Fuzzing proactively tests input-driven code by expanding valid examples into generated cases, exposing edge conditions that ordinary tests may miss. Go includes fuzz testing in its standard testing library. section 8.0 section 8.7.2
A Go fuzz test typically:
- Defines a function whose name begins with Fuzz and accepts *testing.F. section 8.7.2
- Adds seed examples to a corpus with f.Add. section 8.7.2
- Runs the target with generated strings, runes, or other expanded inputs and compares the result against an expected behavior. section 8.7.2
- Uses go test -fuzz=. with a time limit, then preserves minimized failing inputs for reproduction. section 8.7.2
Finding a Unicode bug
A fuzz test compared a custom Contains function with strings.Contains across seeded cases such as ordinary words, empty strings, and rune values. Fuzzing found a failure caused by converting the input string to []byte, which mishandled a multibyte character. Iterating over the string as runes fixed the defect; the minimized input could then be rerun as a regression case. section 8.7.2
Listing 8.38 xstrings_test.go: Fuzz contains function
func FuzzFindInString(f *testing.F) {
cases := []struct { #1
word string
char rune
}{
{
word: "hello world",
char: 'e',
},
{
word: "goodbye",
char: 'e',
},
{
word: "",
char: 'o',
},
{
word: "a longer example",
char: 'z',
},
}
for _, ca := range cases { #2
f.Add(ca.word, ca.char)
}
f.Fuzz(func(t *testing.T, in string, r rune) { #3
contains := xstring.Contains(in, r)
if strings.Contains(in, string(r)) != contains { #4
t.Errorf("failed to find '%b' in %s", r, in)
}
})
}
Shows the seed corpus, f.Add calls, fuzz callback, and comparison against strings.Contains. section 8.7.2
