HappenLee opened a new pull request, #69223:
URL: https://github.com/apache/doris/pull/69223
### What problem does this PR solve?
Issue Number: N/A
Related PR: N/A
Problem Summary:
String functions incur avoidable work on mostly ASCII columns and constant
arguments. This change dispatches ASCII rows in mixed columns to the existing
case-conversion fast paths, adds a table-based translate path for constant
ASCII mappings, appends concat_ws inputs directly, reuses StringOP::fast_repeat
with reserved output storage, and caches constant parse_url part lookup. It
also adds AArch64 NEON implementations for contain_one/contain_zero with
guard-page and boundary tests.
Non-ASCII case conversion retains full ICU root-locale mappings, including
expansions and contextual casing. parse_url still propagates NULL before
reporting an invalid URL part. Unit tests cover mixed translate
replacement/deletion and NULL URL/key inputs with invalid parts. Benchmark
speedups have not been measured in this environment.
### Release note
Reduce processing overhead in string case conversion, translate, concat_ws,
repeat, parse_url, and AArch64 byte-presence checks while preserving SQL
results.
### Check List (For Author)
- Test:
- Unit tests added; execution is pending. Attempted the relevant BE ASAN
tests with `run-be-ut.sh -j 48 --run --filter=...`, but the build stopped
before running tests because the local third-party package lacks
`libpaimon_c.a`.
- Passed `build-support/check-format.sh` with clang-format 16.0.6,
`build-support/check-build-hygiene.sh`, and `git diff --check`.
- Attempted `build-support/run-clang-tidy.sh`; analysis is incomplete
because the failed build did not produce the required PCH, and the local
clang-tidy cannot resolve the compiler resource header `stddef.h`.
- AArch64 execution and performance benchmarks were not run.
- Behavior changed: No SQL semantic changes intended.
- Does this need documentation: No.
### Check List (For Reviewer who merge this PR)
- [ ] Confirm the release note
- [ ] Confirm test cases
- [ ] Confirm document
- [ ] Add branch pick label
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
