fix(webui): 知识库不再把二进制当文本存 + 接上分类/媒体/删除 + 实时计数

后端（handler_memory.go 重写 handleKnowledge）： - ★ multipart 分支原来无条件 file.Read → string(buf[:n]) → 当 Markdown 存进 content.md：传一张 PNG 得到的是一份乱码文本知识，还在 .index.json 占一份 preview，且**没有任何迹象**表明出了问题。现在按 http.DetectContentType 探测的**真实类型**分流（不信客户端声明的 Content-Type——谎报 text/plain 的 PNG 在测试里是真实场景）：媒体入 CAS 按 digest 挂条目 / 文本校验 UTF-8 后存正文 / 都不是则 400 明确拒绝并回传 rejected 清单。 - 状态码语义修正：此前 POST/GET 一律 500、DELETE 一律 404，把「名称非法」这类调用方能自己纠正的错报成服务器故障。现按 errors.Is 分流 400/404/503。 - 搜索支持 category 与 limit；返回 knowledgeView（不泄露服务端绝对路径，不回传几百 KB 的 Dense 浮点数组）。 - 列表端点补 dense 状态（前端据此提示多模态是否就绪）。 - 媒体存储未接线时上传图片返回 503，而非退化成把二进制当文本存。前端（dashboard.js）： - 搜索结果从 <pre>{JSON}</pre> 改为结构化渲染（名称/体积/预览/媒体标记 + 每条删除按钮）。此前前端根本没有删除入口。 - 新增分类输入框（走 category 参数）、多文件上传。 - 计数改实时：原先读 state.kernel 快照，知识条目经工具/上传增删后不会变（实测创建完仍显示 "-"）。切到知识面板时拉 /knowledge 的真实 names.length。 - 顶部输入框变多文件；显示多模态就绪状态（ready/total）。
2026-09-29 22:12:52 +00:00 · 2026-09-26 11:37:13 +08:00
parent 9e282ad774
commit d3a796fde2
4 changed files with 952 additions and 79 deletions
--- a/internal/plugins/webui/handler_memory.go
+++ b/internal/plugins/webui/handler_memory.go
@ -1,12 +1,19 @@
 package webui

 import (
+	"encoding/json"
+	"errors"
+	"fmt"
+	"io"
+	"net/http"
+	"path/filepath"
 	"strconv"
 	"strings"
+	"unicode/utf8"

-	"encoding/json"
+	"gitcode.com/JianFeeeee/HomeAgent/internal/knowledge"
 	sdk "gitcode.com/JianFeeeee/HomeAgent/internal/sdk"
-	"net/http"
+	pubsdk "gitcode.com/JianFeeeee/homeagent-sdk/sdk"
 )

 // 记忆面：图记忆 / 文档记忆 / 文本记忆 / 知识库 / LLM 源 / 变更追踪。
@ -119,79 +126,156 @@ func (h *Handler) handleMemoryGraph(w http.ResponseWriter, r *http.Request) {
 	writeJSON(w, http.StatusOK, map[string]interface{}{"success": true, "data": data})
 }

+// knowledgeWriteReq 是知识写入请求体（JSON 分支）。
+type knowledgeWriteReq struct {
+	Name    string                  `json:"name"`
+	Content string                  `json:"content"`
+	Media   []sdk.KnowledgeMediaRef `json:"media,omitempty"`
+}
+
+// knowledgeView 是返回给前端的知识条目视图。
+//
+// 为何不让前端直接吃 *knowledge.Knowledge：那个结构里有 Path（服务端绝对
+// 路径，不该外泄）、Dense（几百 KB 浮点数组）。前端只需要 name/category/
+// tags/size/updated_at/media 摘要。
+type knowledgeView struct {
+	Name      string               `json:"name"`
+	Category  string               `json:"category,omitempty"`
+	Preview   string               `json:"preview"`
+	Size      int                  `json:"size"`
+	Tags      []string             `json:"tags,omitempty"`
+	UpdatedAt string               `json:"updated_at,omitempty"`
+	Media     []knowledgeMediaView `json:"media,omitempty"`
+}
+
+type knowledgeMediaView struct {
+	Digest string `json:"digest"`
+	MIME   string `json:"mime"`
+	Kind   string `json:"kind,omitempty"`
+}
+
+// knowledgeStatusFor 把内核的 ErrInvalidName / ErrNotFound 映射到正确状态码。
+//
+// 此前一律 500：把「名称非法」「不存在」这种**调用方能自己纠正**的错报成
+// 服务器故障，前端无从区分该改请求还是该报服务器挂了。
+func knowledgeStatusFor(err error) (int, string) {
+	switch {
+	case errors.Is(err, knowledge.ErrInvalidName):
+		return http.StatusBadRequest, err.Error()
+	case errors.Is(err, knowledge.ErrNotFound):
+		return http.StatusNotFound, err.Error()
+	case errors.Is(err, sdk.ErrMediaUnavailable):
+		return http.StatusServiceUnavailable, err.Error()
+	default:
+		return http.StatusInternalServerError, err.Error()
+	}
+}
+
+func toKnowledgeView(k *pubsdk.Knowledge, name string) knowledgeView {
+	v := knowledgeView{Name: name}
+	if k.Content == "" {
+		return v
+	}
+	// 预览按 rune 截断，避免把多字节字符切成乱码
+	r := []rune(k.Content)
+	if len(r) > 200 {
+		v.Preview = string(r[:200]) + "..."
+	} else {
+		v.Preview = k.Content
+	}
+	v.Size = len(k.Content)
+	return v
+}
+
 func (h *Handler) handleKnowledge(w http.ResponseWriter, r *http.Request) {
+	if h.knowledge == nil {
+		writeJSON(w, http.StatusServiceUnavailable, map[string]string{"error": "knowledge not available"})
+		return
+	}
 	switch r.Method {
 	case http.MethodGet:
-		if h.knowledge == nil {
-			writeJSON(w, http.StatusServiceUnavailable, map[string]string{"error": "knowledge not available"})
-			return
-		}
-		query := r.URL.Query().Get("q")
-		if query != "" {
-			results, err := h.knowledge.Search(query, 10)
+		q := r.URL.Query()
+		// category 为空 = 全库；非空则限定在该分类子树内（前缀匹配）
+		category := q.Get("category")
+		if query := strings.TrimSpace(q.Get("q")); query != "" {
+			limit := 10
+			if v, err := strconv.Atoi(q.Get("limit")); err == nil && v > 0 && v <= 100 {
+				limit = v
+			}
+			var results []*pubsdk.Knowledge
+			var err error
+			if scoped, ok := h.knowledge.(interface {
+				SearchIn(query, category string, topK int) ([]*pubsdk.Knowledge, error)
+			}); ok {
+				results, err = scoped.SearchIn(query, category, limit)
+			} else {
+				results, err = h.knowledge.Search(query, limit)
+			}
 			if err != nil {
-				writeJSON(w, http.StatusInternalServerError, map[string]string{"error": err.Error()})
+				code, msg := knowledgeStatusFor(err)
+				writeJSON(w, code, map[string]string{"error": msg})
 				return
 			}
-			writeJSON(w, http.StatusOK, map[string]interface{}{"results": results})
+			views := make([]knowledgeView, 0, len(results))
+			for _, k := range results {
+				views = append(views, toKnowledgeView(k, k.Name))
+			}
+			writeJSON(w, http.StatusOK, map[string]interface{}{"results": views, "category": category})
 			return
 		}
-		categories, err := h.knowledge.List()
+		names, err := h.knowledge.List()
 		if err != nil {
-			writeJSON(w, http.StatusInternalServerError, map[string]string{"error": err.Error()})
+			code, msg := knowledgeStatusFor(err)
+			writeJSON(w, code, map[string]string{"error": msg})
 			return
 		}
-		writeJSON(w, http.StatusOK, map[string]interface{}{
-			"categories": categories,
-			"stats":      h.knowledge.Stats(),
-		})
+		if names == nil {
+			names = []string{}
+		}
+		stats := h.knowledge.Stats()
+		payload := map[string]interface{}{"names": names, "stats": stats}
+		if ds, ok := h.knowledge.(interface {
+			DenseStats() map[string]interface{}
+		}); ok {
+			payload["dense"] = ds.DenseStats()
+		}
+		writeJSON(w, http.StatusOK, payload)

 	case http.MethodPost:
-		if h.knowledge == nil {
-			writeJSON(w, http.StatusServiceUnavailable, map[string]string{"error": "knowledge not available"})
-			return
-		}
+		// multipart 分支：**先看它到底是什么**。
+		//
+		// 旧实现无论传什么都把字节 utf-8 强转后当 Markdown 存进 content.md：
+		// 上传一张 PNG 得到的是一份乱码文本知识，还会在 .index.json 里占一份
+		// preview，且没有任何迹象表明出了问题。现在改为：
+		//   - 文本类（text/* 或 JSON 字节）→ 走原路径存正文
+		//   - 媒体类（image/audio/video）→ 入 media CAS，按 digest 挂到条目上
 		ct := r.Header.Get("Content-Type")
 		if strings.HasPrefix(ct, "multipart/form-data") {
-			if err := r.ParseMultipartForm(10 << 20); err != nil {
-				writeJSON(w, http.StatusBadRequest, map[string]string{"error": err.Error()})
-				return
-			}
-			name := r.FormValue("name")
-			file, _, err := r.FormFile("file")
-			if err != nil {
-				writeJSON(w, http.StatusBadRequest, map[string]string{"error": "file required"})
-				return
-			}
-			defer file.Close()
-			buf := make([]byte, 10<<20)
-			n, _ := file.Read(buf)
-			content := string(buf[:n])
-			if err := h.knowledge.Add(name, content); err != nil {
-				writeJSON(w, http.StatusInternalServerError, map[string]string{"error": err.Error()})
-				return
-			}
-			writeJSON(w, http.StatusCreated, map[string]string{"status": "created", "name": name})
+			h.handleKnowledgeUpload(w, r)
 			return
 		}

-		var req struct {
-			Name    string `json:"name"`
-			Content string `json:"content"`
-		}
+		var req knowledgeWriteReq
 		if err := json.NewDecoder(r.Body).Decode(&req); err != nil {
 			writeJSON(w, http.StatusBadRequest, map[string]string{"error": "invalid request"})
 			return
 		}
-		if req.Name == "" || req.Content == "" {
-			writeJSON(w, http.StatusBadRequest, map[string]string{"error": "name and content required"})
+		if strings.TrimSpace(req.Name) == "" {
+			writeJSON(w, http.StatusBadRequest, map[string]string{"error": "name is required"})
 			return
 		}
-		if err := h.knowledge.Add(req.Name, req.Content); err != nil {
-			writeJSON(w, http.StatusInternalServerError, map[string]string{"error": err.Error()})
+		if strings.TrimSpace(req.Content) == "" && len(req.Media) == 0 {
+			writeJSON(w, http.StatusBadRequest, map[string]string{"error": "content or media is required"})
 			return
 		}
-		writeJSON(w, http.StatusCreated, map[string]string{"status": "created", "name": req.Name})
+		if err := h.knowledge.AddWithMedia(req.Name, req.Content, req.Media); err != nil {
+			code, msg := knowledgeStatusFor(err)
+			writeJSON(w, code, map[string]string{"error": msg})
+			return
+		}
+		writeJSON(w, http.StatusCreated, map[string]interface{}{
+			"status": "created", "name": req.Name, "media": len(req.Media),
+		})

 	case http.MethodDelete:
 		name := r.URL.Query().Get("name")
@ -200,16 +284,160 @@ func (h *Handler) handleKnowledge(w http.ResponseWriter, r *http.Request) {
 			return
 		}
 		if err := h.knowledge.Remove(name); err != nil {
-			writeJSON(w, http.StatusNotFound, map[string]string{"error": err.Error()})
+			// 之前所有失败一律 404，包括名称非法（400 的事）与真实 IO 错误
+			// （500 的事）。"啥都没删"和"服务器坏了"被混为一谈。
+			code, msg := knowledgeStatusFor(err)
+			writeJSON(w, code, map[string]string{"error": msg})
 			return
 		}
-		writeJSON(w, http.StatusOK, map[string]string{"status": "deleted", "name": name})
+		writeJSON(w, http.StatusOK, map[string]interface{}{"status": "deleted", "name": name})

 	default:
 		http.Error(w, "method not allowed", http.StatusMethodNotAllowed)
 	}
 }

+// maxKnowledgeUploadBytes 是知识库单次上传的体积上限。
+// 与 media CAS 的定位一致：知识条目的媒体是引用，不该拖着一堆原始字节。
+const maxKnowledgeUploadBytes = 32 << 20
+
+// handleKnowledgeUpload 处理 multipart 上传，按实际类型分流：媒体入 CAS，
+// 文本存正文。绝不把二进制当文本存。
+func (h *Handler) handleKnowledgeUpload(w http.ResponseWriter, r *http.Request) {
+	if err := r.ParseMultipartForm(maxKnowledgeUploadBytes); err != nil {
+		writeJSON(w, http.StatusBadRequest, map[string]string{"error": "invalid multipart: " + err.Error()})
+		return
+	}
+	name := strings.TrimSpace(r.FormValue("name"))
+	if name == "" {
+		writeJSON(w, http.StatusBadRequest, map[string]string{"error": "name is required"})
+		return
+	}
+
+	files := r.MultipartForm.File["file"]
+	// 也接受通用字段名，避免前端只有 file 字段名不匹配时静默走成"无媒体"
+	if len(files) == 0 {
+		files = r.MultipartForm.File["media"]
+	}
+	if len(files) == 0 {
+		writeJSON(w, http.StatusBadRequest, map[string]string{"error": "file field required"})
+		return
+	}
+	if len(files) > 16 {
+		writeJSON(w, http.StatusBadRequest, map[string]string{"error": "最多 16 个文件"})
+		return
+	}
+
+	var (
+		media    []sdk.KnowledgeMediaRef
+		texts    []string
+		rejected []string
+	)
+	for _, fh := range files {
+		f, err := fh.Open()
+		if err != nil {
+			writeJSON(w, http.StatusBadRequest, map[string]string{"error": "open " + fh.Filename + ": " + err.Error()})
+			return
+		}
+		data, err := io.ReadAll(io.LimitReader(f, maxKnowledgeUploadBytes+1))
+		f.Close()
+		if err != nil {
+			writeJSON(w, http.StatusBadRequest, map[string]string{"error": "read " + fh.Filename + ": " + err.Error()})
+			return
+		}
+		if len(data) == 0 {
+			rejected = append(rejected, fh.Filename+": 空文件")
+			continue
+		}
+		if len(data) > maxKnowledgeUploadBytes {
+			writeJSON(w, http.StatusRequestEntityTooLarge,
+				map[string]string{"error": fmt.Sprintf("%s 超过 %d 上限", fh.Filename, maxKnowledgeUploadBytes)})
+			return
+		}
+
+		mime := fh.Header.Get("Content-Type")
+		if mime == "" {
+			mime = contentTypeByExt(strings.ToLower(filepath.Ext(fh.Filename)))
+		}
+		// 按**探测到的真实类型**判定，而不是信客户端给的 Content-Type：
+		// 声明 text/plain 的 PNG 曾是真实场景，光看头会把二进制当文本存。
+		detected := http.DetectContentType(data)
+		if isMediaMIME(mime) || isMediaMIME(detected) {
+			useMIME := detected
+			if detected == "application/octet-stream" {
+				useMIME = mime
+			}
+			if h.mediaStore == nil {
+				writeJSON(w, http.StatusServiceUnavailable,
+					map[string]string{"error": "媒体存储未初始化，无法保存图片/音视频"})
+				return
+			}
+			digest, err := h.mediaStore.Put(data, useMIME, "webui_knowledge")
+			if err != nil {
+				writeJSON(w, http.StatusInternalServerError, map[string]string{"error": "store media: " + err.Error()})
+				return
+			}
+			media = append(media, sdk.KnowledgeMediaRef{Digest: digest, MIME: useMIME, Kind: mediaKindOf(useMIME)})
+			continue
+		}
+		// 文本类：确认是合法 UTF-8 才当正文，否则拒绝并说明原因
+		if !utf8.Valid(data) {
+			rejected = append(rejected, fmt.Sprintf("%s: 非文本内容且无法识别为媒体（type=%q）", fh.Filename, useMIMEOr(detected, mime)))
+			continue
+		}
+		texts = append(texts, string(data))
+	}
+
+	content := strings.TrimSpace(r.FormValue("content"))
+	if content == "" {
+		content = strings.Join(texts, "\n\n")
+	}
+	if strings.TrimSpace(content) == "" && len(media) == 0 {
+		msg := "没有可写入的内容"
+		if len(rejected) > 0 {
+			msg += "：" + strings.Join(rejected, "; ")
+		}
+		writeJSON(w, http.StatusBadRequest, map[string]string{"error": msg})
+		return
+	}
+
+	if err := h.knowledge.AddWithMedia(name, content, media); err != nil {
+		code, msg := knowledgeStatusFor(err)
+		writeJSON(w, code, map[string]string{"error": msg})
+		return
+	}
+	resp := map[string]interface{}{"status": "created", "name": name, "media": media}
+	if len(rejected) > 0 {
+		resp["rejected"] = rejected
+	}
+	writeJSON(w, http.StatusCreated, resp)
+}
+
+func useMIMEOr(vals ...string) string {
+	for _, v := range vals {
+		if v != "" {
+			return v
+		}
+	}
+	return ""
+}
+
+func isMediaMIME(m string) bool {
+	return strings.HasPrefix(m, "image/") || strings.HasPrefix(m, "audio/") || strings.HasPrefix(m, "video/")
+}
+
+func mediaKindOf(m string) string {
+	switch {
+	case strings.HasPrefix(m, "image/"):
+		return "image"
+	case strings.HasPrefix(m, "audio/"):
+		return "audio"
+	case strings.HasPrefix(m, "video/"):
+		return "video"
+	}
+	return "file"
+}
+
 func (h *Handler) handleTextMemory(w http.ResponseWriter, r *http.Request) {
 	if h.textMem == nil {
 		writeJSON(w, http.StatusServiceUnavailable, map[string]string{"error": "text memory not available"})