feat(random): use crypto/rand for random string generator

[ichigozero]

The current implementation for random package relied on globally seeded value.
If any other part of the programs call rand.Seed, the string generate by random package will also be affected.

To fix this issue, instead of calling rand.Seed during initialization, NewSource is called and the return value is stored inside Random struct.

I made changes by referring to the source code from the link below.
https://github.com/labstack/echo/blob/e6b96f8873fed46e71e0d34cddb81c533167f954/middleware/util.go#L69

I also did some benchmarks before and after the update.

Before

goos: linux
goarch: amd64
pkg: github.com/labstack/gommon/random
cpu: Intel(R) Core(TM) i7-4702MQ CPU @ 2.20GHz
BenchmarkRandomString-8   	 1472410	       806.7 ns/op	      64 B/op	       2 allocs/op
BenchmarkRandomString-8   	 1478504	       812.3 ns/op	      64 B/op	       2 allocs/op
BenchmarkRandomString-8   	 1484530	       806.4 ns/op	      64 B/op	       2 allocs/op
BenchmarkRandomString-8   	 1470031	       801.5 ns/op	      64 B/op	       2 allocs/op
BenchmarkRandomString-8   	 1492831	       809.4 ns/op	      64 B/op	       2 allocs/op

After (with sync.Mutex)

goos: linux
goarch: amd64
pkg: github.com/labstack/gommon/random
cpu: Intel(R) Core(TM) i7-4702MQ CPU @ 2.20GHz
BenchmarkRandomString-8   	 1507052	       771.5 ns/op	      64 B/op	       2 allocs/op
BenchmarkRandomString-8   	 1549957	       775.3 ns/op	      64 B/op	       2 allocs/op
BenchmarkRandomString-8   	 1519629	       774.4 ns/op	      64 B/op	       2 allocs/op
BenchmarkRandomString-8   	 1540794	       774.6 ns/op	      64 B/op	       2 allocs/op
BenchmarkRandomString-8   	 1548172	       777.5 ns/op	      64 B/op	       2 allocs/op

After (with sync.Pool)

goos: linux
goarch: amd64
pkg: github.com/labstack/gommon/random
cpu: Intel(R) Core(TM) i7-4702MQ CPU @ 2.20GHz
BenchmarkRandomString-8   	 2655366	       446.7 ns/op	     112 B/op	       3 allocs/op
BenchmarkRandomString-8   	 2629251	       451.6 ns/op	     112 B/op	       3 allocs/op
BenchmarkRandomString-8   	 2613632	       456.5 ns/op	     112 B/op	       3 allocs/op
BenchmarkRandomString-8   	 2638729	       458.5 ns/op	     112 B/op	       3 allocs/op
BenchmarkRandomString-8   	 2577884	       462.6 ns/op	     112 B/op	       3 allocs/op

[aldas]

maybe https://github.com/labstack/echo/blob/e6b96f8873fed46e71e0d34cddb81c533167f954/middleware/util.go#L69 would be suitable here also.

[ichigozero]

maybe https://github.com/labstack/echo/blob/e6b96f8873fed46e71e0d34cddb81c533167f954/middleware/util.go#L69 would be suitable here also.

Thank you for the comment. I will take a look the link you mentioned and update the PR accordingly.

[ichigozero]

@aldas I have updated the source code as you suggested. Please let me know what you think.

[ichigozero]

@aldas Hi, just wondering what is the progress for this PR?

[aldas]

LGTM