Improve benchmark for ltrim #12513
Improve benchmark for ltrim #12513
Conversation
Rachelint
force-pushed
the
trim-bench
branch
from
aa24474 to
8162cbc
Compare
Rachelint
force-pushed
the
trim-bench
branch
from
8162cbc to
a5625aa
Compare
|
@alamb @Kev1n8 The benchmark is ready, and the result in my local can see in #12395 (comment) |
There was a problem hiding this comment.
Thank you @Rachelint and @Kev1n8 -- looks good to me; I had one question, but I don't think there is anything that would prevent this PR from merging.
| let lens = vec![remaining_len; size]; | ||
| let string_iter = lens.into_iter().map(|len| { |
There was a problem hiding this comment.
I don't understand the use of vec! here - isn't this the same as creating size strings of lenght up to remaining_len?
Maybe this could be somehthing like
let string_iter = (0..size).map(|len|{ ?
There was a problem hiding this comment.
Yes, I want to create size strings, and each one has remaining_len lenght.
🤔 Maybe it is a bit confused, I have improved it and added some comments.
| } | ||
| } | ||
|
|
||
| /// Create args for the ltrim benchmark |
Co-authored-by: Andrew Lamb <[email protected]>
|
Thank you for the comments and benchmarks @Rachelint -- this PR now looks even better 🚀 |
Which issue does this PR close?
Part of #12387
Rationale for this change
We need a benchmark to see the improvements of the new string view trim introduced in #12395 .
What changes are included in this PR?
Improve the ltrim benchmark, so that we can use it to see the improvements of the new string view trim.
Are these changes tested?
Test manually.
Are there any user-facing changes?
No.