test no space

[LeftHandCold]

User description

What type of PR is this?

• API-change
• BUG
• Improvement
• Documentation
• Feature
• Test and CI
• Code Refactoring

Which issue(s) this PR fixes:

issue #22634

What this PR does / why we need it:

test not space

---

PR Type

Bug fix, Enhancement

---

Description

• Revert ORDER BY LIMIT pushdown optimization on IVFFlat index

• Refactor IVFFlat search to use streaming SQL with multi-threaded processing

• Rename plan node parameters from nodes to ns for consistency

• Remove BlockOrderBy and BlockLimit fields from query plan

• Update dependencies and clean up unused imports

---

Diagram Walkthrough

flowchart LR
  A["IVFFlat Search"] -->|"Streaming SQL"| B["Multi-threaded Processing"]
  B -->|"Safe Heap"| C["Result Aggregation"]
  D["Plan Node"] -->|"Remove BlockOrderBy/BlockLimit"| E["Simplified Plan"]
  F["Dependencies"] -->|"Update versions"| G["Aligned packages"]

Loading

File Walkthrough

	Relevant files
Refactoring	2 files compile.go Rename nodes parameter to ns throughout function                  +98/-100 sql_executor.go Simplify plan building by removing statement type switch  +1/-23
Enhancement	3 files search.go Implement streaming SQL with multi-threaded search              +197/-48 ivf_search.go Remove background queries tracking from IVF search              +5/-13    types.go Remove BackgroundQueries field from RuntimeConfig                +2/-4
Tests	4 files search_test.go Add streaming mock functions and cancel test case                +57/-2    ivf_search_test.go Add generic type parameters to vector operations                  +7/-6      pushdown.result Update expected output removing table prefix from sort keys +3/-3      top.result Update expected output removing table prefix from sort keys +3/-3
Bug fix	6 files constant_fold.go Remove BlockLimit and BlockOrderBy constant folding            +1/-17    explain_node.go Remove BlockOrderBy and BlockLimit explain logic                  +1/-37    local_disttae_datasource.go Remove Limit field and condition check from OrderBy handling +3/-3      db_test.go Remove extra nil parameter from BlockDataReadInner calls  +10/-10  stats.go Remove NodeType condition from limit calculation                  +1/-1      plan.proto Remove BlockOrderBy and BlockLimit message fields                +4/-7
Formatting	5 files read.go Remove unused imports and reorganize dependencies                +4/-87    backup.go Reorganize imports and add zap logging import                        +8/-7      reader.go Add zap logging import                                                                      +2/-50    datasource.go Reorganize imports for consistency                                              +3/-3      types.go Add blank line in DataSource interface                                      +1/-1
Dependencies	3 files Makefile Downgrade USearch and related library versions                      +4/-4      go.mod Update Go dependencies and reorganize require blocks          +67/-64  go.sum Update checksums for Go dependencies                                          +22/-22
Configuration changes	1 files .golangci.yml Enable SA1019 staticcheck rule                                                      +1/-1
Additional files	13 files engine_mock.go +0/-12    types.go +0/-10    plan.pb.go +985/-1113 scope.go +0/-4      types.go +0/-3      deepcopy.go +0/-6      pushdown.go +0/-53    query_builder.go +0/-29    index.go +0/-26    types.go +0/-18    table_meta_reader.go +0/-3      txn_table_delegate.go +0/-3      table_reader.go +0/-3

---

[qodo-merge-pro]

PR Compliance Guide 🔍

Below is a summary of compliance checks for this PR:

Security Compliance
⚪	Resource leak risk Description: Streaming SQL consumer may leak memory if producer errors are not fully drained or channels are mishandled; ensure error and result channels are always drained/closed and context cancellations are correctly propagated to avoid unbounded buffering or goroutine leaks. search.go [214-341] Referred Code nthread int64, ) (keys any, distances []float64, err error) { stream_chan := make(chan executor.Result, nthread) error_chan := make(chan error, nthread) distfn, err := metric.ResolveDistanceFn[T](metric.MetricType(idxcfg.Ivfflat.Metric)) if err != nil { return nil, nil, err } centroids_ids, err := idx.findCentroids(proc, query, distfn, idxcfg, rt.Probe, nthread) if err != nil { return nil, nil, err } var instr string for i, c := range centroids_ids { if i > 0 { instr += "," } ... (clipped 107 lines)
	DoS via drain Description: Results from stream channel are drained and closed without a bounded timeout after cancellation, which could allow a malicious or buggy producer to block shutdown leading to denial of service. search.go [318-323] Referred Code if !gStreamClosed.Load() { for res := range stream_chan { res.Close() } }
Ticket Compliance
🟡	🎫 #22634 🔴 Ensure that created database snapshots are not garbage collected (GC) while referenced by a snapshot. Ensure that created table snapshots are not garbage collected (GC) while referenced by a snapshot. Provide tests or CI coverage that validates snapshot retention behavior.
Codebase Duplication Compliance
⚪	Codebase context is not defined Follow the guide to enable codebase context checks.
Custom Compliance
⚪	No custom compliance provided Follow the guide to enable custom compliance check.
Update

Compliance status legend

🟢 - Fully Compliant
🟡 - Partial Compliant
🔴 - Not Compliant
⚪ - Requires Further Human Verification
🏷️ - Compliance label

[qodo-merge-pro]

PR Code Suggestions ✨

Explore these optional code suggestions:

Category	Suggestion	Impact
Possible issue	Avoid reusing protobuf field tags Revert the protobuf field tag for SourceStep from 32 back to 27. Reusing a tag for a field of a different type breaks backward compatibility and can cause data corruption. pkg/pb/plan/plan.pb.go [6351-6358] TblFuncExprList []*Expr `protobuf:"bytes,26,rep,name=tbl_func_expr_list,json=tblFuncExprList,proto3" json:"tbl_func_expr_list,omitempty"` -+ ClusterTable *ClusterTable `protobuf:"bytes,28,opt,name=cluster_table,json=clusterTable,proto3" json:"cluster_table,omitempty"` -+ NotCacheable bool `protobuf:"varint,29,opt,name=not_cacheable,json=notCacheable,proto3" json:"not_cacheable,omitempty"` -+ InsertCtx *InsertCtx `protobuf:"bytes,30,opt,name=insert_ctx,json=insertCtx,proto3" json:"insert_ctx,omitempty"` -+ ReplaceCtx *ReplaceCtx `protobuf:"bytes,31,opt,name=replace_ctx,json=replaceCtx,proto3" json:"replace_ctx,omitempty"` // used to connect two plans[steps] -- SourceStep []int32 `protobuf:"varint,27,rep,packed,name=source_step,json=sourceStep,proto3" json:"source_step,omitempty"` -- ClusterTable *ClusterTable `protobuf:"bytes,28,opt,name=cluster_table,json=clusterTable,proto3" json:"cluster_table,omitempty"` -- NotCacheable bool `protobuf:"varint,29,opt,name=not_cacheable,json=notCacheable,proto3" json:"not_cacheable,omitempty"` -- InsertCtx *InsertCtx `protobuf:"bytes,30,opt,name=insert_ctx,json=insertCtx,proto3" json:"insert_ctx,omitempty"` -- ReplaceCtx *ReplaceCtx `protobuf:"bytes,31,opt,name=replace_ctx,json=replaceCtx,proto3" json:"replace_ctx,omitempty"` -- BlockOrderBy []*OrderBySpec `protobuf:"bytes,32,rep,name=block_order_by,json=blockOrderBy,proto3" json:"block_order_by,omitempty"` -- BlockLimit *Expr `protobuf:"bytes,33,opt,name=block_limit,json=blockLimit,proto3" json:"block_limit,omitempty"` -- PreInsertCtx *PreInsertCtx `protobuf:"bytes,34,opt,name=pre_insert_ctx,json=preInsertCtx,proto3" json:"pre_insert_ctx,omitempty"` -+ SourceStep []int32 `protobuf:"varint,32,rep,packed,name=source_step,json=sourceStep,proto3" json:"source_step,omitempty"` -+ PreInsertCtx *PreInsertCtx `protobuf:"bytes,34,opt,name=pre_insert_ctx,json=preInsertCtx,proto3" json:"pre_insert_ctx,omitempty"` + SourceStep []int32 `protobuf:"varint,27,rep,packed,name=source_step,json=sourceStep,proto3" json:"source_step,omitempty"` + ClusterTable *ClusterTable `protobuf:"bytes,28,opt,name=cluster_table,json=clusterTable,proto3" json:"cluster_table,omitempty"` + NotCacheable bool `protobuf:"varint,29,opt,name=not_cacheable,json=notCacheable,proto3" json:"not_cacheable,omitempty"` + InsertCtx *InsertCtx `protobuf:"bytes,30,opt,name=insert_ctx,json=insertCtx,proto3" json:"insert_ctx,omitempty"` + ReplaceCtx *ReplaceCtx `protobuf:"bytes,31,opt,name=replace_ctx,json=replaceCtx,proto3" json:"replace_ctx,omitempty"` + PreInsertCtx *PreInsertCtx `protobuf:"bytes,34,opt,name=pre_insert_ctx,json=preInsertCtx,proto3" json:"pre_insert_ctx,omitempty"` Apply / Chat Suggestion importance[1-10]: 10 __ Why: The suggestion correctly identifies a critical backward-compatibility issue where a protobuf field tag (32) is reused for a field (SourceStep) with a different type, which can lead to data corruption or panics.	High
	Drain error channel to prevent deadlocks Drain the error_chan completely to handle all reported errors and prevent potential deadlocks from blocked goroutines. pkg/vectorindex/ivfflat/search.go [326-330] // fetch potential remaining errors from error_chan // if error is not nil, fast return - select { - case err = <-error_chan: + for len(error_chan) > 0 { + if err == nil { + err = <-error_chan + } else { + <-error_chan + } + } + if err != nil { return - default: } [To ensure code accuracy, apply this suggestion manually] Suggestion importance[1-10]: 7 __ Why: The suggestion correctly identifies that multiple errors can be sent to error_chan, but only one is read, which could hide other errors and potentially cause goroutines to block if the channel fills up.	Medium
	Add nil check for search vector Add a check in runIvfSearchVector to ensure the retrieved vector fa is not empty before calling the search function. pkg/sql/colexec/table_function/ivf_search.go [231-252] func runIvfSearchVector[T types.RealNumbers](u *ivfSearchState, proc *process.Process, faVec *vector.Vector, nthRow int) (err error) { if faVec.IsNull(uint64(nthRow)) { return nil } fa := vector.GetArrayAt[T](faVec, nthRow) + if len(fa) == 0 { + return nil + } algo, err := u.newIvfAlgoFn(u.idxcfg, u.tblcfg) if err != nil { return err } key := fmt.Sprintf("%s:%d", u.tblcfg.IndexTable, u.idxcfg.Ivfflat.Version) u.keys, u.distances, err = veccache.Cache.Search(proc, key, algo, fa, vectorindex.RuntimeConfig{Limit: uint(u.limit), Probe: uint(u.tblcfg.Nprobe)}) if err != nil { return err } return nil } Apply / Chat Suggestion importance[1-10]: 5 __ Why: The suggestion correctly points out a missing validation for an empty vector slice, which could lead to wasted work or unexpected behavior in the downstream Search function.	Low
High-level	PR content mismatches its description The PR's code changes, which involve refactoring vector index search and the query plan, are inconsistent with its linked ticket that addresses a bug in database snapshots. This mismatch obscures the PR's purpose. Examples: Solution Walkthrough: Before: // PR Description PR for: "Issue #22634: Snapshots do not support databases or tables" // Code Changes - Refactor IVFFlat search to use streaming and multi-threading. - Remove `BlockOrderBy` and `BlockLimit` from the query plan. - Update various dependencies. - ... and other changes unrelated to snapshots. After: // PR Description PR for: "Issue #XXXXX: Revert ORDER BY LIMIT pushdown for IVFFlat and refactor search" // Code Changes - Refactor IVFFlat search to use streaming and multi-threading. - Remove `BlockOrderBy` and `BlockLimit` from the query plan. - Update various dependencies. - ... (code changes remain the same, but now match the description). Suggestion importance[1-10]: 9 __ Why: The suggestion correctly identifies a critical discrepancy between the linked ticket (about snapshots) and the actual code changes (IVFFlat search refactor), which severely hinders understanding the PR's intent and traceability.	High
Update