tl;dr
基本的にio.Writer.Write()を使用するのが高速なようです。
result
$ go test -bench . -benchmem
BenchmarkWrite-4 30000000 48.7 ns/op 16 B/op 1 allocs/op
BenchmarkWriteWithBytes-4 500000000 3.95 ns/op 0 B/op 0 allocs/op
BenchmarkFprintf-4 20000000 91.5 ns/op 0 B/op 0 allocs/op
BenchmarkWriteTo-4 100000000 10.0 ns/op 0 B/op 0 allocs/op
BenchmarkWriteWithBufferBytes-4 300000000 4.31 ns/op 0 B/op 0 allocs/op
source
package main_test
import (
"bytes"
"fmt"
"io"
"net/http"
"testing"
)
var s = "Hello, my world"
var bs = []byte(s)
var buf = bytes.Buffer{}
type NullWriter struct{}
func (w *NullWriter) Write(b []byte) (int, error) {
return len(b), nil
}
func BenchmarkWrite(b *testing.B) {
var w io.Writer = &NullWriter{}
b.ResetTimer()
for i := 0; i < b.N; i++ {
w.Write([]byte(s))
}
}
func BenchmarkWriteWithBytes(b *testing.B) {
var w io.Writer = &NullWriter{}
b.ResetTimer()
for i := 0; i < b.N; i++ {
w.Write(bs)
}
}
func BenchmarkFprintf(b *testing.B) {
var w io.Writer = &NullWriter{}
b.ResetTimer()
for i := 0; i < b.N; i++ {
fmt.Fprintf(w, s)
}
}
func BenchmarkWriteTo(b *testing.B) {
var w io.Writer = &NullWriter{}
b.ResetTimer()
for i := 0; i < b.N; i++ {
buf.WriteTo(w)
}
}
func BenchmarkWriteWithBufferBytes(b *testing.B) {
var w io.Writer = &NullWriter{}
b.ResetTimer()
for i := 0; i < b.N; i++ {
w.Write(buf.Bytes())
}
}