Performance comparison of Golang string assembling method

This is a creation in Article, where the information may have evolved or changed.

There are many ways to assemble strings in the go language, so the question is, which is the best performance?

The following code, respectively, compares the FMT. Sprintf,string +,strings. Join,bytes. Buffer, the method is to loop several times the total time of comparison. The results of operating under Ubuntu 14.04 under VMware are as follows, for reference only:

• FMT. Sprintf and strings. Join speed is quite
• string + one times faster than the above
• bytes. and the Buffer is about 400-500 times faster than the last person.
• If a bytes is temporarily declared within a loop . Buffer to use, will be 50% slower than persistent, but still very fast

The test code is as follows:

Package Mainimport ("bytes" "FMT" "Strings" "Time") func benchmarkstringfunction (n int, index int) (d time.    Duration) {V: = "ni shuo wo shi bu shi tai wu liao le a?" var s string var buf bytes. Buffer T0: = time. Now () for I: = 0; I < n; i++ {Switch Index {case 0://FMT. Sprintf s = fmt. Sprintf ("%s[%s]", S, v) Case 1://string + S = s + "[" + V + "]" Case 2://Strings. Join s = strings. Join ([]string{s, "[", V, "]"}, "") Case 3://Temporary bytes. Buffer B: = bytes.         buffer{} b.writestring ("[") b.writestring (v) b.writestring ("]") s = b.string () Case 4://Stable bytes. Buffer buf. WriteString ("[") buf. WriteString (v) buf. WriteString ("]")} if I = = n-1 {if index = = 4 {//for stable bytes. Buffer s = buf. String ()} FMT. Println (Len (s))//conSume s to avoid compiler optimization}} T1: = time. Now () d = t1. Sub (t0) fmt. Printf ("Time of the-=%v\n", index, D) return D}func main () {k: = 5 D: = [5]time. duration{} for I: = 0; I < K; i++ {D[i] = Benchmarkstringfunction (10000, i)} for I: = 0; i < k-1; i++ {fmt. Printf ("by-%6.1f times of the-%d\n", I, float32 (D[i])/float32 (d[k-1]), k-1)}}

One of the results was as follows:

410000time of way(0)=1.199641573s410000time of way(1)=568.716669ms410000time of way(2)=1.197077483s41time of way(3)=2.277063ms410000time of way(4)=1.398864msway 0 is  857.6 times of way 4way 1 is  406.6 times of way 4way 2 is  855.7 times of way 4way 3 is    1.6 times of way 4