A type switch selects a branch according to the dynamic type stored in an interface value. It suits functions that accept arbitrary values and need type-specific processing, such as converting numeric and string inputs before summing them. Write the switch as `switch val.(type)` and handle each supported type in its own case. section 13.1.1
Summing mixed values
A variadic `sum(v ...interface{}) float64` can range over its arguments and switch on each value's type. The example handles `int`, `int64`, `uint8`, and `string`; numeric values are converted to `float64`, while strings are parsed with `strconv.ParseFloat`. Calling `sum(a, b, c)` with `uint8(2)`, `int(37)`, and `"3.2"` produces `42.200000`. A `default` branch can report and ignore unsupported values. A named type such as `MyInt int64` does not match the `int64` case, so the example prints an unsupported-type message and leaves it out. section 13.1.1
Type switches differ from ordinary value switches: combining several types in one case complicates the value assigned inside that case, so one type per case is typical. When matching by underlying kind is more important than matching an exact type, reflection provides an alternative: `reflect.ValueOf(val).Kind()` can group `reflect.Int` and `reflect.Int64`, and named integer types then participate in the same branch. section 13.1.1
