HappenLee opened a new pull request, #64930:
URL: https://github.com/apache/doris/pull/64930
### What problem does this PR solve?
Issue Number: None
Related PR: None
Problem Summary: sequence_match and sequence_count accepted syntactically
valid but semantically invalid event indexes such as (?0) and indexes larger
than the number of event predicates. In BE, (?0) underflowed before bitset
access and could abort the backend; the upper bound also allowed one
nonexistent event. This change validates the event number range in Nereids and
keeps the BE parser from using event_number - 1 until the range is checked.
### Release note
Fix sequence_match and sequence_count to reject out-of-range event numbers
instead of risking backend abort.
### Check List (For Author)
- Test:
- Unit Test: ./run-be-ut.sh --clean --run --filter=VSequenceMatchTest.*
- Manual test: PATH=/mnt/disk6/common/ldb_toolchain_toucan/bin:$PATH
build-support/clang-format.sh
- Manual test: PATH=/mnt/disk6/common/ldb_toolchain_toucan/bin:$PATH
build-support/check-format.sh
- Manual test: build-support/run-clang-tidy.sh --build-dir
be/ut_build_ASAN
- Manual test: ./build.sh --fe --clean
- Regression test: Not run locally because no isolated worktree Doris
cluster was started; added regression coverage for CI.
- Behavior changed: Yes. Invalid sequence_match and sequence_count event
numbers now return analysis errors or an empty BE-side match instead of
reaching invalid bitset indexes.
- Does this need documentation: No
--
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]
