【Go】单元测试

基本用法

• 测试文件名以_test结尾
• 函数名以Test开始

待测试代码

// split.go
package split

import "strings"

func Split(str string, sep string) []string {
    var ret []string
    index := strings.Index(str, sep)
    for index >= 0 {
        ret = append(ret, str[:index])
        str = str[index+len(sep):]
        index  = strings.Index(str, sep)
    }
    ret = append(ret, str)
    return ret
}

测试代码

// split_test.go
package split

import (
    "reflect"
    "testing"
)

func TestSplit(t *testing.T) {
    got := Split("a:b:c", ":")
    want := []string{"a", "b", "c"}
    if !reflect.DeepEqual(got, want) {
        t.Errorf("expected: %v, got: %v\n", want, got)
    }
}

到所在目录执行

go test -v

测试组

优化多个测试用例的代码

func TestSplitByGroup(t *testing.T) {
    type testCase struct {
        str string
        sep string
        want []string
    }

    testGroup := []testCase {
        {"babcbef", "b", []string{"", "a", "c", "ef"}},
        { "a:b:c", ":", []string{"a", "b", "c"}},
        {"abcef", "bc", []string{"a", "ef"}},
    }

    for _, tg := range testGroup {
        got := Split(tg.str, tg.sep)
        if !reflect.DeepEqual(got, tg.want) {
            t.Errorf("expected: %v, got: %v\n", tg.want, got)
        }
    }
}

子测试

用于区分测试组中，具体执行了哪个测试用例

func TestSplit3(t *testing.T) {
    type testCase struct {
        str  string
        sep  string
        want []string
    }

    testGroup := map[string]testCase{
        "case1": {str: "babcbef", sep: "b", want: []string{"", "a", "c", "ef"}},
        "case2": {str: "a:b:c", sep: ":", want: []string{"a", "b", "c"}},
        "case3": {str: "abcef", sep: "bc", want: []string{"a", "ef1"}},
    }

    for name, tg := range testGroup {
        t.Run(name, func(t *testing.T) {
            got := Split(tg.str, tg.sep)
            if !reflect.DeepEqual(got, tg.want) {
                t.Errorf("expected: %#v, got: %#v\n", tg.want, got)
            }
        })
    }
}

测试结果可以看到是case3结果出错了

image-20221119000527171

如果想跑某一个测试用例

go test -run=Split3/case1
go test -run=Split
// 如果run参数错了，找不到测试用例，会warning
// testing: warning: no tests to run

测试覆盖率

• 函数覆盖率100%
• 代码覆盖着60%

简单使用

go test -cover

image-20221119001017029

使用工具

go test -cover -coverprofile=cover.out
go tool cover -html=cover.out

image-20221119001337342

基准测试

• 函数名以Benchmark开始
• 必须执行b.N次

func BenchmarkSplit(b *testing.B) {
    for i := 0; i < b.N; i++ {
        Split("a:b:c", ":")
    }
}

运行

go test -bench=Split
go test -bench .

image-20221119141057822

看内存信息

go test -bench=Split -benchmem

image-20221119141221966

每次操作使用了三次内存分配

还可以重置时间

b.ResetTimer()

Test Main

当测试文件中有TestMain函数，执行go test就会会调用TestMain，否则会创建一个默认的TestMain；我们自定义TestMain时，需要手动调用m.Run()否则测试函数不会执行

func TestMain(m *testing.M) {
    fmt.Println("before code...")
    retCode := m.Run()
    fmt.Println("retCode", retCode)
    fmt.Println("after code...")
    os.Exit(retCode)
}