For example, this trait enables syntax such as the following test registration in FlatSpec and FixtureFlatSpec:

 "A Stack (when empty)" should "be empty" in { ... }
                        ^
 

It also enables syntax such as the following shared test registration in FlatSpec and FixtureFlatSpec:

 "A Stack (with one item)" should behave like nonEmptyStack(stackWithOneItem, lastValuePushed)
                           ^
 

In addition, it supports the registration of subject descriptions in WordSpec and FixtureWordSpec, such as:

 "A Stack (when empty)" should { ...
                        ^
 

And finally, it also supportds the registration of subject descriptions with after words in WordSpec and FixtureWordSpec. For example:

    def provide = afterWord("provide")

   "The ScalaTest Matchers DSL" should provide {
                                ^
 

The reason this implicit conversion is provided in a separate trait, instead of being provided directly in FlatSpec, WordSpec, FixtureFlatSpec, and FixtureWordSpec, is because an implicit conversion provided directly would conflict with the implicit conversion that provides should methods on String in the ShouldMatchers trait. By contrast, there is no conflict with the separate ShouldVerb trait approach, because:

1. FlatSpec, WordSpec, FixtureFlatSpec, and FixtureWordSpec mix in ShouldVerb directly, and
2. ShouldMatchers extends ShouldVerb, overriding the convertToStringShouldWrapper implicit conversion function.

So whether or not a FlatSpec, WordSpec, FixtureFlatSpec, or FixtureWordSpec mixes in ShouldMatchers, there will only be one implicit conversion in scope that adds should methods to Strings.