Repository navigation
build(deps): [2.6] update Tantivy for phrase-match optimizations - #54029
sre-ci-robot merged 2 commits into
Conversation
Backport master PR milvus-io#54023. Advance the locked Tantivy 0.23 revision to the merge containing zilliztech/tantivy#30 and #33, preserving the legacy Tantivy and rust-stemmers pins. (cherry picked from commit 53af8ef) Signed-off-by: Buqian Zheng <zhengbuqian@gmail.com>
sre-ci-robot
added
the
do-not-merge/need-merge-master-first
|
[ci-v2-notice] To rerun ci-v2 checks, comment with:
If you have any questions or requests, please contact @zhikunyao. |
|
[INFO] PR Label Summary by Default
[WARNING] Milestone not set
You can set milestone by commenting: Use /refresh-label to update related check and label manually |
Codecov Report✅ All modified and coverable lines are covered by tests. Additional details and impacted files@@ Coverage Diff @@
## 2.6 #54029 +/- ##
==========================================
+ Coverage 76.99% 79.13% +2.14%
==========================================
Files 1700 2058 +358
Lines 262533 350600 +88067
==========================================
+ Hits 202142 277463 +75321
- Misses 53550 64827 +11277
- Partials 6841 8310 +1469
🚀 New features to boost your workflow:
|
sre-ci-robot
added
the
low-code-coverage
|
/ci-rerun-e2e-default |
|
/lgtm |
|
[INFO] PR Label Summary by Default
[WARNING] Milestone not set
You can set milestone by commenting: Use /refresh-label to update related check and label manually |
1 similar comment
|
[INFO] PR Label Summary by Default
[WARNING] Milestone not set
You can set milestone by commenting: Use /refresh-label to update related check and label manually |
Advance the nine modern Tantivy workspace pins to fef3c2a57 (zilliztech/tantivy#34). Preserve all other dependencies. Master PR: milvus-io#54023. Signed-off-by: Buqian Zheng <zhengbuqian@gmail.com>
sre-ci-robot
added
low-code-coverage
|
/lgtm |
|
[INFO] PR Label Summary by Default
Use /refresh-label to update related check and label manually |
1 similar comment
|
[INFO] PR Label Summary by Default
Use /refresh-label to update related check and label manually |
|
/approve |
|
[APPROVALNOTIFIER] This PR is APPROVED This pull-request has been approved by: sparknack The full list of commands accepted by this bot can be found here. The pull request process is described here DetailsNeeds approval from an approver in each of these files:
Approvers can indicate their approval by writing |
|
[approval-status] effective-owner-approvals=1 [sparknack(comment)]; do-not-merge/disable-approve-self=not-required; do-not-merge/doc-need-two-approve=disabled; ignored=[none] |
|
[INFO] PR Label Summary by Default
Use /refresh-label to update related check and label manually |
zhengbuqian
removed
the
do-not-merge/need-merge-master-first
4 participants
pr: #54023
Advance the locked Tantivy 0.23 revision to
fef3c2a5748bdb2b2c2f8ca1d9fc125a0d1e8b71, including the no-scoring phrase optimizations from #30, #33, and the repeated-term correctness fix from #34. For example,a awith slop 1 must not match a document containing only onea. Milvus's phrase-match collectors disable scoring and reach the corrected path.Revision and scope
mainHEAD when this update began.596a0f65ba24dd94b15ef97b9b9a909b610b8f22([BUG] Fix Jenkins CI remove local docker images bug #33), whose parent is9ce00de9a822b2a1cfa8d2b91454d7ae017969b9([BUG] Some troubleshoot messages in Milvus do not provide enough information #30). GitHub state and local Git ancestry agree.internal/core/thirdparty/tantivy/tantivy-binding/Cargo.lock(+9/-9). The upstream increment changes four phrase-query source files, with no manifest changes.Current local validation (2026-10-10, fef3c2a57)
Used the separate 2.6 worktree and target directory, with the branch's default features and release profile; no master or 3.0 build outputs were reused.
cargo +1.89 build --release --locked: passed. Generated files unchanged.index_writer_v5::analyzer::analyzer::tests::test_lindera_analyzerrequests IPADIC while the default feature set enables no dictionary, yieldinglindera tokenizer with invalid dict_kind. This reproduces the previously documented configuration limitation; legacy code and pins are unchanged.git diff --checkand parsed lockfile comparison: passed; only nine source hashes changed in this follow-up.Earlier results at 596a0f65 are historical. No full release-branch C++/Go build, live-cluster E2E or local performance benchmark was run. No test/source adjustments or #53986 changes are included. Upstream #34 reports some performance regressions; no universal speedup or measured Milvus gain is claimed. The fix covers unscored literal PhraseQuery, not scored, regex or prefix phrases.