SSpec is a behavior driven development (BDD) framework for Swift and it offers awesome alternative to standard XCTest.
If you worked with behavior driven development (BDD) before, SSpec should be already familiar to you. And it's simple to understand and get started for those who never worked with any BDD frameworks before.
Add SSpec as package dependency in your project's Package.swift
file:
.package(url: "https://github.com/dimakura/SSpec", from: "0.2.3")
You should also put it under test target's dependencies:
.testTarget(
name: "MyPackageTests",
dependencies: [
"SSpec",
// other dependencies
]),
Package manager downloads SSpec for you during build:
swift build
Do not forget to import SSPec to use it:
import SSpec
You can easily understand a lot about SSpec looking at this simple example:
SSpec.run { // 1
it("2 + 2 = 4") { // 2
expect(2 + 2).to.eq(4) // 3
}
}
XCTAssert(SSpec.hasErrors == false)
Standard way to run SSpec is to use SSpec.run
method as this is done at the
first line of our example. Inside closure for this method we define our test
examples.
The first (and only) test example is defined on the second line.
It has title ("2 + 2 = 4"
) and one more closure.
Inside test example's closure at line three, you can see what this example
actually tries to test. It makes sure that 2 + 2
is indeed 4
.
You can place as many examples inside body of SSpec.run {...}
as needed.
You should usually create single SSpec.run {...}
and define all tests inside.
SSpec runs all of the examples and checks their correctness.
If any of the example fails, SSpec will print detailed report of what went wrong, so you can easily locate source of the problem.
Example output with default SSpec reporter looks like this:
To detect failures programmatically, you can use SSpec.hasErrors
property,
which is false
by default, but becomes true
once there's at least one
example failing in your specs.
You can use this property to report failure back to XCTest
:
XCTAssert(SSpec.hasErrors == false)
This is not strictly necessary. But when your CI relies on XCTest's failure, it's a handy way to detect errors.
Two global functions describe
and context
are helpful for grouping similar
tests. They are also help to document your code.
SSpec.run {
describe("Person") {
context("with high salary") {
let person = Person(salary: .High)
it("is rich") {
expect(person.isRich).to.beTrue
}
}
context("with low salary") {
let person = Person(salary: .Low)
it("is not rich") {
expect(person.isRich).to.beFalse
}
it("is poor") {
expect(person.isPoor).to.beTrue
}
}
}
}
You can use describe
and context
interchangeably.
As a rule programmers use describe
for grouping examples at high-level,
splitting them down to more detailed examples with context
s.
If several tests use shared variables or need the same initialization code, you can conveniently use before hook. Before hook runs before each test.
If you need to cleanup after running each test you can do this using after hook.
describe("Event") {
var event: Event!
before {
// This code runs before each example
event = Event(priority: .High, dueDate: Date.tomorrow())
event.save()
}
it("is urgent") {
expect(event.urgent).to.beTrue
}
it("is 1 day away") {
expect(event.daysLeft).to.eq(1)
}
after {
// This code runs after each example
Database.clean()
}
}
Note that if there are before hooks at different levels, they all get executed starting from top-level. For after hook execution starts with the lowest level.
describe("Level 1") {
before {
// this runs first
}
describe("Level 2") {
before {
// this runs second
}
it("Example") {
// this runs after two before hooks
}
after {
// this runs first after the "Example"
}
}
after {
// this runs last
}
}
All values can be tested on presence with beNil
matcher:
it("is nil") {
expect(value).to.beNil
}
To test boolean values, use beTrue
and beFalse
matchers:
it("is true") {
expect(true).to.beTrue
}
it("is false") {
expect(false).to.beFalse
}
TODO:
TODO:
TODO:
TODO:
Swift is strongly typed language. SSpec uses types for extending matchers.
Suppose you created custom class Person
:
class Person {
let name: String
let salary: Double
init(_ name: String, _ salary: Double) {
self.name = name
self.salary = salary
}
var isRich: Bool {
return salary > 50000
}
var isPoor: Bool {
return !isRich
}
}
and want to have your spec to like this:
describe("Person") {
let jack = Person("Jack", 100000)
let jane = Person("Jane", 20000)
describe(jack.name) {
it("is rich") {
expect(jack).to.beRich
}
}
describe(jane.name) {
it("is poor") {
expect(jane).to.bePoor
}
}
}
All you need, is to extend SSExcept<T>
class:
extension SSExpect where T == Person {
var beRich: Void {
// Note that `assert` takes into account negation status.
// If you need to know negation status use `isNegate` variable.
assert(
value!.isRich, // `value` is of type `Person?`
error: "Expected \(value!.name) to be rich", // There's also standard `valueStr` variable and
// `toString(value:)` function you can use.
errorNegate: "Expected \(value!.name) to be poor"
)
}
var bePoor: Void {
assert(
value!.isPoor,
error: "Expected \(value!.name) to be poor",
errorNegate: "Expected \(value!.name) to be rich"
)
}
}
By default SSpec reports progress with dots.
There are other options available:
SSpec.reporter = .Dot // default reporter
SSpec.reporter = .Spec // most document-like reporter
SSpec.reporter = .Progress // progress bar reporter
If you are interested in contributing to SSpec please read about it here.