Comments on: Unit Testing NHibernate DALs – What Are You *Really* Testing? http://lostechies.com/joeybeninghove/2007/05/17/unit-testing-nhibernate-dals-what-are-you-really-testing/ Just another LosTechies site Fri, 15 Oct 2010 23:08:45 +0000 hourly 1 http://wordpress.org/?v=3.4.2 By: Ben http://lostechies.com/joeybeninghove/2007/05/17/unit-testing-nhibernate-dals-what-are-you-really-testing/#comment-7 Ben Wed, 25 Feb 2009 23:41:14 +0000 /blogs/joeydotnet/archive/2007/05/17/unit-testing-nhibernate-dals-what-are-you-really-testing.aspx#comment-7 I know this is an old post, but I'm dealing with these questions now. I know this is an old post, but I’m dealing with these questions now.

]]> By: joeyDotNet http://lostechies.com/joeybeninghove/2007/05/17/unit-testing-nhibernate-dals-what-are-you-really-testing/#comment-6 joeyDotNet Tue, 22 May 2007 09:45:12 +0000 /blogs/joeydotnet/archive/2007/05/17/unit-testing-nhibernate-dals-what-are-you-really-testing.aspx#comment-6 Hey Ben, Yeah, I definitely know what you mean about dealing with broken session factories. I'm actually almost done with a new blog post on various ways to validate NHibernate mapping files that may help. Hopefully I'll get that posted in the next day or two. Hey Ben,
Yeah, I definitely know what you mean about dealing with broken session factories. I’m actually almost done with a new blog post on various ways to validate NHibernate mapping files that may help.

Hopefully I’ll get that posted in the next day or two.

]]> By: Ben Scheirman http://lostechies.com/joeybeninghove/2007/05/17/unit-testing-nhibernate-dals-what-are-you-really-testing/#comment-5 Ben Scheirman Fri, 18 May 2007 10:22:10 +0000 /blogs/joeydotnet/archive/2007/05/17/unit-testing-nhibernate-dals-what-are-you-really-testing.aspx#comment-5 Great post. Often this is a difficult thing to grasp when combining NHibernate and Unit Tests. I generally separate the tests into 2 namespaces (or assemblies) so that TDD.NET can run them independently. I have my Unit Tests that just test the behavior of my constructors, collections, and business rules. Then I have my integration tests that actually test to see if I can save and load a customer, for example. The problem that I face is that while it is easy to to TDD to drive out your mapping in this way, later on when you make a mapping change it will make ALL of your integration tests fail. This goes against one of the core concepts of unit testing. We should be able to identify EXACTLY what went wrong from (hopefully) a single failing test. But since an invalid mapping will render the session factory creation broken, most of the integration tests will break as well. Ideas? Great post. Often this is a difficult thing to grasp when combining NHibernate and Unit Tests.

I generally separate the tests into 2 namespaces (or assemblies) so that TDD.NET can run them independently.

I have my Unit Tests that just test the behavior of my constructors, collections, and business rules. Then I have my integration tests that actually test to see if I can save and load a customer, for example.

The problem that I face is that while it is easy to to TDD to drive out your mapping in this way, later on when you make a mapping change it will make ALL of your integration tests fail. This goes against one of the core concepts of unit testing. We should be able to identify EXACTLY what went wrong from (hopefully) a single failing test.

But since an invalid mapping will render the session factory creation broken, most of the integration tests will break as well.

Ideas?

]]>