Test Anything Protocol (TAP) for Swift

0. Add this project to your Package.swift.

import PackageDescription
let package = Package(
  // ...
  dependencies: [
    .Package(
      url: "https://github.com/dankogai/swift-tap.git", from: "0.6.0"
    )
  ],
  // ...
)

1. write [main.swift] like below:

import TAP
let test = TAP(tests:3)
test.ok(42+0.195 == 42.195, "42 + 0.195 == 42.195") // ok 1 - 42 + 0.195 == 42.195
test.eq(42+0.195,   42.195, "42 + 0.195 is 42.195") // ok 2 - 42 + 0.195 is 42.195
test.ne(42+0.195,   42,     "42 + 0.195 is not 42") // ok 3 - 42 + 0.195 is not 42
test.done() 

2. build and test like below:

$ swift run && prove .build/debug/main
./main .. ok   
All tests successful.
Files=1, Tests=2,  0 wallclock secs ( 0.02 usr +  0.00 sys =  0.02 CPU)
Result: PASS

./main actually prints the following:

1..3
ok 1 - 42 + 0.195 == 42.195
ok 2 - 42 + 0.195 is 42.195
ok 3 - 42 + 0.195 is not 42

You should specify the number of tests to run via TAP(tests:Int) or .plan(_:Int) but this is optional. But when you do so you cannot omit .done() which prints 1..number_of_tests that the TAP protocol demands.

ok 1 - 42 + 0.195 == 42.195
ok 2 - 42 + 0.195 is 42.195
ok 3 - 42 + 0.195 is not 42
1..2

Checks if predicate is true.

func ok(@autoclosure predicate:()->Bool, _ message:String = "")->Bool

Checks if actual == expected. Analogous to .ok(actual == expected) but it is more informative when it is not ok.

On the other hand both actual and expected must be Equatable. So:

• use .eq whenever you can.
• use .ok as a last resort.

public func eq<T:Equatable>(actual:T, _ expected:T, _ message:String = "")->Bool

As you can tell from the type signatures, .eq and .ne accepts arrays and dictionaries if they are equatable.

The opposite of .eq.

public func ne<T:Equatable>(actual:T, _ expected:T, _ message:String = "")->Bool

• Still way too fancy and resource-demanding for some occasions

• https://testanything.org
• https://en.wikipedia.org/wiki/Test_Anything_Protocol