Perf: Optimize substring_index via single-byte fast path and direct indexing#19590
Merged
alamb merged 5 commits intoapache:mainfrom Jan 6, 2026
Merged
Perf: Optimize substring_index via single-byte fast path and direct indexing#19590alamb merged 5 commits intoapache:mainfrom
substring_index via single-byte fast path and direct indexing#19590alamb merged 5 commits intoapache:mainfrom
Conversation
uzqw
reviewed
Jan 1, 2026
uzqw
left a comment
uzqw
left a comment
There was a problem hiding this comment.
Hi! I noticed some Clippy warnings when running locally:
cargo clippy -p datafusion-functions --all-targets --all-features
| match string.get(string.len().saturating_sub(length)..) { | ||
| Some(substring) => builder.append_value(substring), | ||
| None => builder.append_null(), | ||
| if n > 0 { |
There was a problem hiding this comment.
this else { if .. } block can be collapsed
|
|
||
| for batch_size in batch_sizes { | ||
| group.bench_function( | ||
| &format!("substr_index_{}_single_delimiter", batch_size), |
There was a problem hiding this comment.
variables can be used directly in the format! string
| ); | ||
|
|
||
| group.bench_function( | ||
| &format!("substr_index_{}_long_delimiter", batch_size), |
There was a problem hiding this comment.
variables can be used directly in the format! string
Contributor
Author
|
Contributor
Author
|
Hi @Jefffrey , could you please help trigger the workflows for this PR? I'd appreciate a review when you have a moment. Thanks! |
Jefffrey
approved these changes
Jan 2, 2026
Contributor
|
run benchmarks |
|
🤖 |
|
🤖: Benchmark completed Details
|
Contributor
|
run benchmark substr_index |
This comment was marked as outdated.
This comment was marked as outdated.
alamb
reviewed
Jan 5, 2026
| match string.get(..length) { | ||
| Some(substring) => builder.append_value(substring), | ||
| None => builder.append_null(), | ||
| let result_idx = if delimiter.len() == 1 { |
Contributor
There was a problem hiding this comment.
I double checked that this checks for length in bytes
https://doc.rust-lang.org/std/string/struct.String.html#method.len
| T::Native: OffsetSizeTrait, | ||
| { | ||
| let mut builder = StringBuilder::new(); | ||
| let num_rows = string_array.len(); |
Contributor
There was a problem hiding this comment.
If you wanted to make this really fast, you could also implement special case code for the common cases where delimiter and count were scalar values -- I suspect you could make it quite fast.
Contributor
Author
There was a problem hiding this comment.
Good point! I’ll try adding a special-case implementation for that. Thanks!
|
🤖 |
|
🤖: Benchmark completed Details
|
Contributor
|
Onwards 🚀 |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
6 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Which issue does this PR close?
Rationale for this change
This PR improves the performance of the
substring_indexfunction by optimizingdelimiter search and substring extraction:
.,,), avoiding UTF-8 pattern matching overhead.match_indices/rmatch_indices.What changes are included in this PR?
delimiter.len() == 1using byte-based search.match_indicesandrmatch_indicesfor more efficient positioning.Benchmarks
Are these changes tested?
Are there any user-facing changes?
No.