# Strange behaviour in Unit Test **URL:** \ **Category:** Rails\ **Created:** [February 8, 2007, 11:02pm UTC](https://www.ruby-forum.com/t/strange-behaviour-in-unit-test/84975 "2007-02-08T23:02:20Z")\ **Posts on this page:** 7\ **Page:** 1
**Author:** ![Joram](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/j/e8c25b/32.png) [@Joram](https://www.ruby-forum.com/u/Joram)\ **Post date:** [February 8, 2007, 11:02pm UTC](https://www.ruby-forum.com/t/strange-behaviour-in-unit-test/84975/1 "2007-02-08T23:02:20Z")
Hi, I’m having a weird problem in one of my unit tests. this code generates an assertion error on the second line assert !@valid\_category.update\_attributes(:name =\> nil) assert\_equal old\_name, @valid\_category.name However, when I write the following, the error is no more… assert !@valid\_category.update\_attributes(:name =\> nil) @valid\_category.reload assert\_equal old\_name, @valid\_category.name This sure can’t be normal? I tried to write out the name of the @valid\_category between the 2 lines, and its empty. Another question: Is it correct that the update\_attribute method bypasses validation? According to the API this is the fact, but why would that be, updates should also be valid?? Thanks! Joram ---
**Author:** ![Joram](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/j/e8c25b/32.png) [@Joram](https://www.ruby-forum.com/u/Joram)\ **Post date:** [February 8, 2007, 11:07pm UTC](https://www.ruby-forum.com/t/strange-behaviour-in-unit-test/84975/2 "2007-02-08T23:07:20Z")
That is normal behavior. #update\_attributes actually does update the attributes, but returns on whether the save to the database was successful or not. One reason it’s done this way is to allow for field re-filling on forms when there are errors in the form. Jason ---
**Author:** ![Joram](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/j/e8c25b/32.png) [@Joram](https://www.ruby-forum.com/u/Joram)\ **Post date:** [February 8, 2007, 11:08pm UTC](https://www.ruby-forum.com/t/strange-behaviour-in-unit-test/84975/3 "2007-02-08T23:08:20Z")
Sorry, hit the send button too quickly. As to your second question, update\_attributes does not bypass validations. Nothing in AR bypasses the validations unless you specifically tell it to do so. Jason ---
**Author:** ![Joram](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/j/e8c25b/32.png) [@Joram](https://www.ruby-forum.com/u/Joram)\ **Post date:** [February 8, 2007, 11:17pm UTC](https://www.ruby-forum.com/t/strange-behaviour-in-unit-test/84975/4 "2007-02-08T23:17:21Z")
Thanks for the information, Jason, it does make sense to me now. About the validation, I found this in the API: "update\_attribute(name, value) Updates a single attribute and saves the record. This is especially useful for boolean flags on existing records. Note: This method is overwritten by the Validation module that’ll make sure that updates made with this method doesn’t get subjected to validation checks. Hence, attributes can be updated even if the full object isn’t valid. " I’m getting an MySql exception when I use this method, stating that I want to insert a null value into a column where null is not allowed. ---
**Author:** ![Joram](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/j/e8c25b/32.png) [@Joram](https://www.ruby-forum.com/u/Joram)\ **Post date:** [February 8, 2007, 11:31pm UTC](https://www.ruby-forum.com/t/strange-behaviour-in-unit-test/84975/5 "2007-02-08T23:31:22Z")
> I’m getting an MySql exception when I use this method, stating that I > want to insert a null value into a column where null is not allowed. Yes, update\_attribute does bypass validation. You are getting an error because of a null restriction on the column in MySQL. If it wasn’t for that the nil name would have persisted. Only use update attribute if you don’t want validations to run. Aaron ---
**Author:** ![Joram](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/j/e8c25b/32.png) [@Joram](https://www.ruby-forum.com/u/Joram)\ **Post date:** [February 8, 2007, 11:25pm UTC](https://www.ruby-forum.com/t/strange-behaviour-in-unit-test/84975/6 "2007-02-08T23:25:22Z")
> assert !@valid\_category.update\_attributes(:name =\> nil) > assert\_equal old\_name, @valid\_category.name When an update fails validation the changes that you made are still in memory, they just don’t persist to the database. Think about when you fill out a form, click submit, and an validation error occurs. All the values you entered are returned and populate the form so you don’t have to enter them again. @valid\_category.valid? =\> tells you if the current version is valid puts @valid\_category.errors.inspect =\> to see the validation errors You can even do something like this to check that the error you expected occurred: assert\_equal( 1, @valid\_category.errors.length ) assert\_equal( ‘name must have a value’, @valid\_category.errors.on(:name)) > assert !@valid\_category.update\_attributes(:name =\> nil) > @valid\_category.reload > assert\_equal old\_name, @valid\_category.name This works because reload does exactly that, it reloads @valid\_category with the version from the database. Your update\_attributes failed validation so the database version was left unchanged. Aaron ---
**Author:** ![Joram](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/j/e8c25b/32.png) [@Joram](https://www.ruby-forum.com/u/Joram)\ **Post date:** [February 8, 2007, 11:38pm UTC](https://www.ruby-forum.com/t/strange-behaviour-in-unit-test/84975/7 "2007-02-08T23:38:23Z")
Thank you Aaron and Jason for your quick reply! It’s all clear to me now 🙂