Why assertEquals(a, b) instead of assert(a == b)?

Question

Why most (all?) unit test frameworks have large APIs with separate functions for specifying different types of boolean conditions (eg. assertEquals, assertNotEqual, etc) instead of using single assert function (or language construct) with desired boolean expression?

Answer 1

A simple assert will only throw AssertionError stating that the asserted conditation evaluated to false:

assert "foo".equals("boo")

java.lang.AssertionError: assertion failed

(not to mention assert string1 == string2 is incorrect due to reference comparison)

By passing both a and b the library can include them in the error message. Here: FEST assertions:

assertThat("foo").isEqualTo("boo");

//throws:
Exception in thread "main" org.junit.ComparisonFailure:
expected:<'[b]oo'> but was:<'[f]oo'>

Note that some languages are more powerful:

In Groovy (example from: Groovy 1.7 Power Assert):

a = 10
b = 9

assert 91 == a * b

yields:

Assertion failed: 

assert 91 == a * b
          |  | | |
          |  10| 9
          |    90
          false

    at ConsoleScript2.run(ConsoleScript2:4)

In Scala (ScalaTest) there is a special === operator:

assert(1 === 2)

yields 1 did not equal 2.