That microbench is taken out of context, «faster» there does not actually mean anything.
In an actual application where you would actually use the result instead of throwing it out immediately, char-by-char concatenation would consume up to an order of magnitude more memory and slow everything down.
My code to repoduce that is in a comment below, in another thread here.
Yep. Also remember your usage scenario.
Are you in a browser ?
How large is the string ?
How many strings ?
Because if you're only reversing a 100 chars string, even 1000 chars, the implementation (amongst thoses proposed in this thread) doesn't matter much performance wise (memory, duration, cpu cycles)...
For clarity and code comprehension, it's something else sure ^
This is a showcase, not a contest. Also, there is no point using reverse() to do a reverse operation. It's the same as trying to explain some term using the same term in a sentence.
This would be best solution. Yes it’s a re implementation of array reverse but it skips the split and join and only uses 2 bytes of memory and n/2 time
regular for loop will beat all these methods by huge margin
function reverse(s) {
var o = '';
for (var i = s.length - 1; i >= 0; i--)
o += s[i];
return o;
}
For further actions, you may consider blocking this person and/or reporting abuse
We're a place where coders share, stay up-to-date and grow their careers.
Have you looked at the performance impact of each method?
Performace of reversing »Hello«?
the first is the most straightforward, while the second is going to be the fastest by far compared to #1 and #3
No, it isn't. Symbol-by-symbol string concatenation comes with significant penalties outside of microbenchmarks that immediately throw the result out.
Perhaps, but reduce is more than a power of magnitude faster than split reverse join. It’s on jsperf
That microbench is taken out of context, «faster» there does not actually mean anything.
In an actual application where you would actually use the result instead of throwing it out immediately, char-by-char concatenation would consume up to an order of magnitude more memory and slow everything down.
My code to repoduce that is in a comment below, in another thread here.
I think something like this would actually be the most performant:
Wouldn't it still take quadratic time?
My attempt:
That basically reimplements Array#reverse() from JS land at the second step.
Array#reverse() operates in-place, so there is no benefit here — you are doing exactly what
.reverse()
already does.This approach is unlikely to be more performant than the built-in method, and just takes more code to write and maintain.
Yep. Also remember your usage scenario.
Are you in a browser ?
How large is the string ?
How many strings ?
Because if you're only reversing a 100 chars string, even 1000 chars, the implementation (amongst thoses proposed in this thread) doesn't matter much performance wise (memory, duration, cpu cycles)...
For clarity and code comprehension, it's something else sure ^
This is a showcase, not a contest. Also, there is no point using
reverse()
to do a reverse operation. It's the same as trying to explain some term using the same term in a sentence.This would be best solution. Yes it’s a re implementation of array reverse but it skips the split and join and only uses 2 bytes of memory and n/2 time
Oh it’s not what I thought it was
regular for loop will beat all these methods by huge margin
function reverse(s) {
var o = '';
for (var i = s.length - 1; i >= 0; i--)
o += s[i];
return o;
}