# Unit test question! **URL:** \ **Category:** Rails\ **Created:** [June 13, 2007, 6:25am UTC](https://www.ruby-forum.com/t/unit-test-question/106674 "2007-06-13T06:25:14Z")\ **Posts on this page:** 4\ **Page:** 1
**Author:** ![DBC\_User](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/d/ee7513/32.png) [@DBC\_User](https://www.ruby-forum.com/u/DBC_User)\ **Post date:** [June 13, 2007, 6:25am UTC](https://www.ruby-forum.com/t/unit-test-question/106674/1 "2007-06-13T06:25:14Z")
Hi, I have a model and I am trying to do the testing. So my test code is something like the following def test\_first assert\_nil Person.find(1) end in this I am expecting the test to run against the development database as only development database has the table. But when I ran, it is failing because the test table is not found. Why is model testing is pointing to test database anyway? Thanks. ---
**Author:** ![DBC\_User](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/d/ee7513/32.png) [@DBC\_User](https://www.ruby-forum.com/u/DBC_User)\ **Post date:** [June 13, 2007, 6:39am UTC](https://www.ruby-forum.com/t/unit-test-question/106674/2 "2007-06-13T06:39:15Z")
On 6/12/07, DBC User [removed\_email\_address@domain.invalid](mailto:removed_email_address@domain.invalid) wrote: > in this I am expecting the test to run against the development > database as only development database has the table. But when I ran, > it is failing because the test table is not found. Why is model > testing is pointing to test database anyway? > > Thanks. > > > Why are you expecting it to run against the development database? It will run against whatever database is specified for the test environment in database.yml. Take a look at [http://glu.ttono.us/articles/2006/05/22/guide-environments-in-rails-1-1](http://glu.ttono.us/articles/2006/05/22/guide-environments-in-rails-1-1) to understand Rails environments. Pat ---
**Author:** ![DBC\_User](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/d/ee7513/32.png) [@DBC\_User](https://www.ruby-forum.com/u/DBC_User)\ **Post date:** [June 13, 2007, 5:01pm UTC](https://www.ruby-forum.com/t/unit-test-question/106674/3 "2007-06-13T17:01:28Z")
On 6/13/07, DBC User [removed\_email\_address@domain.invalid](mailto:removed_email_address@domain.invalid) wrote: > Hi, > > I have a model and I am trying to do the testing. So my test code is > something like the following > > def test\_first > assert\_nil Person.find(1) > end This particular test isn’t going to work. find(id) will raise an exception if the row can’t be found. If it can be found, it will return the Person object. It will never return nil. So you have a couple of options: 1. Use “assert\_raise”: 2. Use find\_by\_id, which _will_ return nil if the row isn’t found: > in this I am expecting the test to run against the development > database as only development database has the table. But when I ran, > it is failing because the test table is not found. Why is model > testing is pointing to test database anyway? It’s because your test file includes “test\_helper.rb” (as it should), which sets the RAILS\_ENV variable to “test”. You’re fighting Rails here. If you create an empty test database and then run “rake test”, it will copy the development structure to the test database and run the tests over there for you. The reason for a separate test database is that you want a controlled set of data to work with. The data in your development database is constantly changing, so your test results wouldn’t be stable. ---
**Author:** ![DBC\_User](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/d/ee7513/32.png) [@DBC\_User](https://www.ruby-forum.com/u/DBC_User)\ **Post date:** [June 14, 2007, 5:36am UTC](https://www.ruby-forum.com/t/unit-test-question/106674/4 "2007-06-14T05:36:22Z")
DBC User wrote: > testing is pointing to test database anyway? > Because manual and automated tests should use different databases. When you manually test, you need a generally sloppy set of records to noodle around with. Suppose you set up a given Person record, such as “Mr. Frootloop”. Each time you change the code and manually test, you want the same record to come back in the same state as you left it. You should run your automated tests much more often than you manually test. Your automated should have an exact set of data fixtures, and you don’t need the tests to die just because you manually changed a record. So you need your tests to erase the entire database each time they run. Put these rules together, and you need two databases. Your “manual test fixture”, Mr Frootloop, will not disappear just because you ran the automated tests. Tip: Configure the automated tests to run automatically, each time you change the code. Then get in the habit of saving your files deliberately, and trying to predict the result of each test run. If you guess wrong, you can always Undo your last edit until the tests pass again. This is a lot better than endlessly debugging! – Phlip [Test Driven Ajax (on Rails) [Book]](http://www.oreilly.com/catalog/9780596510657/) “Test Driven Ajax (on Rails)” assert\_xpath, assert\_javascript, & assert\_ajax