（用户看到一坨
+ // 机器码）；创建后不刷新计数（1644 行读的是 state.kernel 快照，创建
+ // 完仍是旧值）；没有任何删除入口，也没有媒体上传。
+ function knowEsc(v) {
+ return String(v == null ? "" : v)
+ .replace(/&/g, "&")
+ .replace(//g, ">");
+ }
+
+ async function refreshKnowledgeCount() {
+ try {
+ var d = await api("/knowledge");
+ var el = document.getElementById("know-count");
+ if (el && d && d.names) el.textContent = d.names.length;
+ // 稠密路（多模态）状态提示
+ var hint = document.getElementById("know-media-hint");
+ if (hint && d && d.dense) {
+ if (d.dense.enabled) {
+ hint.textContent =
+ __("多模态已就绪 ", "Multimodal ready ") +
+ (d.dense.ready || 0) +
+ "/" +
+ ((d.dense.ready || 0) + (d.dense.stale || 0));
+ } else {
+ hint.textContent = __(
+ "未接入多模态：图片只记录不参与召回",
+ "Multimodal off: images recorded but not searchable",
+ );
+ }
+ }
+ } catch (e) {
+ /* 计数刷新失败不该打断用户操作 */
+ }
+ }
+
+ function renderKnowledgeResults(views, q) {
+ if (!views || !views.length) {
+ return (
+ '' +
+ __("未找到相关知识", "No matching knowledge") +
+ "
"
+ );
+ }
+ var html = "";
+ views.forEach(function (v) {
+ html += '';
+ html += '';
+ html += '' + knowEsc(v.name) + "";
+ if (v.size) {
+ html += '' + v.size + " B";
+ }
+ html +=
+ '";
+ html += "";
+ if (v.media && v.media.length) {
+ html += '';
+ v.media.forEach(function (m) {
+ var kind = m.kind || "file";
+ html +=
+ '[' + knowEsc(kind) + "] " + knowEsc(m.digest.slice(0, 8)) + "…";
+ });
+ html += "";
+ }
+ if (v.preview) {
+ html +=
+ '' +
+ knowEsc(v.preview) +
+ "";
+ }
+ html += "";
+ });
+ return html;
+ }
+
async function searchKnowledgeChat() {
var q = document.getElementById("know-query")?.value;
+ var cat = document.getElementById("know-category")?.value || "";
var r = document.getElementById("know-result-chat");
- if (!r || !q) return;
+ if (!r) return;
+ if (!q) {
+ r.innerHTML =
+ '' +
+ __("请输入查询关键词", "Enter a keyword") +
+ "
";
+ return;
+ }
r.innerHTML = '';
try {
- var data = await api("/knowledge?q=" + encodeURIComponent(q));
- r.innerHTML =
- '' +
- escHtml(JSON.stringify(data, null, 2)) +
- "
";
+ var url = "/knowledge?q=" + encodeURIComponent(q);
+ if (cat) url += "&category=" + encodeURIComponent(cat);
+ var data = await api(url);
+ r.innerHTML = renderKnowledgeResults(data && data.results, q);
} catch (e) {
r.innerHTML =
- '' +
- __("搜索失败: ", "Search failed: ") +
- escHtml(e.message) +
- "
";
+ '' + __("搜索失败: ", "Search failed: ") + knowEsc(e.message) + "
";
+ }
+ }
+
+ async function deleteKnowledge(name) {
+ if (
+ !confirm(
+ __("确定删除知识「", "Delete knowledge \"") + name + __("」？此操作不可撤销。", "\"? This cannot be undone."),
+ )
+ ) {
+ return;
+ }
+ try {
+ await api("/knowledge?name=" + encodeURIComponent(name), { method: "DELETE" });
+ toast(__("知识「", 'Knowledge "') + name + __("」已删除", '" deleted'));
+ refreshKnowledgeCount();
+ searchKnowledgeChat();
+ } catch (e) {
+ toast(__("删除失败: ", "Delete failed: ") + e.message, true);
}
}
async function createKnowledgeChat() {
var name = document.getElementById("know-name")?.value;
var content = document.getElementById("know-content")?.value;
- if (!name || !content) {
+ var fileInput = document.getElementById("know-media");
+ var files = fileInput && fileInput.files ? fileInput.files : null;
+ if (!name) {
+ toast(__("名称不能为空", "Name is required"), true);
+ return;
+ }
+ if ((!content || !content.trim()) && (!files || !files.length)) {
toast(
- __("名称和内容不能为空", "Name and content cannot be empty"),
+ __("内容与媒体至少要有一项", "Content or media is required"),
true,
);
return;
}
try {
- var r = await api("/knowledge", {
- method: "POST",
- body: JSON.stringify({ name: name, content: content }),
- });
- if (r.status || r.id) {
- toast(
- __("知识「", 'Knowledge "') + name + __("」已创建", '" created'),
- );
- document.getElementById("know-name").value = "";
- document.getElementById("know-content").value = "";
+ var r;
+ if (files && files.length) {
+ // 有文件走 multipart：服务端按**探测到的真实类型**分流，
+ // 图片/音视频入媒体库并按 digest 挂到条目上，文本存正文。
+ var fd = new FormData();
+ fd.append("name", name);
+ if (content) fd.append("content", content);
+ for (var i = 0; i < files.length; i++) fd.append("file", files[i]);
+ r = await api("/knowledge", { method: "POST", body: fd });
} else {
- toast(__("创建失败", "Create failed"), true);
+ r = await api("/knowledge", {
+ method: "POST",
+ body: JSON.stringify({ name: name, content: content }),
+ });
}
+ var msg = __("知识「", 'Knowledge "') + name + __("」已创建", '" created');
+ if (r && r.media) msg += __("，含 ", " with ") + r.media + __(" 个媒体", " media item(s)");
+ if (r && r.rejected && r.rejected.length) {
+ msg += __("；", "; ") + r.rejected.length + __(" 项被跳过", " skipped");
+ }
+ toast(msg);
+ document.getElementById("know-name").value = "";
+ document.getElementById("know-content").value = "";
+ if (fileInput) fileInput.value = "";
+ refreshKnowledgeCount();
} catch (e) {
toast(__("创建失败: ", "Create failed: ") + e.message, true);
}
}
function switchChatPanel(tab, el) {
+ // 切到知识面板时拉实时计数：面板里的数字来自 state.kernel 快照，
+ // 而知识条目会经工具/上传增删，快照不会自己变（实测创建后仍显示 "-"）。
+ if (tab === "knowledge") refreshKnowledgeCount();
var panels = {
chat: document.getElementById("chat-panel-chat"),
starmap: document.getElementById("chat-panel-starmap"),
diff --git a/internal/plugins/webui/handler.go b/internal/plugins/webui/handler.go
index 64d93d6..5555ba6 100644
--- a/internal/plugins/webui/handler.go
+++ b/internal/plugins/webui/handler.go
@@ -100,6 +100,7 @@ type Handler struct {
status sdk.StatusAPI
term sdk.TerminalAPI
llm sdk.LLMAPI
+ mediaStore sdk.MediaAPI
// hostPort 是 webui 实际监听的 ":port"（用于推导服务入口链接）。
hostPort string
@@ -161,6 +162,7 @@ func NewHandler(s *sdk.PluginSDK) *Handler {
st sdk.StatusAPI
term sdk.TerminalAPI
llm sdk.LLMAPI
+ ms sdk.MediaAPI
)
if s != nil {
sup, mem, idx = s.Supervisor(), s.Memory(), s.Indexer()
@@ -169,6 +171,7 @@ func NewHandler(s *sdk.PluginSDK) *Handler {
se, pm = s.Settings(), s.PluginMgr()
st, llm = s.Status(), s.LLM()
term = s.Terminal()
+ ms = s.Media()
}
h := &Handler{
sdk: s,
@@ -186,6 +189,7 @@ func NewHandler(s *sdk.PluginSDK) *Handler {
status: st,
term: term,
llm: llm,
+ mediaStore: ms,
sessions: make(map[string]time.Time),
loginLimiter: newLoginLimiter(loginMaxFails, loginWindow),
pendingIdx: -1,
diff --git a/internal/plugins/webui/handler_knowledge_test.go b/internal/plugins/webui/handler_knowledge_test.go
new file mode 100644
index 0000000..f2e855d
--- /dev/null
+++ b/internal/plugins/webui/handler_knowledge_test.go
@@ -0,0 +1,508 @@
+package webui
+
+import (
+ "bytes"
+ "encoding/json"
+ "image"
+ "image/color"
+ "image/png"
+ "mime/multipart"
+ "net/http"
+ "net/http/httptest"
+ "strings"
+ "testing"
+ "time"
+
+ "path/filepath"
+
+ "gitcode.com/JianFeeeee/HomeAgent/internal/knowledge"
+ "gitcode.com/JianFeeeee/HomeAgent/internal/memory/media"
+ "gitcode.com/JianFeeeee/HomeAgent/internal/sdk"
+ "gitcode.com/JianFeeeee/HomeAgent/internal/supervisor"
+ "gitcode.com/JianFeeeee/HomeAgent/pkg/types"
+)
+
+// newKnowledgeHandler 构造一个带知识库（可选媒体存储）的 Handler。
+func newKnowledgeHandler(t *testing.T, withMedia bool) (*Handler, *knowledge.Store, func()) {
+ t.Helper()
+ ks := knowledge.NewStore(t.TempDir())
+ if err := ks.Start(); err != nil {
+ t.Fatal(err)
+ }
+
+ cfg := &types.Config{Daemon: types.DaemonConfig{
+ CheckInterval: time.Minute, HeartbeatInterval: 30 * time.Second,
+ }}
+ sup := supervisor.New(cfg)
+ sup.Start()
+
+ sdkCfg := sdk.SDKConfig{
+ Supervisor: supervisor.NewSDKAdapter(sup),
+ Knowledge: sdk.NewKnowledge(ks),
+ Config: sdk.NewConfig(cfg),
+ }
+ if withMedia {
+ ms, err := media.New(t.TempDir())
+ if err != nil {
+ t.Fatal(err)
+ }
+ sdkCfg.Media = sdk.NewMedia(ms)
+ }
+ h := NewHandler(testSDK(sdkCfg))
+ return h, ks, func() { sup.Shutdown() }
+}
+
+// 上传 PNG 绝不能被当文本存成乱码正文。
+//
+// 修复前的行为：任何文件都是 file.Read 后 string(buf[:n]) 直接当 Markdown
+// 存进 content.md —— 传张图得到一份乱码文本知识，且无任何迹象。
+func TestKnowledgeUploadPNGIsNotStoredAsText(t *testing.T) {
+ h, ks, done := newKnowledgeHandler(t, true)
+ defer done()
+
+ raw := makePNG(t)
+
+ var buf bytes.Buffer
+ mw := multipart.NewWriter(&buf)
+ _ = mw.WriteField("name", "cat-photo")
+ fw, err := mw.CreateFormFile("file", "cat.png")
+ if err != nil {
+ t.Fatal(err)
+ }
+ if _, err := fw.Write(raw); err != nil {
+ t.Fatal(err)
+ }
+ mw.Close()
+
+ req := httptest.NewRequest(http.MethodPost, "/api/v1/knowledge", &buf)
+ req.Header.Set("Content-Type", mw.FormDataContentType())
+ w := httptest.NewRecorder()
+ h.handleKnowledge(w, req)
+
+ if w.Code != http.StatusCreated {
+ t.Fatalf("应 201，实为 %d: %s", w.Code, w.Body.String())
+ }
+
+ k := knowledgeEntry(t, ks, "cat-photo")
+ if k == nil {
+ t.Fatalf("条目未创建，List=%v", ks.List())
+ }
+ // 正文里绝不能出现 PNG 字节被强转后的乱码
+ if k.Content != "" {
+ for i, r := range k.Content {
+ if r == 0xFFFD || r == 0 {
+ t.Fatalf("正文含二进制强转的乱码（偏移 %d），说明 PNG 被当文本存了", i)
+ }
+ }
+ if !isPrintableOrSpace(k.Content) {
+ t.Errorf("正文含不可打印字符，非文本被当 Markdown 存了")
+ }
+ }
+ if len(k.Media) != 1 {
+ t.Fatalf("应挂 1 个媒体，实为 %+v", k.Media)
+ }
+ if k.Media[0].Digest == "" {
+ t.Errorf("媒体 digest 为空: %+v", k.Media[0])
+ }
+ if !strings.HasPrefix(k.Media[0].MIME, "image/") {
+ t.Errorf("媒体 MIME 应为 image/*，实为 %q", k.Media[0].MIME)
+ }
+}
+
+// 声明 text/plain 但实际是 PNG（Content-Type 不可信）也必须入 CAS。
+func TestKnowledgeUploadDetectsRealTypeOverClaimedHeader(t *testing.T) {
+ h, ks, done := newKnowledgeHandler(t, true)
+ defer done()
+
+ raw := makePNG(t)
+ var buf bytes.Buffer
+ mw := multipart.NewWriter(&buf)
+ _ = mw.WriteField("name", "liar")
+ // 故意谎报为纯文本
+ fw, _ := mw.CreateFormFile("file", "x.png")
+ fw.Write(raw)
+ mw.Close()
+
+ req := httptest.NewRequest(http.MethodPost, "/api/v1/knowledge", &buf)
+ req.Header.Set("Content-Type", mw.FormDataContentType())
+ w := httptest.NewRecorder()
+ h.handleKnowledge(w, req)
+ if w.Code != http.StatusCreated {
+ t.Fatalf("应 201，实为 %d: %s", w.Code, w.Body.String())
+ }
+ k := knowledgeEntry(t, ks, "liar")
+ if k == nil || len(k.Media) != 1 {
+ t.Fatalf("谎报 Content-Type 的 PNG 应被探测为媒体并入 CAS，实为 %+v", k)
+ }
+}
+
+// 纯文本上传仍走原路径存正文。
+func TestKnowledgeUploadTextStaysAsContent(t *testing.T) {
+ h, ks, done := newKnowledgeHandler(t, true)
+ defer done()
+
+ var buf bytes.Buffer
+ mw := multipart.NewWriter(&buf)
+ _ = mw.WriteField("name", "notes")
+ fw, _ := mw.CreateFormFile("file", "n.txt")
+ fw.Write([]byte("这是正文内容"))
+ mw.Close()
+
+ req := httptest.NewRequest(http.MethodPost, "/api/v1/knowledge", &buf)
+ req.Header.Set("Content-Type", mw.FormDataContentType())
+ w := httptest.NewRecorder()
+ h.handleKnowledge(w, req)
+ if w.Code != http.StatusCreated {
+ t.Fatalf("应 201，实为 %d: %s", w.Code, w.Body.String())
+ }
+ k := knowledgeEntry(t, ks, "notes")
+ if k == nil {
+ t.Fatal("条目未创建")
+ }
+ if !strings.Contains(k.Content, "这是正文内容") {
+ t.Errorf("文本应存为正文，实为 %q", k.Content)
+ }
+ if len(k.Media) != 0 {
+ t.Errorf("文本不该有媒体，实为 %+v", k.Media)
+ }
+}
+
+// 非文本也非媒体的垃圾字节应被明确拒绝，而不是当文本存。
+func TestKnowledgeUploadRejectsBinaryGarbage(t *testing.T) {
+ h, ks, done := newKnowledgeHandler(t, true)
+ defer done()
+
+ // 非法 UTF-8 且探测不出媒体
+ garbage := []byte{0xFF, 0xFE, 0x00, 0x01, 0x02, 0x03, 0xFF, 0xFE, 0x00}
+ var buf bytes.Buffer
+ mw := multipart.NewWriter(&buf)
+ _ = mw.WriteField("name", "junk")
+ fw, _ := mw.CreateFormFile("file", "j.bin")
+ fw.Write(garbage)
+ mw.Close()
+
+ req := httptest.NewRequest(http.MethodPost, "/api/v1/knowledge", &buf)
+ req.Header.Set("Content-Type", mw.FormDataContentType())
+ w := httptest.NewRecorder()
+ h.handleKnowledge(w, req)
+
+ if w.Code != http.StatusBadRequest {
+ t.Fatalf("垃圾字节应 400，实为 %d: %s", w.Code, w.Body.String())
+ }
+ if knowledgeEntry(t, ks, "junk") != nil {
+ t.Error("被拒绝的条目不该落库")
+ }
+}
+
+// 媒体存储未初始化时上传图片必须 503，不能静默把二进制当文本存。
+func TestKnowledgeUploadMediaWithoutStoreIsUnavailable(t *testing.T) {
+ h, _, done := newKnowledgeHandler(t, false)
+ defer done()
+
+ raw := makePNG(t)
+ var buf bytes.Buffer
+ mw := multipart.NewWriter(&buf)
+ _ = mw.WriteField("name", "x")
+ fw, _ := mw.CreateFormFile("file", "a.png")
+ fw.Write(raw)
+ mw.Close()
+
+ req := httptest.NewRequest(http.MethodPost, "/api/v1/knowledge", &buf)
+ req.Header.Set("Content-Type", mw.FormDataContentType())
+ w := httptest.NewRecorder()
+ h.handleKnowledge(w, req)
+ if w.Code != http.StatusServiceUnavailable {
+ t.Errorf("无媒体存储应 503，实为 %d: %s", w.Code, w.Body.String())
+ }
+}
+
+// 状态码语义：名称非法是 400，不存在是 404，缺名是 400。
+func TestKnowledgeStatusCodes(t *testing.T) {
+ h, _, done := newKnowledgeHandler(t, true)
+ defer done()
+
+ // 非法名称（..）
+ body := `{"name":"../evil","content":"x"}`
+ req := httptest.NewRequest(http.MethodPost, "/api/v1/knowledge", strings.NewReader(body))
+ req.Header.Set("Content-Type", "application/json")
+ w := httptest.NewRecorder()
+ h.handleKnowledge(w, req)
+ if w.Code != http.StatusBadRequest {
+ t.Errorf("非法名称应 400，实为 %d: %s", w.Code, w.Body.String())
+ }
+
+ // 缺名
+ req = httptest.NewRequest(http.MethodPost, "/api/v1/knowledge", strings.NewReader(`{"content":"x"}`))
+ req.Header.Set("Content-Type", "application/json")
+ w = httptest.NewRecorder()
+ h.handleKnowledge(w, req)
+ if w.Code != http.StatusBadRequest {
+ t.Errorf("缺名应 400，实为 %d", w.Code)
+ }
+
+ // 删除不存在
+ req = httptest.NewRequest(http.MethodDelete, "/api/v1/knowledge?name=nope", nil)
+ w = httptest.NewRecorder()
+ h.handleKnowledge(w, req)
+ if w.Code != http.StatusNotFound {
+ t.Errorf("删除不存在应 404，实为 %d", w.Code)
+ }
+}
+
+// 搜索结果应带分类/预览等前端需要的字段，而不是裸 JSON。
+func TestKnowledgeSearchReturnsViews(t *testing.T) {
+ h, ks, done := newKnowledgeHandler(t, true)
+ defer done()
+ if err := ks.Add("tech/go/并发", "goroutine 调度 GMP 抢占"); err != nil {
+ t.Fatal(err)
+ }
+
+ req := httptest.NewRequest(http.MethodGet, "/api/v1/knowledge?q=GMP", nil)
+ w := httptest.NewRecorder()
+ h.handleKnowledge(w, req)
+ if w.Code != http.StatusOK {
+ t.Fatalf("应 200，实为 %d: %s", w.Code, w.Body.String())
+ }
+ var resp struct {
+ Results []knowledgeView `json:"results"`
+ }
+ if err := json.Unmarshal(w.Body.Bytes(), &resp); err != nil {
+ t.Fatal(err)
+ }
+ if len(resp.Results) == 0 {
+ t.Fatal("应命中")
+ }
+ r := resp.Results[0]
+ if r.Name != "tech/go/并发" {
+ t.Errorf("name 不对: %q", r.Name)
+ }
+ if r.Preview == "" {
+ t.Error("缺 preview")
+ }
+ // 不得泄露服务端绝对路径
+ if strings.Contains(r.Preview, string(filepath.Separator)) && strings.Contains(r.Preview, "tmp") {
+ t.Errorf("疑似泄露服务端路径: %q", r.Preview)
+ }
+}
+
+// 分类过滤参数应真正生效。
+func TestKnowledgeSearchCategoryFilter(t *testing.T) {
+ h, ks, done := newKnowledgeHandler(t, true)
+ defer done()
+ for _, e := range []struct{ n, c string }{
+ {"tech/go/a", "并发 调度"},
+ {"life/b", "作息 睡眠"},
+ } {
+ if err := ks.Add(e.n, e.c); err != nil {
+ t.Fatal(err)
+ }
+ }
+ req := httptest.NewRequest(http.MethodGet, "/api/v1/knowledge?q=%E5%B9%B2%E8%8D%89&category=tech", nil)
+ w := httptest.NewRecorder()
+ h.handleKnowledge(w, req)
+ if w.Code != http.StatusOK {
+ t.Fatalf("应 200，实为 %d: %s", w.Code, w.Body.String())
+ }
+ var resp struct {
+ Results []knowledgeView `json:"results"`
+ }
+ json.Unmarshal(w.Body.Bytes(), &resp)
+ for _, r := range resp.Results {
+ if !strings.HasPrefix(r.Name, "tech/") {
+ t.Errorf("分类过滤失效，混入 %q", r.Name)
+ }
+ }
+}
+
+// 列表端点应返回 names + stats（供前端刷新计数），并附稠密路状态。
+func TestKnowledgeListReturnsStats(t *testing.T) {
+ h, ks, done := newKnowledgeHandler(t, true)
+ defer done()
+ _ = ks.Add("a", "A")
+ _ = ks.Add("b", "B")
+
+ req := httptest.NewRequest(http.MethodGet, "/api/v1/knowledge", nil)
+ w := httptest.NewRecorder()
+ h.handleKnowledge(w, req)
+ if w.Code != http.StatusOK {
+ t.Fatalf("应 200，实为 %d", w.Code)
+ }
+ var resp struct {
+ Names []string `json:"names"`
+ Stats map[string]interface{} `json:"stats"`
+ Dense map[string]interface{} `json:"dense"`
+ }
+ if err := json.Unmarshal(w.Body.Bytes(), &resp); err != nil {
+ t.Fatal(err)
+ }
+ if len(resp.Names) != 2 {
+ t.Errorf("names 应有 2 条，实为 %v", resp.Names)
+ }
+ if resp.Stats == nil {
+ t.Error("缺 stats")
+ }
+ if resp.Dense == nil {
+ t.Error("缺 dense 状态（前端要据此提示多模态是否就绪）")
+ }
+}
+
+// knowledgeEntry 按名字取一条知识（含媒体引用与正文）。
+// Store 没有导出的 Items()，测试里用「全库检索 + 名字匹配」拿到同一条。
+func knowledgeEntry(t *testing.T, ks *knowledge.Store, name string) *knowledge.Knowledge {
+ t.Helper()
+ for _, k := range ks.Search(name, 100) {
+ if k.Name == name {
+ return k
+ }
+ }
+ // 检索可能因分词而漏，退回遍历 List + 逐条检索
+ for _, n := range ks.List() {
+ if n != name {
+ continue
+ }
+ for _, k := range ks.Search(n, 100) {
+ if k.Name == name {
+ return k
+ }
+ }
+ }
+ return nil
+}
+
+func makePNG(t *testing.T) []byte {
+ t.Helper()
+ img := image.NewRGBA(image.Rect(0, 0, 4, 4))
+ img.Set(0, 0, color.RGBA{R: 255, A: 255})
+ var buf bytes.Buffer
+ if err := png.Encode(&buf, img); err != nil {
+ t.Fatal(err)
+ }
+ return buf.Bytes()
+}
+
+func isPrintableOrSpace(s string) bool {
+ for _, r := range s {
+ if r < 0x20 && r != '\n' && r != '\t' && r != '\r' {
+ return false
+ }
+ }
+ return true
+}
+
+// 前端实际发的 multipart 形状：name + 可选 content + 多个 file 字段。
+// 前端在 dashboard.js 的 createKnowledgeChat 里用 FormData 组装，
+// 逐个 append("file", files[i]) —— 这里逐字复刻，确认服务端吃得下。
+func TestKnowledgeUploadMultipleFilesFromFrontend(t *testing.T) {
+ h, ks, done := newKnowledgeHandler(t, true)
+ defer done()
+
+ var buf bytes.Buffer
+ mw := multipart.NewWriter(&buf)
+ _ = mw.WriteField("name", "cat-dog")
+ _ = mw.WriteField("content", "两只动物")
+ for i, name := range []string{"a.png", "b.png"} {
+ fw, _ := mw.CreateFormFile("file", name)
+ fw.Write(makePNG(t))
+ _ = i
+ }
+ mw.Close()
+
+ req := httptest.NewRequest(http.MethodPost, "/api/v1/knowledge", &buf)
+ req.Header.Set("Content-Type", mw.FormDataContentType())
+ w := httptest.NewRecorder()
+ h.handleKnowledge(w, req)
+ if w.Code != http.StatusCreated {
+ t.Fatalf("应 201，实为 %d: %s", w.Code, w.Body.String())
+ }
+ var resp struct {
+ Media []sdk.KnowledgeMediaRef `json:"media"`
+ }
+ if err := json.Unmarshal(w.Body.Bytes(), &resp); err != nil {
+ t.Fatal(err)
+ }
+ // 同一张图传两次 → 内容寻址去重，两个 digest 相同（这是 CAS 的预期行为）
+ if len(resp.Media) != 2 {
+ t.Fatalf("应返回 2 个媒体，实为 %d", len(resp.Media))
+ }
+ if resp.Media[0].Digest != resp.Media[1].Digest {
+ t.Errorf("相同内容应去重为同一 digest，实为 %s vs %s", resp.Media[0].Digest, resp.Media[1].Digest)
+ }
+ k := knowledgeEntry(t, ks, "cat-dog")
+ if k == nil {
+ t.Fatal("条目未创建")
+ }
+ if k.Content != "两只动物" {
+ t.Errorf("显式 content 应被采用，实为 %q", k.Content)
+ }
+}
+
+// 文本文件不带 content 字段时，文本内容应被采纳为正文。
+func TestKnowledgeUploadTextFileBecomesContent(t *testing.T) {
+ h, ks, done := newKnowledgeHandler(t, true)
+ defer done()
+
+ var buf bytes.Buffer
+ mw := multipart.NewWriter(&buf)
+ _ = mw.WriteField("name", "fromfile")
+ fw, _ := mw.CreateFormFile("file", "a.md")
+ fw.Write([]byte("# 标题\n\n正文"))
+ mw.Close()
+
+ req := httptest.NewRequest(http.MethodPost, "/api/v1/knowledge", &buf)
+ req.Header.Set("Content-Type", mw.FormDataContentType())
+ w := httptest.NewRecorder()
+ h.handleKnowledge(w, req)
+ if w.Code != http.StatusCreated {
+ t.Fatalf("应 201，实为 %d: %s", w.Code, w.Body.String())
+ }
+ k := knowledgeEntry(t, ks, "fromfile")
+ if k == nil || !strings.Contains(k.Content, "正文") {
+ t.Errorf("文本文件内容应成为正文，实为 %+v", k)
+ }
+}
+
+// 上传的媒体必须能在媒体库里取回字节（digest 有效），否则引用是死的。
+func TestKnowledgeUploadedMediaRetrievable(t *testing.T) {
+ h, ks, done := newKnowledgeHandler(t, true)
+ defer done()
+
+ raw := makePNG(t)
+ var buf bytes.Buffer
+ mw := multipart.NewWriter(&buf)
+ _ = mw.WriteField("name", "m")
+ fw, _ := mw.CreateFormFile("file", "a.png")
+ fw.Write(raw)
+ mw.Close()
+
+ req := httptest.NewRequest(http.MethodPost, "/api/v1/knowledge", &buf)
+ req.Header.Set("Content-Type", mw.FormDataContentType())
+ w := httptest.NewRecorder()
+ h.handleKnowledge(w, req)
+ if w.Code != http.StatusCreated {
+ t.Fatal(w.Body.String())
+ }
+ var resp struct {
+ Media []sdk.KnowledgeMediaRef `json:"media"`
+ }
+ json.Unmarshal(w.Body.Bytes(), &resp)
+ if len(resp.Media) != 1 {
+ t.Fatalf("应 1 个媒体，实为 %d", len(resp.Media))
+ }
+ // Handler 持有的媒体存储应能按 digest 取回原始字节
+ got, err := h.mediaStore.Get(resp.Media[0].Digest)
+ if err != nil {
+ t.Fatalf("媒体不可取回: %v", err)
+ }
+ if !bytes.Equal(got, raw) {
+ t.Error("取回的字节与上传的不一致")
+ }
+ info, err := h.mediaStore.Stat(resp.Media[0].Digest)
+ if err != nil {
+ t.Fatal(err)
+ }
+ if info.Kind != "image" {
+ t.Errorf("kind 应为 image，实为 %q", info.Kind)
+ }
+ // 条目上也应持久化了引用（重启不丢）
+ _ = ks.Flush()
+}
diff --git a/internal/plugins/webui/handler_memory.go b/internal/plugins/webui/handler_memory.go
index 1b2efd2..b96ad33 100644
--- 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"})