One way to narrow down a problem when debugging a test is to add logging with e.g. fmt.Printf().
The problem with this approach is lack of selectivity: imagine you have 100 tests and only 1 test fails. For debugging the issue you only need to see logs when executing that 1 test but you’ll drown in log output from all 100 tests.
My solution: control logging state with global variable verboseLog and allow toggling this flag per test.
Something like this:
var (
verboseLog = false
)
func myCodeWithBugs(s string) {
if verboseLog {
fmt.Printf("s: %s\n", s)
}
...
}
func TestMyCode(t *testing.T) {
var tests = []struct {
... test fields
debug bool
}{
{ ..., false }, // without verbose logging
{ ..., true }, // with verbose logging
}
for _, test := range tests {
verboseLog = test.debug
...
}
}