There's no right answer to this, it really depends. Some people (e.g Roy Osherove) say you should only test code which has conditional logic (IF statements etc), which may or may not include your DAL. Some people (often those doing TDD) will say you should test everything, including the DAL, and aim for 100% code coverage.
Personally I only test it if it has logic in, so end up with some DAL methods tested and some not. Most of the time you just end up checking that your BL calls your DAL, which has some merit but I don't find necessary. I think it makes more sense to have integration tests which cover the app end-to-end, including the database, which covers things like GetUserById.
Either way, and you probably know this already, but make sure your unit tests don't touch an actual database. (No problem doing this, but that's an integration test not a unit test, as it takes a lot longer and involves complex setup, and should be run separately).
It is a good practice to write unit test for every layer, even the DAL.
I don't think running tests on the real db is a good idea, you might ruin important data. We used to set up a copy of the db for tests with just enough data in it to run tests on.
In our test project we had a special web.config file with test settings, like a ConnectionString to our test db.
In my experience it was helpful to test each layer on its own. Integrating it and test again. Integration test normally does not test all aspects. Sometimes if the Data Access Layer (I don't know nHibernate) is generated code or sort of generic code it looks like overkill. But I have seen it more than once that systematic testing pays off.
Is it redundancy? In my opinion it is not.
Is it common practice? Hard to tell. I would say no. I have seen it in some projects but not in all projects I worked in. Was often dependend on time/resources and mentality of the team / individiual developer.
Is it better to have test database, or create database data during test? This is quite a different question. Cannot be answered easily. Depends on your project. Create a new one is good but sometimes throws up unreal bugs (although bugs). It is depending on your project (product development or a proprietary development). Usually in an proprietary on site development a database gets migrated into from somewhere. So a second test is definitely needed with the migrated data. But this is rather at a system test level.
Unit testing the DAL is worth it as mentioned if there is logic in there, for example if using the same StoredProc for insert & update its worth knowing that an insert works, a subsequent call updates the previous and a select returns it and not a list. In your case SaveUser method probably inserts first time around and subsequently updates, its nice to know that this is whats being done at unit test stage.
If you're using a framework like iBatis or Hibernate where you can implement typehandlers its worth confirming that the handlers handle values in a way that's acceptable to your underlying DB.
As for testing against an actual DB if you use a framework like Spring you can avail of the supported database unit test classes with auto rollback of transactions so you run your tests and the DB is unaffected afterwards. See here for information. Others probably offer similiar support.
I used TDD as a development style on some projects in the past two years, but I always get stuck on the same point: how can I test the integration of the various parts of my program?
What I am currently doing is writing a testcase per class (this is my rule of thumb: a "unit" is a class, and each class has one or more testcases). I try to resolve dependencies by using mocks and stubs and this works really well as each class can be tested independently. After some coding, all important classes are tested. I then "wire" them together using an IoC container. And here I am stuck: How to test if the wiring was successfull and the objects interact the way I want?
An example: Think of a web application. There is a controller class which takes an array of ids, uses a repository to fetch the records based on these ids and then iterates over the records and writes them as a string to an outfile.
To make it simple, there would be three classes: Controller, Repository, OutfileWriter. Each of them is tested in isolation.
What I would do in order to test the "real" application: making the http request (either manually or automated) with some ids from the database and then look in the filesystem if the file was written. Of course this process could be automated, but still: doesn´t that duplicate the test-logic? Is this what is called an "integration test"? In a book i recently read about Unit Testing it seemed to me that integration testing was more of an anti-pattern?
IMO, and I have no literature to back me on this, but the key difference between our various forms of testing is scope,
Unit testing is testing isolated pieces of functionality [typically a method or stateful class]
Integration testing is testing the interaction of two or more dependent pieces [typically a service and consumer, or even a database connection, or connection to some other remote service]
System integration testing is testing of a system end to end [a special case of integration testing]
If you are familiar with unit testing, then it should come as no surprise that there is no such thing as a perfect or 'magic-bullet' test. Integration and system integration testing is very much like unit testing, in that each is a suite of tests set to verify a certain kind of behavior.
For each test, you set the scope which then dictates the input and expected output. You then execute the test, and evaluate the actual to the expected.
In practice, you may have a good idea how the system works, and so writing typical positive and negative path tests will come naturally. However, for any application of sufficient complexity, it is unreasonable to expect total coverage of every possible scenario.
Unfortunately, this means unexpected scenarios will crop up in Quality Assurance [QA], PreProduction [PP], and Production [Prod] cycles. At which point, your attempts to replicate these scenarios in dev should make their way into your integration and system integration suites as automated tests.
Hope this helps, :)
ps: pet-peeve #1: managers or devs calling integration and system integration tests "unit tests" simply because nUnit or MsTest was used to automate it ...
What you describe is indeed integration testing (more or less). And no, it is not an antipattern, but a necessary part of the sw development lifecycle.
Any reasonably complicated program is more than the sum of its parts. So however well you unit test it, you still have not much clue about whether the whole system is going to work as expected.
There are several aspects of why it is so:
unit tests are performed in an isolated environment, so they can't say anything about how the parts of the program are working together in real life
the "unit tester hat" easily limits one's view, so there are whole classes of factors which the developers simply don't recognize as something that needs to be tested*
even if they do, there are things which can't be reasonably tested in unit tests - e.g. how do you test whether your app server survives under high load, or if the DB connection goes down in the middle of a request?
* One example I just read from Luke Hohmann's book Beyond Software Architecture: in an app which applied strong antipiracy defense by creating and maintaining a "snapshot" of the IDs of HW components in the actual machine, the developers had the code very well covered with unit tests. Then QA managed to crash the app in 10 minutes by trying it out on a machine without a network card. As it turned out, since the developers were working on Macs, they took it for granted that the machine has a network card whose MAC address can be incorporated into the snapshot...
What I would do in order to test the
"real" application: making the http
request (either manually or automated)
with some ids from the database and
then look in the filesystem if the
file was written. Of course this
process could be automated, but still:
doesn´t that duplicate the test-logic?
Maybe you are duplicated code, but you are not duplicating efforts. Unit tests and integrations tests serve two different purposes, and usually both purposes are desired in the SDLC. If possible factor out code used for both unit/integration tests into a common library. I would also try to have separate projects for your unit/integration tests b/c
your unit tests should be ran separately (fast and no dependencies). Your integration tests will be more brittle and break often so you probably will have a different policy for running/maintaining those tests.
Is this what is called an "integration
test"?
Yes indeed it is.
In an integration test, just as in a unit test you need to validate what happened in the test. In your example you specified an OutfileWriter, You would need some mechanism to verify that the file and data is good. You really want to automate this so you might want to have a:
Class OutFilevalidator {
function isCorrect(fName, dataList) {
// open file read data and
// validation logic
}
You might review "Taming the Beast", a presentation by Markus Clermont and John Thomas about automated testing of AJAX applications.
YouTube Video
Very rough summary of a relevant piece: you want to use the smallest testing technique you can for any specific verification. Spelling the same idea another way, you are trying to minimize the time required to run all of the tests, without sacrificing any information.
The larger tests, therefore are mostly about making sure that the plumbing is right - is Tab A actually in slot A, rather than slot B; do both components agree that length is measured in meters, rather than feet, and so on.
There's going to be duplication in which code paths are executed, and possibly you will reuse some of the setup and verification code, but I wouldn't normally expect your integration tests to include the same level of combinatoric explosion that would happen at a unit level.
Driving your TDD with BDD would cover most of this for you. You can use Cucumber / SpecFlow, with WatiR / WatiN. For each feature it has one or more scenarios, and you work on one scenario (behaviour) at a time, and when it passes, you move onto the next scenario until the feature is complete.
To complete a scenario, you have to use TDD to drive the code necessary to make each step in the current scenario pass. The scenarios are agnostic to your back end implementation, however they verify that your implementation works; if there is something that isn't working in the web app for that feature, the behaviour needs to be in a scenario.
You can of course use integration testing, as others pointed out.
I have a simple project, mostly consisting of back-end service code. I have this fully unit-tested, including my DAL layer...
Now I have to write the front-end. I re-use what business objects I can in my front-end, and at one point I have a grid that renders some output. I have my DAL object with some function called DisplayRecords(id) which displays the records for a given ID...
All of this DAL objects are unit tested. But is it worth it to write a unit test for the DisplayRecords() function? This function is calling a stored proc, which is doing some joins. This means that my unit-test would have to set-up multiple tables, one with 15 columns, and its return value is a DataSet (this is the only function in my DAL that returns a datset - because it wasnt worth it to create an object just for this one grid)...
Is stuff like this even worth testing? What about front-end logic in general - do people tend to skip unit tests for the ASP.NET front-end, similar to how people 'skip' the logic for private functions? I know the latter is a bit different - testing behavior vs implementation and all... but, am just curious what the general rule-of-thumb is?
Thanks very much
There are a few things that weigh into whether you should write tests:
It's all about confidence. You build tests so that you have confidence to make changes. Can you confidently make changes without tests?
How important is this code to the consumers of the application? If this is critical and central to everything, test it.
How embarrassing is it if you have regressions? On my last project, my goal was no regressions-- I didn't want the client to have to report the same bug twice. So every important bug got a test to reproduce it before it was fixed.
How hard is it to write the test? There are many tools that can help ease the pain:
Selenium is well understood and straightforward to set up. Can be a little expensive to maintain a large test suite in selenium. You'll need the fixture data for this to work.
Use a mock to stub out your DAL call, assuming its tested elsewhere. That way you can save time creating all the fixture data. This is a common pattern in testing Java/Spring controllers.
Break the code down in other ways simply so that it can be tested. For example, extract out the code that formats a specific grid cell, and write unit tests around that, independent of the view code or real data.
I tend to make quick Selenium tests and just sit and watch the app do its thing - that's a fast validation method which avoids all the manual clicking.
Fully automated UI testing is tedious and should IMO only be done in more mature apps where the UI won't change much. Regarding the 'in-between' code, I would test it if it is reused and/or complicated/ introducing new logic, but if its just more or less a new sequence of DAL method calls and specific to a single view I would skip it.