feat: Phase 6.6 知识库 RAG 增强 — 文档索引 + 语义检索 + KnowledgeProvider

- rag.Embedder: LLM API 文本向量化 (OpenAI-compatible) - rag.KnowledgeStore: 文档分块 + 重叠窗口 + 余弦相似度搜索 - rag.Retriever: 高级知识检索 + 格式化摘要 - KnowledgeProvider: 子会话提供者，整合入编排管线 - knowledge_search / knowledge_ingest 工具 - EnrichmentData 管线全线支持 KnowledgeInfo Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
2026-05-23 22:33:26 +08:00
parent 9a8fb8d0ce
commit cd83eec39e
10 changed files with 752 additions and 3 deletions
@@ -0,0 +1,96 @@
+package subsession
+
+import (
+	"context"
+	"fmt"
+	"time"
+
+	"github.com/yourname/cyrene-ai/ai-core/internal/model"
+	"github.com/yourname/cyrene-ai/ai-core/internal/rag"
+	"github.com/yourname/cyrene-ai/pkg/logger"
+)
+
+// KnowledgeProvider searches the knowledge base for relevant information.
+type KnowledgeProvider struct {
+	retriever *rag.Retriever
+}
+
+// NewKnowledgeProvider creates a knowledge subsession provider.
+func NewKnowledgeProvider(retriever *rag.Retriever) *KnowledgeProvider {
+	return &KnowledgeProvider{retriever: retriever}
+}
+
+func (p *KnowledgeProvider) Type() model.SubSessionType {
+	return model.SubSessionKnowledge
+}
+
+func (p *KnowledgeProvider) CanHandle(_ context.Context, intent *model.IntentResult, _ string) bool {
+	if intent == nil {
+		return true
+	}
+	// Activate for technical questions, how-to queries, and factual questions
+	switch intent.Primary {
+	case "knowledge", "technical", "how_to", "factual", "research":
+		return true
+	case "chat":
+		// For general chat, only search if there might be relevant info
+		return false
+	}
+	return true
+}
+
+func (p *KnowledgeProvider) Priority() int {
+	return 3
+}
+
+func (p *KnowledgeProvider) Timeout() time.Duration {
+	return 15 * time.Second
+}
+
+func (p *KnowledgeProvider) CreateContext(ctx context.Context, params CreateContextParams) ([]model.LLMMessage, error) {
+	return []model.LLMMessage{
+		{Role: model.RoleSystem, Content: "知识库检索子会话"},
+		{Role: model.RoleUser, Content: params.UserMessage},
+	}, nil
+}
+
+func (p *KnowledgeProvider) Execute(ctx context.Context, subCtx []model.LLMMessage) (*model.SubSessionResult, error) {
+	userMessage := ""
+	for i := len(subCtx) - 1; i >= 0; i-- {
+		if subCtx[i].Role == model.RoleUser {
+			userMessage = subCtx[i].Content
+			break
+		}
+	}
+	if userMessage == "" {
+		return nil, fmt.Errorf("无法提取用户消息")
+	}
+
+	result := &model.SubSessionResult{
+		Type:       model.SubSessionKnowledge,
+		Confidence: 0,
+	}
+
+	if p.retriever == nil {
+		result.Summary = "（知识库未就绪）"
+		return result, nil
+	}
+
+	retrieval, err := p.retriever.Retrieve(ctx, userMessage, 3)
+	if err != nil {
+		logger.Printf("[knowledge-subsession] 知识检索失败: %v", err)
+		result.Error = fmt.Sprintf("检索失败: %v", err)
+		result.Summary = "（知识库检索失败）"
+		return result, nil
+	}
+
+	if len(retrieval.Results) == 0 {
+		result.Summary = "（未找到相关知识）"
+		return result, nil
+	}
+
+	result.Summary = retrieval.Summary
+	result.Confidence = 0.6
+	logger.Printf("[knowledge-subsession] 完成: 找到 %d 条知识", len(retrieval.Results))
+	return result, nil
+}