Add files via upload

This commit is contained in:
公明
2026-07-08 15:20:59 +08:00
committed by GitHub
parent 784be8a162
commit 62fafbf09b
5 changed files with 192 additions and 34 deletions
+43 -7
View File
@@ -4,6 +4,7 @@ import (
"context"
"fmt"
"net/http"
"strings"
"time"
"cyberstrike-ai/internal/audit"
@@ -316,20 +317,55 @@ func (h *KnowledgeHandler) DeleteItem(c *gin.Context) {
c.JSON(http.StatusOK, gin.H{"message": "删除成功"})
}
// RebuildIndex 重建索引
func (h *KnowledgeHandler) RebuildIndex(c *gin.Context) {
// 异步重建索引
// StartIndex 构建知识库向量索引。默认仅补齐尚无向量的知识项;mode=full 时全量重建。
func (h *KnowledgeHandler) StartIndex(c *gin.Context) {
if err := h.indexer.TryBeginIndexRun(); err != nil {
c.JSON(http.StatusConflict, gin.H{"error": "已有索引任务正在进行,请等待完成"})
return
}
mode := strings.TrimSpace(c.Query("mode"))
if mode == "" {
mode = "missing"
}
if mode != "full" && mode != "missing" {
h.indexer.FinishIndexRun()
c.JSON(http.StatusBadRequest, gin.H{"error": "无效的 mode 参数,可选值:missing、full"})
return
}
fullRebuild := mode == "full"
message := "索引构建已开始,将在后台进行"
auditAction := "index_build"
auditDetail := "构建知识库索引"
if fullRebuild {
message = "全量索引重建已开始,将在后台进行"
auditAction = "index_rebuild_full"
auditDetail = "全量重建知识库索引"
}
go func() {
defer h.indexer.FinishIndexRun()
ctx := context.Background()
if err := h.indexer.RebuildIndex(ctx); err != nil {
h.logger.Error("重建索引失败", zap.Error(err))
var err error
if fullRebuild {
err = h.indexer.RunRebuildIndex(ctx)
} else {
err = h.indexer.RunIndexMissing(ctx)
}
if err != nil {
if fullRebuild {
h.logger.Error("全量重建索引失败", zap.Error(err))
} else {
h.logger.Error("构建知识库索引失败", zap.Error(err))
}
}
}()
if h.audit != nil {
h.audit.RecordOK(c, "knowledge", "index_rebuild", "重建知识库索引", "knowledge", "", nil)
h.audit.RecordOK(c, "knowledge", auditAction, auditDetail, "knowledge", "", nil)
}
c.JSON(http.StatusOK, gin.H{"message": "索引重建已开始,将在后台进行"})
c.JSON(http.StatusOK, gin.H{"message": message, "mode": mode})
}
// ScanKnowledgeBase 扫描知识库
+23 -4
View File
@@ -4416,16 +4416,35 @@ func (h *OpenAPIHandler) GetOpenAPISpec(c *gin.Context) {
"/api/knowledge/index": map[string]interface{}{
"post": map[string]interface{}{
"tags": []string{"知识库"},
"summary": "建索引",
"description": "重新构建知识库索引",
"operationId": "rebuildIndex",
"summary": "建索引",
"description": "构建知识库向量索引。默认仅处理尚无向量的知识项;mode=full 时全量重建。",
"operationId": "startKnowledgeIndex",
"parameters": []map[string]interface{}{
{
"name": "mode",
"in": "query",
"required": false,
"description": "索引模式:missing(默认,补齐缺失向量)或 full(全量重建)",
"schema": map[string]interface{}{
"type": "string",
"enum": []string{"missing", "full"},
"default": "missing",
},
},
},
"responses": map[string]interface{}{
"200": map[string]interface{}{
"description": "重建索引任务已启动",
"description": "索引任务已启动",
},
"400": map[string]interface{}{
"description": "无效的 mode 参数",
},
"401": map[string]interface{}{
"description": "未授权",
},
"409": map[string]interface{}{
"description": "已有索引任务正在进行",
},
},
},
},
+1 -1
View File
@@ -43,7 +43,7 @@ var apiDocI18nSummaryToKey = map[string]string{
"设置对话置顶": "pinConversation", "设置分组置顶": "pinGroup", "设置分组中对话的置顶": "pinGroupConversation",
"获取分类": "getCategories", "列出知识项": "listKnowledgeItems", "创建知识项": "createKnowledgeItem",
"获取知识项": "getKnowledgeItem", "更新知识项": "updateKnowledgeItem", "删除知识项": "deleteKnowledgeItem",
"获取索引状态": "getIndexStatus", "建索引": "rebuildIndex", "扫描知识库": "scanKnowledgeBase",
"获取索引状态": "getIndexStatus", "建索引": "startKnowledgeIndex", "扫描知识库": "scanKnowledgeBase",
"搜索知识库": "searchKnowledgeBase", "基础搜索": "basicSearch", "按风险类型搜索": "searchByRiskType",
"获取检索日志": "getRetrievalLogs", "删除检索日志": "deleteRetrievalLog",
"MCP端点": "mcpEndpoint", "列出所有工具": "listAllTools", "调用工具": "invokeTool", "初始化连接": "initConnection",
+105 -22
View File
@@ -213,9 +213,13 @@ func (idx *Indexer) HasIndex() (bool, error) {
return count > 0, nil
}
// RebuildIndex 重建所有索引
func (idx *Indexer) RebuildIndex(ctx context.Context) error {
func (idx *Indexer) beginIndexRun() error {
idx.rebuildMu.Lock()
defer idx.rebuildMu.Unlock()
if idx.isRebuilding {
return fmt.Errorf("索引任务已在进行中")
}
idx.isRebuilding = true
idx.rebuildTotalItems = 0
idx.rebuildCurrent = 0
@@ -223,41 +227,124 @@ func (idx *Indexer) RebuildIndex(ctx context.Context) error {
idx.rebuildStartTime = time.Now()
idx.rebuildLastItemID = ""
idx.rebuildLastChunks = 0
idx.rebuildMu.Unlock()
return nil
}
// TryBeginIndexRun 同步占用索引任务槽位;调用方必须在后台任务结束时调用 FinishIndexRun。
func (idx *Indexer) TryBeginIndexRun() error {
return idx.beginIndexRun()
}
func (idx *Indexer) FinishIndexRun() {
idx.rebuildMu.Lock()
idx.isRebuilding = false
idx.rebuildMu.Unlock()
}
func (idx *Indexer) resetLastError() {
idx.mu.Lock()
idx.lastError = ""
idx.lastErrorTime = time.Time{}
idx.errorCount = 0
idx.mu.Unlock()
}
rows, err := idx.db.Query("SELECT id FROM knowledge_base_items")
func (idx *Indexer) setIndexRunTotal(total int) {
idx.rebuildMu.Lock()
idx.rebuildTotalItems = total
idx.rebuildMu.Unlock()
}
// IndexMissing 为尚无向量的知识项构建索引(默认推荐路径,适合冷启动与中断续跑)。
func (idx *Indexer) IndexMissing(ctx context.Context) error {
if err := idx.beginIndexRun(); err != nil {
return err
}
defer idx.FinishIndexRun()
return idx.runIndexMissing(ctx)
}
// RebuildIndex 全量重建所有知识项索引(显式 opt-in,成本更高)。
func (idx *Indexer) RebuildIndex(ctx context.Context) error {
if err := idx.beginIndexRun(); err != nil {
return err
}
defer idx.FinishIndexRun()
return idx.runRebuildIndex(ctx)
}
// RunRebuildIndex 在已占用索引任务槽位后执行全量重建(供 HTTP handler 后台任务使用)。
func (idx *Indexer) RunRebuildIndex(ctx context.Context) error {
return idx.runRebuildIndex(ctx)
}
// RunIndexMissing 在已占用索引任务槽位后执行缺失索引补齐(供 HTTP handler 后台任务使用)。
func (idx *Indexer) RunIndexMissing(ctx context.Context) error {
return idx.runIndexMissing(ctx)
}
func (idx *Indexer) runRebuildIndex(ctx context.Context) error {
idx.resetLastError()
rows, err := idx.db.QueryContext(ctx, "SELECT id FROM knowledge_base_items ORDER BY updated_at ASC, id ASC")
if err != nil {
idx.rebuildMu.Lock()
idx.isRebuilding = false
idx.rebuildMu.Unlock()
return fmt.Errorf("查询知识项失败:%w", err)
}
defer rows.Close()
itemIDs, err := scanKnowledgeItemIDs(rows)
if err != nil {
return err
}
idx.setIndexRunTotal(len(itemIDs))
idx.logger.Info("开始重建索引", zap.Int("totalItems", len(itemIDs)))
return idx.indexItemIDs(ctx, itemIDs, "索引重建完成")
}
func (idx *Indexer) runIndexMissing(ctx context.Context) error {
idx.resetLastError()
rows, err := idx.db.QueryContext(ctx, `
SELECT i.id
FROM knowledge_base_items i
LEFT JOIN knowledge_embeddings e ON e.item_id = i.id
WHERE e.item_id IS NULL
ORDER BY i.updated_at ASC, i.id ASC
`)
if err != nil {
return fmt.Errorf("查询未索引知识项失败:%w", err)
}
defer rows.Close()
itemIDs, err := scanKnowledgeItemIDs(rows)
if err != nil {
return fmt.Errorf("扫描未索引知识项 ID 失败:%w", err)
}
idx.setIndexRunTotal(len(itemIDs))
idx.logger.Info("开始补齐缺失索引", zap.Int("totalItems", len(itemIDs)))
return idx.indexItemIDs(ctx, itemIDs, "索引构建完成")
}
func scanKnowledgeItemIDs(rows *sql.Rows) ([]string, error) {
var itemIDs []string
for rows.Next() {
var id string
if err := rows.Scan(&id); err != nil {
idx.rebuildMu.Lock()
idx.isRebuilding = false
idx.rebuildMu.Unlock()
return fmt.Errorf("扫描知识项 ID 失败:%w", err)
return nil, fmt.Errorf("扫描知识项 ID 失败:%w", err)
}
itemIDs = append(itemIDs, id)
}
if err := rows.Err(); err != nil {
return nil, fmt.Errorf("扫描知识项 ID 失败:%w", err)
}
return itemIDs, nil
}
idx.rebuildMu.Lock()
idx.rebuildTotalItems = len(itemIDs)
idx.rebuildMu.Unlock()
idx.logger.Info("开始重建索引", zap.Int("totalItems", len(itemIDs)))
func (idx *Indexer) indexItemIDs(ctx context.Context, itemIDs []string, doneMessage string) error {
failedCount := 0
consecutiveFailures := 0
maxConsecutiveFailures := 5
@@ -329,11 +416,7 @@ func (idx *Indexer) RebuildIndex(ctx context.Context) error {
}
}
idx.rebuildMu.Lock()
idx.isRebuilding = false
idx.rebuildMu.Unlock()
idx.logger.Info("索引重建完成", zap.Int("totalItems", len(itemIDs)), zap.Int("failedCount", failedCount))
idx.logger.Info(doneMessage, zap.Int("totalItems", len(itemIDs)), zap.Int("failedCount", failedCount))
return nil
}
@@ -0,0 +1,20 @@
package knowledge
import "testing"
func TestIndexerRejectsConcurrentIndexRuns(t *testing.T) {
idx := &Indexer{}
if err := idx.beginIndexRun(); err != nil {
t.Fatalf("first index run should start: %v", err)
}
if err := idx.beginIndexRun(); err == nil {
t.Fatal("second index run should be rejected while one is active")
}
idx.FinishIndexRun()
if err := idx.beginIndexRun(); err != nil {
t.Fatalf("index run should start again after finish: %v", err)
}
idx.FinishIndexRun()
}