Package testkit provides all sorts of test helpers which take *testing.T
instance and report errors if something goes wrong. See GoDoc for the
documentation and examples.
In my opinion constantly checking errors while setting up a test reduces its readability.
Example test:
func Test_MyFunction(t *testing.T) {
// --- Given ---
fil, err := ioutil.TempFile(t.TempDir(), "test_*.txt")
if err != nil {
t.Fatal(err)
}
defer fil.Close()
content := []byte("line1\nline2\nend")
if _, err := fil.Write(content); err != nil {
t.Fatal(err)
}
// --- When ---
// Do some operations on the test file.
// --- Then ---
chk, err := os.Open(fil.Name())
if err != nil {
t.Fatal(err)
}
got, err := ioutil.ReadAll(chk)
if err != nil {
t.Fatal(err)
}
expected := []byte{0, 1, 2, 3}
if !bytes.Equal(expected, got) {
t.Errorf("expected %v got %v", expected, got)
}
}
The same test using testkit:
import (
kit "github.com/rzajac/testkit"
)
func Test_MyFunction(t *testing.T) {
// --- Given ---
content := []byte("line1\nline2\nend")
pth := kit.TempFileBuf(t, t.TempDir(), content)
// --- When ---
// Do some operations on the test file.
// --- Then ---
got := kit.ReadFile(t, pth)
expected := []byte{0, 1, 2, 3}
if !bytes.Equal(expected, got) {
t.Errorf("expected %v got %v", expected, got)
}
}
TempFileBuf creates and writes to temporary file from a buffer. Returns
path to created file. It registers cleanup function with t removing the created
file. Calls t.Fatal() on error. Similarly ReadFile is a wrapper
around ioutil.ReadFile() which calls t.Fatal() on error.
BSD-2-Clause