Non-allocating way to parse section of string as number?

General Usage Performance
, ,
1

Given a big string with numbers in it like

input = ";(,)<mul(595,110)~ #(select()-?who():mul(732,729)+/;%@mul(924,700)"

and I know at execution time exactly where a number is in the string, I can parse that section of the string like so

i = 10
parse(Int, input[i:i+2])

but this allocates

Benchmark: 2605 samples with 1264 evaluations
 min    19.896 ns (1 allocs: 32 bytes)
 median 23.248 ns (1 allocs: 32 bytes)
 mean   29.374 ns (1 allocs: 32 bytes, 0.08% gc time)
 max    7.418 μs (1 allocs: 32 bytes, 99.26% gc time)

which can (in my experiments) easily add up over a very big string and cause a slow down.

My solution was to just implement a very naive parser that doesn’t allocate

arr = codeunits(input) #outside the call to naiveparseint and used many times
function naiveparseint(arr, xstart, xend)
	output::Int = 0
	pow = 0
	for i in xend:-1:xstart
		output += (arr[i]-0x30)*10^pow
		pow += 1
	end
	return output
end

but it feels like there must be a better way to do this in Julia.

Any ideas?

2

Instead of pow = 0, *10^pow, pow +=1, you can do mult = 1, *mult, mult *= 10

3

Try

parse(Int, @view input[i:i+2])