Find the cause of the linear increase in text benchmark execution time

[hackerwins]

Description:

At #246 we introduced benchmark tests.

However, it was not possible to complete the 250,000 text benchmark test. If we check the execution time for each section by 1000, we can see that it increases linearly.

    editing_bench_test.go:86: processing... 1000 833.432413ms
    editing_bench_test.go:86: processing... 2000 2.486363836s
    editing_bench_test.go:86: processing... 3000 3.766670666s
    editing_bench_test.go:86: processing... 4000 5.264808552s
    editing_bench_test.go:86: processing... 5000 7.45837556s
    ...

We need to use the profiler to see what is causing the execution speed to increase.

We can check performance bottlenecks in the browser with the following command:

$ make bench
$ go tool pprof ./test/bench/text-editing.prof
(pprof) web

Why:

• We can improve the running speed of Yorkie Text or Document.

[hackerwins]

In addition, Ticket.Key is a function that is called very frequently, so I think we should find a way to replace fmt.Sprintf.

[hackerwins]

Now it seems to be somewhat resolved, so I'll close it.