What is means of the "Bencher.bytes" ?
If the concept of throughput is relevant to what you're benching (which is the case for parsing strings, to an extent) you can set the bencher.bytes
field to the amount of data used / consumed per iteration, and at the end the benchmark report will print a throughput in bytes/second or somesuch.
与恶龙缠斗过久,自身亦成为恶龙;凝视深渊过久,深渊将回以凝视…