performance opt: using copy instead of append

[wathenjiang]

Please see https://gist.github.com/spongecaptain/c5eedde03caaaa5f3de0331a50d24baa

Using copy instead of append brings performance improvements.

Original version benchmark test result:

$ go test -v -run=none -bench=. -cpu 1,2,4,8,16 -benchmem

goos: darwin
goarch: arm64
pkg: github.com/embano1/memlog
BenchmarkLog_write
BenchmarkLog_write              12658945                83.79 ns/op           89 B/op          1 allocs/op
BenchmarkLog_write-2            14599183                81.88 ns/op           89 B/op          1 allocs/op
BenchmarkLog_write-4            14565168                81.85 ns/op           89 B/op          1 allocs/op
BenchmarkLog_write-8            14652230                80.89 ns/op           89 B/op          1 allocs/op
BenchmarkLog_write-16           14582964                81.61 ns/op           89 B/op          1 allocs/op
BenchmarkLog_read
BenchmarkLog_read               23078679                51.61 ns/op           32 B/op          1 allocs/op
BenchmarkLog_read-2             23851464                49.38 ns/op           32 B/op          1 allocs/op
BenchmarkLog_read-4             23998719                50.11 ns/op           32 B/op          1 allocs/op
BenchmarkLog_read-8             23799724                50.11 ns/op           32 B/op          1 allocs/op
BenchmarkLog_read-16            23617048                49.77 ns/op           32 B/op          1 allocs/op
PASS
ok      github.com/embano1/memlog       12.563s

Using copy version:

$ go test -v -run=none -bench=. -cpu 1,2,4,8,16 -benchmem

goos: darwin
goarch: arm64
pkg: github.com/embano1/memlog
BenchmarkLog_write
BenchmarkLog_write              15381063                78.47 ns/op           89 B/op          1 allocs/op
BenchmarkLog_write-2            15534843                77.78 ns/op           89 B/op          1 allocs/op
BenchmarkLog_write-4            15860680                75.88 ns/op           89 B/op          1 allocs/op
BenchmarkLog_write-8            15975698                75.16 ns/op           89 B/op          1 allocs/op
BenchmarkLog_write-16           15493441                76.15 ns/op           89 B/op          1 allocs/op
BenchmarkLog_read
BenchmarkLog_read               25497036                47.53 ns/op           32 B/op          1 allocs/op
BenchmarkLog_read-2             26456144                45.67 ns/op           32 B/op          1 allocs/op
BenchmarkLog_read-4             25824014                45.50 ns/op           32 B/op          1 allocs/op
BenchmarkLog_read-8             26311268                45.73 ns/op           32 B/op          1 allocs/op
BenchmarkLog_read-16            26083293                45.62 ns/op           32 B/op          1 allocs/op
PASS
ok      github.com/embano1/memlog       13.762s

[codecov-commenter]

Codecov Report

Merging #38 (5fd6552) into main (d374463) will increase coverage by 0.09%.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##             main      #38      +/-   ##
==========================================
+ Coverage   87.21%   87.31%   +0.09%     
==========================================
  Files           7        7              
  Lines         266      268       +2     
==========================================
+ Hits          232      234       +2     
  Misses         19       19              
  Partials       15       15              

Impacted Files	Coverage Δ
memlog.go	85.18% <100.00%> (+0.27%)	⬆️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update d374463...5fd6552. Read the comment docs.

[embano1]

You're the best! Thx a ton and keep em coming :)