Files
wehub-resource-sync 498b235461
Build and test / Build and test AMD64 Ubuntu 22.04 (push) Failing after 0s
Publish Builder / amazonlinux2023 (push) Failing after 1s
Build and test / UT for Go (push) Has been skipped
Publish KRTE Images / KRTE (push) Failing after 1s
Build and test / Integration Test (push) Has been skipped
Build and test / Upload Code Coverage (push) Has been skipped
Publish Builder / rockylinux9 (push) Failing after 1s
Publish Builder / ubuntu22.04 (push) Failing after 0s
Publish Builder / ubuntu24.04 (push) Failing after 0s
Publish Gpu Builder / publish-gpu-builder (push) Failing after 1s
Publish Test Images / PyTest (push) Failing after 0s
Build and test / UT for Cpp (push) Has been cancelled
chore: import upstream snapshot with attribution
2026-07-13 12:31:17 +08:00

201 lines
6.9 KiB
Go

// Licensed to the LF AI & Data foundation under one
// or more contributor license agreements. See the NOTICE file
// distributed with this work for additional information
// regarding copyright ownership. The ASF licenses this file
// to you under the Apache License, Version 2.0 (the
// "License"); you may not use this file except in compliance
// with the License. You may obtain a copy of the License at
//
// http://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the License is distributed on an "AS IS" BASIS,
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
// See the License for the specific language governing permissions and
// limitations under the License.
package tasks
import (
"context"
"fmt"
"github.com/apache/arrow/go/v17/arrow"
"google.golang.org/protobuf/proto"
"github.com/milvus-io/milvus/internal/querynodev2/segments"
"github.com/milvus-io/milvus/internal/util/function/chain"
"github.com/milvus-io/milvus/internal/util/function/chain/expr"
"github.com/milvus-io/milvus/internal/util/function/chain/types"
"github.com/milvus-io/milvus/internal/util/segcore"
"github.com/milvus-io/milvus/pkg/v3/proto/planpb"
"github.com/milvus-io/milvus/pkg/v3/util/merr"
)
const (
boostScoreColumnPrefix = "boost_score_"
functionScoreColumn = "function_score"
)
func boostScoreColumn(index int) string {
return fmt.Sprintf("%s%d", boostScoreColumnPrefix, index)
}
func extractPlanScorers(serializedPlan []byte) ([]*planpb.ScoreFunction, error) {
plan, err := extractPlanWithScorers(serializedPlan)
if err != nil || plan == nil {
return nil, err
}
return plan.GetScorers(), nil
}
func extractPlanWithScorers(serializedPlan []byte) (*planpb.PlanNode, error) {
if len(serializedPlan) == 0 {
return nil, nil
}
plan := &planpb.PlanNode{}
if err := proto.Unmarshal(serializedPlan, plan); err != nil {
return nil, err
}
return plan, nil
}
func functionModeToScoreCombineMode(mode planpb.FunctionMode) (string, error) {
switch mode {
case planpb.FunctionMode_FunctionModeMultiply:
return expr.ModeMultiply, nil
case planpb.FunctionMode_FunctionModeSum:
return expr.ModeSum, nil
default:
return "", merr.WrapErrServiceInternal(fmt.Sprintf("boost_score: unknown function mode %s", mode.String()))
}
}
func boostModeToScoreCombineMode(mode planpb.BoostMode) (string, error) {
switch mode {
case planpb.BoostMode_BoostModeMultiply:
return expr.ModeMultiply, nil
case planpb.BoostMode_BoostModeSum:
return expr.ModeSum, nil
default:
return "", merr.WrapErrServiceInternal(fmt.Sprintf("boost_score: unknown boost mode %s", mode.String()))
}
}
var boostScoreRunnerFactory = newSegmentBoostScoreRunner
type boostScoreFunc func(context.Context, segments.Segment, *segcore.SearchRequest, *planpb.ScoreFunction, *arrow.Chunked) (*arrow.Chunked, error)
func newSegmentBoostScoreRunner(scoreFunc boostScoreFunc, segment segments.Segment, searchReq *segcore.SearchRequest, scorer *planpb.ScoreFunction) expr.BoostScoreRunner {
return func(ctx context.Context, offsets *arrow.Chunked) (*arrow.Chunked, error) {
return scoreFunc(ctx, segment, searchReq, scorer, offsets)
}
}
func buildBoostScoreChain(
df *chain.DataFrame,
segment segments.Segment,
searchReq *segcore.SearchRequest,
scorers []*planpb.ScoreFunction,
scoreFunc boostScoreFunc,
functionMode string,
boostMode string,
) (*chain.FuncChain, error) {
boostChain := chain.NewFuncChainWithAllocator(defaultAllocator).
SetName("l0-rerank").
SetStage(types.StageL0Rerank)
boostScoreColumns, err := appendBoostScoreColumns(boostChain, segment, searchReq, scorers, scoreFunc)
if err != nil {
return nil, err
}
functionScoreCol, err := appendFunctionScoreColumn(boostChain, boostScoreColumns, functionMode)
if err != nil {
return nil, err
}
if err := appendFinalBoostScore(boostChain, functionScoreCol, boostMode); err != nil {
return nil, err
}
return appendL0RerankReduceContract(boostChain), nil
}
func appendBoostScoreColumns(
boostChain *chain.FuncChain,
segment segments.Segment,
searchReq *segcore.SearchRequest,
scorers []*planpb.ScoreFunction,
scoreFunc boostScoreFunc,
) ([]string, error) {
boostScoreColumns := make([]string, 0, len(scorers))
for scorerIdx, scorer := range scorers {
outputCol := boostScoreColumn(scorerIdx)
boostExpr, err := expr.NewBoostScoreExpr(boostScoreRunnerFactory(scoreFunc, segment, searchReq, scorer))
if err != nil {
return nil, err
}
boostChain.Map(boostExpr, []string{types.SegOffsetFieldName}, []string{outputCol})
boostScoreColumns = append(boostScoreColumns, outputCol)
}
return boostScoreColumns, nil
}
func appendFunctionScoreColumn(boostChain *chain.FuncChain, boostScoreColumns []string, functionMode string) (string, error) {
if len(boostScoreColumns) == 1 {
return boostScoreColumns[0], nil
}
functionCombineExpr, err := expr.NewNumCombineExpr(functionMode, nil, expr.WithNullPolicy(expr.NumCombineNullSkip))
if err != nil {
return "", err
}
boostChain.Map(functionCombineExpr, boostScoreColumns, []string{functionScoreColumn})
return functionScoreColumn, nil
}
func appendFinalBoostScore(boostChain *chain.FuncChain, functionScoreCol string, boostMode string) error {
finalCombineExpr, err := expr.NewNumCombineExpr(boostMode, nil, expr.WithNullPolicy(expr.NumCombineNullSkip))
if err != nil {
return err
}
boostChain.Map(finalCombineExpr,
[]string{types.ScoreFieldName, functionScoreCol},
[]string{types.ScoreFieldName})
return nil
}
func (t *SearchTask) applyBoostScores(segDFs []*chain.DataFrame, searchedSegments []segments.Segment, searchReq *segcore.SearchRequest) error {
plan, err := extractPlanWithScorers(t.req.GetReq().GetSerializedExprPlan())
if err != nil {
return merr.WrapErrServiceInternal(fmt.Sprintf("boost_score: failed to parse search plan scorers: %v", err))
}
return t.applyBoostScoresWithPlan(segDFs, plan, searchedSegments, searchReq)
}
func (t *SearchTask) applyBoostScoresWithPlan(segDFs []*chain.DataFrame, plan *planpb.PlanNode, searchedSegments []segments.Segment, searchReq *segcore.SearchRequest) error {
if len(segDFs) != len(searchedSegments) {
return merr.WrapErrServiceInternal(fmt.Sprintf("boost_score: DataFrame count %d does not match segment count %d", len(segDFs), len(searchedSegments)))
}
if plan == nil || len(plan.GetScorers()) == 0 {
return nil
}
scorers := plan.GetScorers()
functionMode, err := functionModeToScoreCombineMode(plan.GetScoreOption().GetFunctionMode())
if err != nil {
return err
}
boostMode, err := boostModeToScoreCombineMode(plan.GetScoreOption().GetBoostMode())
if err != nil {
return err
}
scoreFunc := segments.AsyncComputeScorerScoresOnChunkedOffsets
return executeL0RerankChains(t.ctx, segDFs, func(_ context.Context, i int, df *chain.DataFrame) (*chain.FuncChain, error) {
return buildBoostScoreChain(df, searchedSegments[i], searchReq, scorers, scoreFunc, functionMode, boostMode)
}, "boost_score")
}