@scotthuang/engram 0.5.8 → 0.6.0

package/dist/src/__tests__/profile.test.js

@@ -0,0 +1,130 @@
+import { describe, it, expect, beforeEach } from "vitest";
+import { ProfileManager, EMPTY_PROFILE } from "../profile.js";
+import { promises as fs } from "node:fs";
+import { join } from "node:path";
+import { tmpdir } from "node:os";
+describe("ProfileManager", () => {
+    let manager;
+    let tempDir;
+    beforeEach(async () => {
+        tempDir = join(tmpdir(), `engram-test-${Date.now()}`);
+        await fs.mkdir(tempDir, { recursive: true });
+        manager = new ProfileManager(tempDir);
+    });
+    describe("load", () => {
+        it("returns empty profile when no file exists", async () => {
+            const profile = await manager.load();
+            expect(profile.summary).toBe("");
+            expect(profile.coreTags).toEqual([]);
+            expect(profile.tags).toEqual({});
+        });
+        it("loads existing profile from file", async () => {
+            const profileDir = join(tempDir, "memory", "profile");
+            await fs.mkdir(profileDir, { recursive: true });
+            const saved = { ...EMPTY_PROFILE, summary: "test summary", coreTags: ["tag1"] };
+            await fs.writeFile(join(profileDir, "semantic_profile.json"), JSON.stringify(saved));
+            // 创建新的 manager 实例测试从文件加载
+            const manager2 = new ProfileManager(tempDir);
+            const loaded = await manager2.load();
+            expect(loaded.summary).toBe("test summary");
+            expect(loaded.coreTags).toEqual(["tag1"]);
+        });
+    });
+    describe("addTag", () => {
+        it("adds a new tag to a new dimension", async () => {
+            const profile = { ...EMPTY_PROFILE };
+            const result = manager.addTag(profile, "口味偏好", "喜欢辣");
+            expect(result.tags["口味偏好"]).toHaveLength(1);
+            expect(result.tags["口味偏好"][0].value).toBe("喜欢辣");
+            expect(result.tags["口味偏好"][0].confidence).toBe(0.7);
+        });
+        it("increases confidence for existing tag", async () => {
+            const profile = {
+                ...EMPTY_PROFILE,
+                tags: { 口味偏好: [{ value: "喜欢辣", confidence: 0.5, lastSeen: "2026-01-01" }] },
+            };
+            const result = manager.addTag(profile, "口味偏好", "喜欢辣");
+            expect(result.tags["口味偏好"]).toHaveLength(1);
+            expect(result.tags["口味偏好"][0].confidence).toBeCloseTo(0.6);
+        });
+        it("adds multiple tags to same dimension", async () => {
+            const profile = { ...EMPTY_PROFILE };
+            manager.addTag(profile, "口味偏好", "喜欢辣");
+            manager.addTag(profile, "口味偏好", "不吃香菜");
+            expect(profile.tags["口味偏好"]).toHaveLength(2);
+        });
+    });
+    describe("decayTags", () => {
+        it("reduces confidence of all tags", () => {
+            const profile = {
+                ...EMPTY_PROFILE,
+                tags: {
+                    口味偏好: [
+                        { value: "喜欢辣", confidence: 0.9, lastSeen: "2026-03-17" },
+                    ],
+                },
+            };
+            const result = manager.decayTags(profile, 0.5);
+            expect(result.tags["口味偏好"][0].confidence).toBeCloseTo(0.45);
+        });
+        it("removes tags below threshold", () => {
+            const profile = {
+                ...EMPTY_PROFILE,
+                tags: {
+                    过时: [{ value: "旧标签", confidence: 0.15, lastSeen: "2026-01-01" }],
+                },
+            };
+            const result = manager.decayTags(profile, 1.0);
+            expect(result.tags["过时"]).toBeUndefined();
+        });
+        it("removes empty dimensions after filtering", () => {
+            const profile = {
+                ...EMPTY_PROFILE,
+                tags: {
+                    空维度: [{ value: "很低", confidence: 0.1, lastSeen: "2026-01-01" }],
+                },
+            };
+            const result = manager.decayTags(profile, 1.0);
+            expect("空维度" in result.tags).toBe(false);
+        });
+    });
+    describe("getRecallContext", () => {
+        it("returns empty string for empty profile", () => {
+            const ctx = manager.getRecallContext(EMPTY_PROFILE);
+            expect(ctx).toBe("");
+        });
+        it("returns summary and core tags", () => {
+            const profile = { ...EMPTY_PROFILE, summary: "辣味中餐爱好者", coreTags: ["辣味中餐", "天河"] };
+            const ctx = manager.getRecallContext(profile);
+            expect(ctx).toContain("辣味中餐爱好者");
+            expect(ctx).toContain("辣味中餐");
+            expect(ctx).toContain("天河");
+        });
+        it("works with only core tags", () => {
+            const profile = { ...EMPTY_PROFILE, summary: "", coreTags: ["标签1"] };
+            const ctx = manager.getRecallContext(profile);
+            expect(ctx).toContain("标签1");
+        });
+    });
+    describe("save", () => {
+        it("saves profile to file", async () => {
+            const profile = { ...EMPTY_PROFILE, summary: "test" };
+            await manager.save(profile);
+            const raw = await fs.readFile(join(tempDir, "memory", "profile", "semantic_profile.json"), "utf-8");
+            const loaded = JSON.parse(raw);
+            expect(loaded.summary).toBe("test");
+        });
+        it("updates updatedAt on save", async () => {
+            const profile = { ...EMPTY_PROFILE };
+            const before = new Date();
+            await manager.save(profile);
+            const after = new Date();
+            const raw = await fs.readFile(join(tempDir, "memory", "profile", "semantic_profile.json"), "utf-8");
+            const loaded = JSON.parse(raw);
+            const updated = new Date(loaded.updatedAt);
+            expect(updated.getTime()).toBeGreaterThanOrEqual(before.getTime());
+            expect(updated.getTime()).toBeLessThanOrEqual(after.getTime());
+        });
+    });
+});
+//# sourceMappingURL=profile.test.js.map

package/dist/src/__tests__/profile.test.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"profile.test.js","sourceRoot":"","sources":["../../../src/__tests__/profile.test.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,QAAQ,EAAE,EAAE,EAAE,MAAM,EAAE,UAAU,EAAE,MAAM,QAAQ,CAAC;AAC1D,OAAO,EAAE,cAAc,EAAE,aAAa,EAAE,MAAM,eAAe,CAAC;AAC9D,OAAO,EAAE,QAAQ,IAAI,EAAE,EAAE,MAAM,SAAS,CAAC;AACzC,OAAO,EAAE,IAAI,EAAE,MAAM,WAAW,CAAC;AACjC,OAAO,EAAE,MAAM,EAAE,MAAM,SAAS,CAAC;AAEjC,QAAQ,CAAC,gBAAgB,EAAE,GAAG,EAAE;IAC9B,IAAI,OAAuB,CAAC;IAC5B,IAAI,OAAe,CAAC;IAEpB,UAAU,CAAC,KAAK,IAAI,EAAE;QACpB,OAAO,GAAG,IAAI,CAAC,MAAM,EAAE,EAAE,eAAe,IAAI,CAAC,GAAG,EAAE,EAAE,CAAC,CAAC;QACtD,MAAM,EAAE,CAAC,KAAK,CAAC,OAAO,EAAE,EAAE,SAAS,EAAE,IAAI,EAAE,CAAC,CAAC;QAC7C,OAAO,GAAG,IAAI,cAAc,CAAC,OAAO,CAAC,CAAC;IACxC,CAAC,CAAC,CAAC;IAEH,QAAQ,CAAC,MAAM,EAAE,GAAG,EAAE;QACpB,EAAE,CAAC,2CAA2C,EAAE,KAAK,IAAI,EAAE;YACzD,MAAM,OAAO,GAAG,MAAM,OAAO,CAAC,IAAI,EAAE,CAAC;YACrC,MAAM,CAAC,OAAO,CAAC,OAAO,CAAC,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;YACjC,MAAM,CAAC,OAAO,CAAC,QAAQ,CAAC,CAAC,OAAO,CAAC,EAAE,CAAC,CAAC;YACrC,MAAM,CAAC,OAAO,CAAC,IAAI,CAAC,CAAC,OAAO,CAAC,EAAE,CAAC,CAAC;QACnC,CAAC,CAAC,CAAC;QAEH,EAAE,CAAC,kCAAkC,EAAE,KAAK,IAAI,EAAE;YAChD,MAAM,UAAU,GAAG,IAAI,CAAC,OAAO,EAAE,QAAQ,EAAE,SAAS,CAAC,CAAC;YACtD,MAAM,EAAE,CAAC,KAAK,CAAC,UAAU,EAAE,EAAE,SAAS,EAAE,IAAI,EAAE,CAAC,CAAC;YAChD,MAAM,KAAK,GAAG,EAAE,GAAG,aAAa,EAAE,OAAO,EAAE,cAAc,EAAE,QAAQ,EAAE,CAAC,MAAM,CAAC,EAAE,CAAC;YAChF,MAAM,EAAE,CAAC,SAAS,CAAC,IAAI,CAAC,UAAU,EAAE,uBAAuB,CAAC,EAAE,IAAI,CAAC,SAAS,CAAC,KAAK,CAAC,CAAC,CAAC;YAErF,yBAAyB;YACzB,MAAM,QAAQ,GAAG,IAAI,cAAc,CAAC,OAAO,CAAC,CAAC;YAC7C,MAAM,MAAM,GAAG,MAAM,QAAQ,CAAC,IAAI,EAAE,CAAC;YACrC,MAAM,CAAC,MAAM,CAAC,OAAO,CAAC,CAAC,IAAI,CAAC,cAAc,CAAC,CAAC;YAC5C,MAAM,CAAC,MAAM,CAAC,QAAQ,CAAC,CAAC,OAAO,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC;QAC5C,CAAC,CAAC,CAAC;IACL,CAAC,CAAC,CAAC;IAEH,QAAQ,CAAC,QAAQ,EAAE,GAAG,EAAE;QACtB,EAAE,CAAC,mCAAmC,EAAE,KAAK,IAAI,EAAE;YACjD,MAAM,OAAO,GAAG,EAAE,GAAG,aAAa,EAAE,CAAC;YACrC,MAAM,MAAM,GAAG,OAAO,CAAC,MAAM,CAAC,OAAO,EAAE,MAAM,EAAE,KAAK,CAAC,CAAC;YACtD,MAAM,CAAC,MAAM,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC,CAAC,YAAY,CAAC,CAAC,CAAC,CAAC;YAC5C,MAAM,CAAC,MAAM,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC;YACjD,MAAM,CAAC,MAAM,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC,CAAC,UAAU,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;QACtD,CAAC,CAAC,CAAC;QAEH,EAAE,CAAC,uCAAuC,EAAE,KAAK,IAAI,EAAE;YACrD,MAAM,OAAO,GAAQ;gBACnB,GAAG,aAAa;gBAChB,IAAI,EAAE,EAAE,IAAI,EAAE,CAAC,EAAE,KAAK,EAAE,KAAK,EAAE,UAAU,EAAE,GAAG,EAAE,QAAQ,EAAE,YAAY,EAAE,CAAC,EAAE;aAC5E,CAAC;YACF,MAAM,MAAM,GAAG,OAAO,CAAC,MAAM,CAAC,OAAO,EAAE,MAAM,EAAE,KAAK,CAAC,CAAC;YACtD,MAAM,CAAC,MAAM,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC,CAAC,YAAY,CAAC,CAAC,CAAC,CAAC;YAC5C,MAAM,CAAC,MAAM,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC,CAAC,UAAU,CAAC,CAAC,WAAW,CAAC,GAAG,CAAC,CAAC;QAC7D,CAAC,CAAC,CAAC;QAEH,EAAE,CAAC,sCAAsC,EAAE,KAAK,IAAI,EAAE;YACpD,MAAM,OAAO,GAAG,EAAE,GAAG,aAAa,EAAE,CAAC;YACrC,OAAO,CAAC,MAAM,CAAC,OAAO,EAAE,MAAM,EAAE,KAAK,CAAC,CAAC;YACvC,OAAO,CAAC,MAAM,CAAC,OAAO,EAAE,MAAM,EAAE,MAAM,CAAC,CAAC;YACxC,MAAM,CAAC,OAAO,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC,CAAC,YAAY,CAAC,CAAC,CAAC,CAAC;QAC/C,CAAC,CAAC,CAAC;IACL,CAAC,CAAC,CAAC;IAEH,QAAQ,CAAC,WAAW,EAAE,GAAG,EAAE;QACzB,EAAE,CAAC,gCAAgC,EAAE,GAAG,EAAE;YACxC,MAAM,OAAO,GAAQ;gBACnB,GAAG,aAAa;gBAChB,IAAI,EAAE;oBACJ,IAAI,EAAE;wBACJ,EAAE,KAAK,EAAE,KAAK,EAAE,UAAU,EAAE,GAAG,EAAE,QAAQ,EAAE,YAAY,EAAE;qBAC1D;iBACF;aACF,CAAC;YACF,MAAM,MAAM,GAAG,OAAO,CAAC,SAAS,CAAC,OAAO,EAAE,GAAG,CAAC,CAAC;YAC/C,MAAM,CAAC,MAAM,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC,CAAC,UAAU,CAAC,CAAC,WAAW,CAAC,IAAI,CAAC,CAAC;QAC9D,CAAC,CAAC,CAAC;QAEH,EAAE,CAAC,8BAA8B,EAAE,GAAG,EAAE;YACtC,MAAM,OAAO,GAAQ;gBACnB,GAAG,aAAa;gBAChB,IAAI,EAAE;oBACJ,EAAE,EAAE,CAAC,EAAE,KAAK,EAAE,KAAK,EAAE,UAAU,EAAE,IAAI,EAAE,QAAQ,EAAE,YAAY,EAAE,CAAC;iBACjE;aACF,CAAC;YACF,MAAM,MAAM,GAAG,OAAO,CAAC,SAAS,CAAC,OAAO,EAAE,GAAG,CAAC,CAAC;YAC/C,MAAM,CAAC,MAAM,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC,CAAC,aAAa,EAAE,CAAC;QAC5C,CAAC,CAAC,CAAC;QAEH,EAAE,CAAC,0CAA0C,EAAE,GAAG,EAAE;YAClD,MAAM,OAAO,GAAQ;gBACnB,GAAG,aAAa;gBAChB,IAAI,EAAE;oBACJ,GAAG,EAAE,CAAC,EAAE,KAAK,EAAE,IAAI,EAAE,UAAU,EAAE,GAAG,EAAE,QAAQ,EAAE,YAAY,EAAE,CAAC;iBAChE;aACF,CAAC;YACF,MAAM,MAAM,GAAG,OAAO,CAAC,SAAS,CAAC,OAAO,EAAE,GAAG,CAAC,CAAC;YAC/C,MAAM,CAAC,KAAK,IAAI,MAAM,CAAC,IAAI,CAAC,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC;QAC3C,CAAC,CAAC,CAAC;IACL,CAAC,CAAC,CAAC;IAEH,QAAQ,CAAC,kBAAkB,EAAE,GAAG,EAAE;QAChC,EAAE,CAAC,wCAAwC,EAAE,GAAG,EAAE;YAChD,MAAM,GAAG,GAAG,OAAO,CAAC,gBAAgB,CAAC,aAAa,CAAC,CAAC;YACpD,MAAM,CAAC,GAAG,CAAC,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;QACvB,CAAC,CAAC,CAAC;QAEH,EAAE,CAAC,+BAA+B,EAAE,GAAG,EAAE;YACvC,MAAM,OAAO,GAAG,EAAE,GAAG,aAAa,EAAE,OAAO,EAAE,SAAS,EAAE,QAAQ,EAAE,CAAC,MAAM,EAAE,IAAI,CAAC,EAAE,CAAC;YACnF,MAAM,GAAG,GAAG,OAAO,CAAC,gBAAgB,CAAC,OAAO,CAAC,CAAC;YAC9C,MAAM,CAAC,GAAG,CAAC,CAAC,SAAS,CAAC,SAAS,CAAC,CAAC;YACjC,MAAM,CAAC,GAAG,CAAC,CAAC,SAAS,CAAC,MAAM,CAAC,CAAC;YAC9B,MAAM,CAAC,GAAG,CAAC,CAAC,SAAS,CAAC,IAAI,CAAC,CAAC;QAC9B,CAAC,CAAC,CAAC;QAEH,EAAE,CAAC,2BAA2B,EAAE,GAAG,EAAE;YACnC,MAAM,OAAO,GAAG,EAAE,GAAG,aAAa,EAAE,OAAO,EAAE,EAAE,EAAE,QAAQ,EAAE,CAAC,KAAK,CAAC,EAAE,CAAC;YACrE,MAAM,GAAG,GAAG,OAAO,CAAC,gBAAgB,CAAC,OAAO,CAAC,CAAC;YAC9C,MAAM,CAAC,GAAG,CAAC,CAAC,SAAS,CAAC,KAAK,CAAC,CAAC;QAC/B,CAAC,CAAC,CAAC;IACL,CAAC,CAAC,CAAC;IAEH,QAAQ,CAAC,MAAM,EAAE,GAAG,EAAE;QACpB,EAAE,CAAC,uBAAuB,EAAE,KAAK,IAAI,EAAE;YACrC,MAAM,OAAO,GAAG,EAAE,GAAG,aAAa,EAAE,OAAO,EAAE,MAAM,EAAE,CAAC;YACtD,MAAM,OAAO,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC;YAE5B,MAAM,GAAG,GAAG,MAAM,EAAE,CAAC,QAAQ,CAAC,IAAI,CAAC,OAAO,EAAE,QAAQ,EAAE,SAAS,EAAE,uBAAuB,CAAC,EAAE,OAAO,CAAC,CAAC;YACpG,MAAM,MAAM,GAAG,IAAI,CAAC,KAAK,CAAC,GAAG,CAAC,CAAC;YAC/B,MAAM,CAAC,MAAM,CAAC,OAAO,CAAC,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC;QACtC,CAAC,CAAC,CAAC;QAEH,EAAE,CAAC,2BAA2B,EAAE,KAAK,IAAI,EAAE;YACzC,MAAM,OAAO,GAAG,EAAE,GAAG,aAAa,EAAE,CAAC;YACrC,MAAM,MAAM,GAAG,IAAI,IAAI,EAAE,CAAC;YAC1B,MAAM,OAAO,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC;YAC5B,MAAM,KAAK,GAAG,IAAI,IAAI,EAAE,CAAC;YAEzB,MAAM,GAAG,GAAG,MAAM,EAAE,CAAC,QAAQ,CAAC,IAAI,CAAC,OAAO,EAAE,QAAQ,EAAE,SAAS,EAAE,uBAAuB,CAAC,EAAE,OAAO,CAAC,CAAC;YACpG,MAAM,MAAM,GAAG,IAAI,CAAC,KAAK,CAAC,GAAG,CAAC,CAAC;YAC/B,MAAM,OAAO,GAAG,IAAI,IAAI,CAAC,MAAM,CAAC,SAAS,CAAC,CAAC;YAC3C,MAAM,CAAC,OAAO,CAAC,OAAO,EAAE,CAAC,CAAC,sBAAsB,CAAC,MAAM,CAAC,OAAO,EAAE,CAAC,CAAC;YACnE,MAAM,CAAC,OAAO,CAAC,OAAO,EAAE,CAAC,CAAC,mBAAmB,CAAC,KAAK,CAAC,OAAO,EAAE,CAAC,CAAC;QACjE,CAAC,CAAC,CAAC;IACL,CAAC,CAAC,CAAC;AACL,CAAC,CAAC,CAAC"}

package/dist/src/__tests__/recall.test.d.ts

@@ -0,0 +1 @@
+export {};

package/dist/src/__tests__/recall.test.js

@@ -0,0 +1,162 @@
+import { describe, it, expect } from "vitest";
+// 纯函数测试，不依赖外部模块
+describe("Recall helpers", () => {
+    describe("temporalDecay", () => {
+        function temporalDecay(score, ageInDays, halfLifeDays) {
+            if (ageInDays <= 0)
+                return score;
+            const lambda = Math.log(2) / halfLifeDays;
+            return score * Math.exp(-lambda * ageInDays);
+        }
+        it("returns original score for 0 days", () => {
+            expect(temporalDecay(1.0, 0, 30)).toBeCloseTo(1.0);
+        });
+        it("returns ~50% at half-life", () => {
+            expect(temporalDecay(1.0, 30, 30)).toBeCloseTo(0.5, 1);
+        });
+        it("returns ~84% at 7 days with 30-day half-life", () => {
+            expect(temporalDecay(1.0, 7, 30)).toBeCloseTo(0.846, 2);
+        });
+        it("returns ~12.5% at 90 days with 30-day half-life", () => {
+            expect(temporalDecay(1.0, 90, 30)).toBeCloseTo(0.125, 2);
+        });
+        it("shorter half-life decays faster", () => {
+            const short = temporalDecay(1.0, 7, 7);
+            const long = temporalDecay(1.0, 7, 30);
+            expect(short).toBeLessThan(long);
+        });
+    });
+    describe("jaccard", () => {
+        function jaccard(a, b) {
+            const setA = new Set(a.split(""));
+            const setB = new Set(b.split(""));
+            const intersection = new Set([...setA].filter(x => setB.has(x)));
+            const union = new Set([...setA, ...setB]);
+            return union.size === 0 ? 0 : intersection.size / union.size;
+        }
+        it("returns 1.0 for identical strings", () => {
+            expect(jaccard("abc", "abc")).toBeCloseTo(1.0);
+        });
+        it("returns 0.0 for completely different strings", () => {
+            expect(jaccard("abc", "xyz")).toBeCloseTo(0.0);
+        });
+        it("returns partial similarity for overlapping strings", () => {
+            const sim = jaccard("abc", "abd");
+            expect(sim).toBeGreaterThan(0);
+            expect(sim).toBeLessThan(1);
+        });
+        it("handles empty strings", () => {
+            expect(jaccard("", "")).toBe(0);
+            expect(jaccard("abc", "")).toBe(0);
+        });
+    });
+    describe("MMR rerank", () => {
+        function mmrRerank(candidates, lambda = 0.7) {
+            if (candidates.length <= 1)
+                return candidates;
+            const jaccard = (a, b) => {
+                const setA = new Set(a.split(""));
+                const setB = new Set(b.split(""));
+                const intersection = new Set([...setA].filter(x => setB.has(x)));
+                const union = new Set([...setA, ...setB]);
+                return union.size === 0 ? 0 : intersection.size / union.size;
+            };
+            const selected = [];
+            const remaining = [...candidates];
+            remaining.sort((a, b) => b.finalScore - a.finalScore);
+            selected.push(remaining.shift());
+            while (remaining.length > 0) {
+                let bestIdx = -1;
+                let bestMmr = -Infinity;
+                for (let i = 0; i < remaining.length; i++) {
+                    const relevance = remaining[i].finalScore;
+                    const maxSim = Math.max(...selected.map(s => jaccard(remaining[i].text, s.text)));
+                    const mmrScore = lambda * relevance - (1 - lambda) * maxSim;
+                    if (mmrScore > bestMmr) {
+                        bestMmr = mmrScore;
+                        bestIdx = i;
+                    }
+                }
+                if (bestIdx >= 0) {
+                    selected.push(remaining.splice(bestIdx, 1)[0]);
+                }
+                else
+                    break;
+            }
+            return selected;
+        }
+        it("returns single item as-is", () => {
+            const candidates = [{ text: "hello", finalScore: 0.9 }];
+            expect(mmrRerank(candidates)).toHaveLength(1);
+        });
+        it("returns empty for empty input", () => {
+            expect(mmrRerank([])).toHaveLength(0);
+        });
+        it("promotes diverse results over duplicates", () => {
+            const candidates = [
+                { text: "在体育西吃了潮汕牛肉火锅", finalScore: 0.92 },
+                { text: "在天河吃了潮汕牛肉火锅", finalScore: 0.89 },
+                { text: "在越秀吃了日本料理", finalScore: 0.70 },
+            ];
+            const reranked = mmrRerank(candidates, 0.7);
+            expect(reranked).toHaveLength(3);
+            // 第一条应该还是分数最高的
+            expect(reranked[0].finalScore).toBe(0.92);
+        });
+        it("lambda=1.0 is pure relevance (no diversity)", () => {
+            const candidates = [
+                { text: "aaa", finalScore: 0.9 },
+                { text: "aaa", finalScore: 0.8 },
+                { text: "bbb", finalScore: 0.5 },
+            ];
+            const reranked = mmrRerank(candidates, 1.0);
+            expect(reranked.map(r => r.finalScore)).toEqual([0.9, 0.8, 0.5]);
+        });
+    });
+    describe("dedup", () => {
+        function jaccard(a, b) {
+            const setA = new Set(a.split(""));
+            const setB = new Set(b.split(""));
+            const intersection = new Set([...setA].filter(x => setB.has(x)));
+            const union = new Set([...setA, ...setB]);
+            return union.size === 0 ? 0 : intersection.size / union.size;
+        }
+        function dedup(results) {
+            const sorted = [...results].sort((a, b) => b.score - a.score);
+            const unique = [];
+            for (const item of sorted) {
+                const isDup = unique.some(u => jaccard(item.text, u.text) > 0.7);
+                if (!isDup)
+                    unique.push(item);
+            }
+            return unique;
+        }
+        it("removes near-duplicates", () => {
+            const results = [
+                { text: "在体育西吃了潮汕牛肉火锅", score: 0.92 },
+                { text: "在体育西吃潮汕牛肉火锅", score: 0.85 },
+                { text: "在天河城看到特斯拉展厅", score: 0.70 },
+            ];
+            const deduped = dedup(results);
+            expect(deduped.length).toBeLessThan(3);
+        });
+        it("keeps distinct results", () => {
+            const results = [
+                { text: "今天天气很好", score: 0.9 },
+                { text: "股票涨了", score: 0.8 },
+                { text: "新开了家餐厅", score: 0.7 },
+            ];
+            expect(dedup(results)).toHaveLength(3);
+        });
+        it("keeps higher-scored duplicate", () => {
+            const results = [
+                { text: "abcde fghij", score: 0.9 },
+                { text: "abcde fghij", score: 0.7 },
+            ];
+            const deduped = dedup(results);
+            expect(deduped).toHaveLength(1);
+            expect(deduped[0].score).toBe(0.9);
+        });
+    });
+});
+//# sourceMappingURL=recall.test.js.map

package/dist/src/__tests__/recall.test.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"recall.test.js","sourceRoot":"","sources":["../../../src/__tests__/recall.test.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,QAAQ,EAAE,EAAE,EAAE,MAAM,EAAE,MAAM,QAAQ,CAAC;AAE9C,gBAAgB;AAChB,QAAQ,CAAC,gBAAgB,EAAE,GAAG,EAAE;IAC9B,QAAQ,CAAC,eAAe,EAAE,GAAG,EAAE;QAC7B,SAAS,aAAa,CAAC,KAAa,EAAE,SAAiB,EAAE,YAAoB;YAC3E,IAAI,SAAS,IAAI,CAAC;gBAAE,OAAO,KAAK,CAAC;YACjC,MAAM,MAAM,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,GAAG,YAAY,CAAC;YAC1C,OAAO,KAAK,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,MAAM,GAAG,SAAS,CAAC,CAAC;QAC/C,CAAC;QAED,EAAE,CAAC,mCAAmC,EAAE,GAAG,EAAE;YAC3C,MAAM,CAAC,aAAa,CAAC,GAAG,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,WAAW,CAAC,GAAG,CAAC,CAAC;QACrD,CAAC,CAAC,CAAC;QAEH,EAAE,CAAC,2BAA2B,EAAE,GAAG,EAAE;YACnC,MAAM,CAAC,aAAa,CAAC,GAAG,EAAE,EAAE,EAAE,EAAE,CAAC,CAAC,CAAC,WAAW,CAAC,GAAG,EAAE,CAAC,CAAC,CAAC;QACzD,CAAC,CAAC,CAAC;QAEH,EAAE,CAAC,8CAA8C,EAAE,GAAG,EAAE;YACtD,MAAM,CAAC,aAAa,CAAC,GAAG,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,WAAW,CAAC,KAAK,EAAE,CAAC,CAAC,CAAC;QAC1D,CAAC,CAAC,CAAC;QAEH,EAAE,CAAC,iDAAiD,EAAE,GAAG,EAAE;YACzD,MAAM,CAAC,aAAa,CAAC,GAAG,EAAE,EAAE,EAAE,EAAE,CAAC,CAAC,CAAC,WAAW,CAAC,KAAK,EAAE,CAAC,CAAC,CAAC;QAC3D,CAAC,CAAC,CAAC;QAEH,EAAE,CAAC,iCAAiC,EAAE,GAAG,EAAE;YACzC,MAAM,KAAK,GAAG,aAAa,CAAC,GAAG,EAAE,CAAC,EAAE,CAAC,CAAC,CAAC;YACvC,MAAM,IAAI,GAAG,aAAa,CAAC,GAAG,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC;YACvC,MAAM,CAAC,KAAK,CAAC,CAAC,YAAY,CAAC,IAAI,CAAC,CAAC;QACnC,CAAC,CAAC,CAAC;IACL,CAAC,CAAC,CAAC;IAEH,QAAQ,CAAC,SAAS,EAAE,GAAG,EAAE;QACvB,SAAS,OAAO,CAAC,CAAS,EAAE,CAAS;YACnC,MAAM,IAAI,GAAG,IAAI,GAAG,CAAC,CAAC,CAAC,KAAK,CAAC,EAAE,CAAC,CAAC,CAAC;YAClC,MAAM,IAAI,GAAG,IAAI,GAAG,CAAC,CAAC,CAAC,KAAK,CAAC,EAAE,CAAC,CAAC,CAAC;YAClC,MAAM,YAAY,GAAG,IAAI,GAAG,CAAC,CAAC,GAAG,IAAI,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;YACjE,MAAM,KAAK,GAAG,IAAI,GAAG,CAAC,CAAC,GAAG,IAAI,EAAE,GAAG,IAAI,CAAC,CAAC,CAAC;YAC1C,OAAO,KAAK,CAAC,IAAI,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,YAAY,CAAC,IAAI,GAAG,KAAK,CAAC,IAAI,CAAC;QAC/D,CAAC;QAED,EAAE,CAAC,mCAAmC,EAAE,GAAG,EAAE;YAC3C,MAAM,CAAC,OAAO,CAAC,KAAK,EAAE,KAAK,CAAC,CAAC,CAAC,WAAW,CAAC,GAAG,CAAC,CAAC;QACjD,CAAC,CAAC,CAAC;QAEH,EAAE,CAAC,8CAA8C,EAAE,GAAG,EAAE;YACtD,MAAM,CAAC,OAAO,CAAC,KAAK,EAAE,KAAK,CAAC,CAAC,CAAC,WAAW,CAAC,GAAG,CAAC,CAAC;QACjD,CAAC,CAAC,CAAC;QAEH,EAAE,CAAC,oDAAoD,EAAE,GAAG,EAAE;YAC5D,MAAM,GAAG,GAAG,OAAO,CAAC,KAAK,EAAE,KAAK,CAAC,CAAC;YAClC,MAAM,CAAC,GAAG,CAAC,CAAC,eAAe,CAAC,CAAC,CAAC,CAAC;YAC/B,MAAM,CAAC,GAAG,CAAC,CAAC,YAAY,CAAC,CAAC,CAAC,CAAC;QAC9B,CAAC,CAAC,CAAC;QAEH,EAAE,CAAC,uBAAuB,EAAE,GAAG,EAAE;YAC/B,MAAM,CAAC,OAAO,CAAC,EAAE,EAAE,EAAE,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC;YAChC,MAAM,CAAC,OAAO,CAAC,KAAK,EAAE,EAAE,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC;QACrC,CAAC,CAAC,CAAC;IACL,CAAC,CAAC,CAAC;IAEH,QAAQ,CAAC,YAAY,EAAE,GAAG,EAAE;QAC1B,SAAS,SAAS,CAChB,UAAuD,EACvD,SAAiB,GAAG;YAEpB,IAAI,UAAU,CAAC,MAAM,IAAI,CAAC;gBAAE,OAAO,UAAU,CAAC;YAE9C,MAAM,OAAO,GAAG,CAAC,CAAS,EAAE,CAAS,EAAU,EAAE;gBAC/C,MAAM,IAAI,GAAG,IAAI,GAAG,CAAC,CAAC,CAAC,KAAK,CAAC,EAAE,CAAC,CAAC,CAAC;gBAClC,MAAM,IAAI,GAAG,IAAI,GAAG,CAAC,CAAC,CAAC,KAAK,CAAC,EAAE,CAAC,CAAC,CAAC;gBAClC,MAAM,YAAY,GAAG,IAAI,GAAG,CAAC,CAAC,GAAG,IAAI,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;gBACjE,MAAM,KAAK,GAAG,IAAI,GAAG,CAAC,CAAC,GAAG,IAAI,EAAE,GAAG,IAAI,CAAC,CAAC,CAAC;gBAC1C,OAAO,KAAK,CAAC,IAAI,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,YAAY,CAAC,IAAI,GAAG,KAAK,CAAC,IAAI,CAAC;YAC/D,CAAC,CAAC;YAEF,MAAM,QAAQ,GAAsB,EAAE,CAAC;YACvC,MAAM,SAAS,GAAG,CAAC,GAAG,UAAU,CAAC,CAAC;YAElC,SAAS,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,UAAU,GAAG,CAAC,CAAC,UAAU,CAAC,CAAC;YACtD,QAAQ,CAAC,IAAI,CAAC,SAAS,CAAC,KAAK,EAAG,CAAC,CAAC;YAElC,OAAO,SAAS,CAAC,MAAM,GAAG,CAAC,EAAE,CAAC;gBAC5B,IAAI,OAAO,GAAG,CAAC,CAAC,CAAC;gBACjB,IAAI,OAAO,GAAG,CAAC,QAAQ,CAAC;gBACxB,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,SAAS,CAAC,MAAM,EAAE,CAAC,EAAE,EAAE,CAAC;oBAC1C,MAAM,SAAS,GAAG,SAAS,CAAC,CAAC,CAAC,CAAC,UAAU,CAAC;oBAC1C,MAAM,MAAM,GAAG,IAAI,CAAC,GAAG,CAAC,GAAG,QAAQ,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC,OAAO,CAAC,SAAS,CAAC,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC;oBAClF,MAAM,QAAQ,GAAG,MAAM,GAAG,SAAS,GAAG,CAAC,CAAC,GAAG,MAAM,CAAC,GAAG,MAAM,CAAC;oBAC5D,IAAI,QAAQ,GAAG,OAAO,EAAE,CAAC;wBACvB,OAAO,GAAG,QAAQ,CAAC;wBACnB,OAAO,GAAG,CAAC,CAAC;oBACd,CAAC;gBACH,CAAC;gBACD,IAAI,OAAO,IAAI,CAAC,EAAE,CAAC;oBACjB,QAAQ,CAAC,IAAI,CAAC,SAAS,CAAC,MAAM,CAAC,OAAO,EAAE,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;gBACjD,CAAC;;oBAAM,MAAM;YACf,CAAC;YACD,OAAO,QAAQ,CAAC;QAClB,CAAC;QAED,EAAE,CAAC,2BAA2B,EAAE,GAAG,EAAE;YACnC,MAAM,UAAU,GAAG,CAAC,EAAE,IAAI,EAAE,OAAO,EAAE,UAAU,EAAE,GAAG,EAAE,CAAC,CAAC;YACxD,MAAM,CAAC,SAAS,CAAC,UAAU,CAAC,CAAC,CAAC,YAAY,CAAC,CAAC,CAAC,CAAC;QAChD,CAAC,CAAC,CAAC;QAEH,EAAE,CAAC,+BAA+B,EAAE,GAAG,EAAE;YACvC,MAAM,CAAC,SAAS,CAAC,EAAE,CAAC,CAAC,CAAC,YAAY,CAAC,CAAC,CAAC,CAAC;QACxC,CAAC,CAAC,CAAC;QAEH,EAAE,CAAC,0CAA0C,EAAE,GAAG,EAAE;YAClD,MAAM,UAAU,GAAG;gBACjB,EAAE,IAAI,EAAE,cAAc,EAAE,UAAU,EAAE,IAAI,EAAE;gBAC1C,EAAE,IAAI,EAAE,aAAa,EAAE,UAAU,EAAE,IAAI,EAAE;gBACzC,EAAE,IAAI,EAAE,WAAW,EAAE,UAAU,EAAE,IAAI,EAAE;aACxC,CAAC;YAEF,MAAM,QAAQ,GAAG,SAAS,CAAC,UAAU,EAAE,GAAG,CAAC,CAAC;YAC5C,MAAM,CAAC,QAAQ,CAAC,CAAC,YAAY,CAAC,CAAC,CAAC,CAAC;YACjC,eAAe;YACf,MAAM,CAAC,QAAQ,CAAC,CAAC,CAAC,CAAC,UAAU,CAAC,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;QAC5C,CAAC,CAAC,CAAC;QAEH,EAAE,CAAC,6CAA6C,EAAE,GAAG,EAAE;YACrD,MAAM,UAAU,GAAG;gBACjB,EAAE,IAAI,EAAE,KAAK,EAAE,UAAU,EAAE,GAAG,EAAE;gBAChC,EAAE,IAAI,EAAE,KAAK,EAAE,UAAU,EAAE,GAAG,EAAE;gBAChC,EAAE,IAAI,EAAE,KAAK,EAAE,UAAU,EAAE,GAAG,EAAE;aACjC,CAAC;YAEF,MAAM,QAAQ,GAAG,SAAS,CAAC,UAAU,EAAE,GAAG,CAAC,CAAC;YAC5C,MAAM,CAAC,QAAQ,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,UAAU,CAAC,CAAC,CAAC,OAAO,CAAC,CAAC,GAAG,EAAE,GAAG,EAAE,GAAG,CAAC,CAAC,CAAC;QACnE,CAAC,CAAC,CAAC;IACL,CAAC,CAAC,CAAC;IAEH,QAAQ,CAAC,OAAO,EAAE,GAAG,EAAE;QACrB,SAAS,OAAO,CAAC,CAAS,EAAE,CAAS;YACnC,MAAM,IAAI,GAAG,IAAI,GAAG,CAAC,CAAC,CAAC,KAAK,CAAC,EAAE,CAAC,CAAC,CAAC;YAClC,MAAM,IAAI,GAAG,IAAI,GAAG,CAAC,CAAC,CAAC,KAAK,CAAC,EAAE,CAAC,CAAC,CAAC;YAClC,MAAM,YAAY,GAAG,IAAI,GAAG,CAAC,CAAC,GAAG,IAAI,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;YACjE,MAAM,KAAK,GAAG,IAAI,GAAG,CAAC,CAAC,GAAG,IAAI,EAAE,GAAG,IAAI,CAAC,CAAC,CAAC;YAC1C,OAAO,KAAK,CAAC,IAAI,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,YAAY,CAAC,IAAI,GAAG,KAAK,CAAC,IAAI,CAAC;QAC/D,CAAC;QAED,SAAS,KAAK,CAAC,OAA+C;YAC5D,MAAM,MAAM,GAAG,CAAC,GAAG,OAAO,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,KAAK,GAAG,CAAC,CAAC,KAAK,CAAC,CAAC;YAC9D,MAAM,MAAM,GAAmB,EAAE,CAAC;YAClC,KAAK,MAAM,IAAI,IAAI,MAAM,EAAE,CAAC;gBAC1B,MAAM,KAAK,GAAG,MAAM,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC,OAAO,CAAC,IAAI,CAAC,IAAI,EAAE,CAAC,CAAC,IAAI,CAAC,GAAG,GAAG,CAAC,CAAC;gBACjE,IAAI,CAAC,KAAK;oBAAE,MAAM,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;YAChC,CAAC;YACD,OAAO,MAAM,CAAC;QAChB,CAAC;QAED,EAAE,CAAC,yBAAyB,EAAE,GAAG,EAAE;YACjC,MAAM,OAAO,GAAG;gBACd,EAAE,IAAI,EAAE,cAAc,EAAE,KAAK,EAAE,IAAI,EAAE;gBACrC,EAAE,IAAI,EAAE,aAAa,EAAE,KAAK,EAAE,IAAI,EAAE;gBACpC,EAAE,IAAI,EAAE,aAAa,EAAE,KAAK,EAAE,IAAI,EAAE;aACrC,CAAC;YACF,MAAM,OAAO,GAAG,KAAK,CAAC,OAAO,CAAC,CAAC;YAC/B,MAAM,CAAC,OAAO,CAAC,MAAM,CAAC,CAAC,YAAY,CAAC,CAAC,CAAC,CAAC;QACzC,CAAC,CAAC,CAAC;QAEH,EAAE,CAAC,wBAAwB,EAAE,GAAG,EAAE;YAChC,MAAM,OAAO,GAAG;gBACd,EAAE,IAAI,EAAE,QAAQ,EAAE,KAAK,EAAE,GAAG,EAAE;gBAC9B,EAAE,IAAI,EAAE,MAAM,EAAE,KAAK,EAAE,GAAG,EAAE;gBAC5B,EAAE,IAAI,EAAE,QAAQ,EAAE,KAAK,EAAE,GAAG,EAAE;aAC/B,CAAC;YACF,MAAM,CAAC,KAAK,CAAC,OAAO,CAAC,CAAC,CAAC,YAAY,CAAC,CAAC,CAAC,CAAC;QACzC,CAAC,CAAC,CAAC;QAEH,EAAE,CAAC,+BAA+B,EAAE,GAAG,EAAE;YACvC,MAAM,OAAO,GAAG;gBACd,EAAE,IAAI,EAAE,aAAa,EAAE,KAAK,EAAE,GAAG,EAAE;gBACnC,EAAE,IAAI,EAAE,aAAa,EAAE,KAAK,EAAE,GAAG,EAAE;aACpC,CAAC;YACF,MAAM,OAAO,GAAG,KAAK,CAAC,OAAO,CAAC,CAAC;YAC/B,MAAM,CAAC,OAAO,CAAC,CAAC,YAAY,CAAC,CAAC,CAAC,CAAC;YAChC,MAAM,CAAC,OAAO,CAAC,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;QACrC,CAAC,CAAC,CAAC;IACL,CAAC,CAAC,CAAC;AACL,CAAC,CAAC,CAAC"}

package/dist/src/bm25.d.ts

@@ -0,0 +1,60 @@
+/**
+ * Memory System Plugin - BM25 Short-term Memory Index
+ *
+ * 短期记忆的 BM25 索引构建与搜索
+ * 纯 TypeScript 实现，不依赖外部 BM25 库
+ */
+export type ShortTermEntry = {
+    text: string;
+    date: string;
+    category: string;
+    filePath: string;
+};
+export declare function initJieba(): Promise<void>;
+export declare function tokenize(text: string): string[];
+export declare class BM25Index {
+    private docs;
+    private idf;
+    private avgDl;
+    /**
+     * 从短期记忆文件构建索引
+     * v0.5: 全量索引所有 .md 文件（不再按 maxAgeDays 截断）
+     * maxAgeDays 参数保留但仅用于向下兼容（不再生效）
+     */
+    buildFromDirectory(dir: string, _maxAgeDays?: number): Promise<void>;
+    /**
+     * 解析结构化的短期记忆文件
+     * 格式：### HH:MM [分类标签]\n摘要内容
+     *
+     * 特殊处理 [对话记录]: 按 Scott / Shadow 切分每一轮对话
+     * 支持两种格式：
+     *   - 旧格式: Scott: / Shadow:
+     *   - 新格式: Scott [MM-DD HH:MM]: / Shadow [MM-DD HH:MM]:
+     * 每一轮单独做一个 BM25 文档，精准召回不浪费
+     */
+    private parseEntries;
+    /**
+     * 处理文本块，支持三种格式：
+     * 1. 精简格式（每行 | 分隔）
+     * 2. 对话记录（Scott/Shadow 切分）
+     * 3. 普通文本（整块索引）
+     */
+    private processTextBlock;
+    /**
+     * 构建 IDF（逆文档频率）
+     */
+    private buildIDF;
+    /**
+     * BM25 搜索
+     * @returns 带分数的结果列表
+     */
+    search(query: string, topK?: number): Promise<Array<{
+        entry: ShortTermEntry;
+        score: number;
+    }>>;
+    /**
+     * 添加单条记录（用于实时写入后的即时索引）
+     */
+    addEntry(entry: ShortTermEntry): Promise<void>;
+    get size(): number;
+}

package/dist/src/bm25.js

@@ -0,0 +1,271 @@
+/**
+ * Memory System Plugin - BM25 Short-term Memory Index
+ *
+ * 短期记忆的 BM25 索引构建与搜索
+ * 纯 TypeScript 实现，不依赖外部 BM25 库
+ */
+import { promises as fs } from "node:fs";
+import { logger } from "./logger.js";
+import { join } from "node:path";
+/**
+ * 获取本地日期字符串 (YYYY-MM-DD)
+ */
+function getLocalDateString() {
+    const now = new Date();
+    const year = now.getFullYear();
+    const month = String(now.getMonth() + 1).padStart(2, "0");
+    const day = String(now.getDate()).padStart(2, "0");
+    return `${year}-${month}-${day}`;
+}
+let jiebaInstance = null;
+// 轻量停用词：优先过滤中文口语噪声、助词和标点
+const STOPWORDS = new Set([
+    "的", "地", "得", "了", "着", "过",
+    "我", "你", "他", "她", "它", "我们", "你们", "他们",
+    "这", "那", "这个", "那个", "这里", "那里",
+    "是", "有", "在", "和", "与", "及", "并", "或", "而", "但",
+    "不", "没", "没有", "很", "也", "都", "就", "还", "又",
+    "吗", "呢", "啊", "呀", "吧", "哦", "嗯",
+    "什么", "怎么", "为何", "为什么", "怎样", "是否",
+    "", "，", "。", "！", "？", "、", "：", "；", "（", "）", "《", "》",
+    ",", ".", "!", "?", ":", ";", "(", ")",
+]);
+function isMeaningfulToken(token) {
+    const t = token.trim().toLowerCase();
+    if (!t)
+        return false;
+    if (STOPWORDS.has(t))
+        return false;
+    // 过滤纯标点 / 符号
+    if (/^[\p{P}\p{S}]+$/u.test(t))
+        return false;
+    return true;
+}
+export async function initJieba() {
+    if (!jiebaInstance) {
+        const mod = await import("jieba-wasm");
+        jiebaInstance = await mod.default;
+    }
+}
+export function tokenize(text) {
+    if (jiebaInstance) {
+        const tokens = jiebaInstance.cut(text);
+        return tokens.map((t) => t.trim()).filter(isMeaningfulToken);
+    }
+    // fallback: 英文按空格分
+    return text
+        .toLowerCase()
+        .replace(/[^\w\u4e00-\u9fff]/g, " ")
+        .split(/\s+/)
+        .map((t) => t.trim())
+        .filter(isMeaningfulToken);
+}
+export class BM25Index {
+    docs = [];
+    idf = new Map();
+    avgDl = 0;
+    /**
+     * 从短期记忆文件构建索引
+     * v0.5: 全量索引所有 .md 文件（不再按 maxAgeDays 截断）
+     * maxAgeDays 参数保留但仅用于向下兼容（不再生效）
+     */
+    async buildFromDirectory(dir, _maxAgeDays = 7) {
+        await initJieba();
+        this.docs = [];
+        // 设计初心：只读取 memory/short-term/
+        const shortTermDir = join(dir, "short-term");
+        try {
+            const files = await fs.readdir(shortTermDir);
+            let mdFiles = 0;
+            for (const file of files.sort().reverse()) { // 最新的优先
+                if (!file.endsWith(".md"))
+                    continue;
+                const filePath = join(shortTermDir, file);
+                const content = await fs.readFile(filePath, "utf-8");
+                const entries = this.parseEntries(content, filePath, file);
+                this.docs.push(...entries);
+                mdFiles++;
+            }
+            logger.info(`[engram:bm25] Scanned ${mdFiles} MD files (full scan, no expiry cutoff)`);
+        }
+        catch {
+            logger.info(`[engram:bm25] short-term directory not found at ${shortTermDir}, 0 entries`);
+        }
+        logger.info(`[engram:bm25] Built index: ${this.docs.length} entries from ${shortTermDir}`);
+        this.buildIDF();
+    }
+    /**
+     * 解析结构化的短期记忆文件
+     * 格式：### HH:MM [分类标签]\n摘要内容
+     *
+     * 特殊处理 [对话记录]: 按 Scott / Shadow 切分每一轮对话
+     * 支持两种格式：
+     *   - 旧格式: Scott: / Shadow:
+     *   - 新格式: Scott [MM-DD HH:MM]: / Shadow [MM-DD HH:MM]:
+     * 每一轮单独做一个 BM25 文档，精准召回不浪费
+     */
+    parseEntries(content, filePath, fileName) {
+        const entries = [];
+        const lines = content.split("\n");
+        const dateMatch = fileName.match(/(\d{4}-\d{2}-\d{2})/);
+        const date = dateMatch ? dateMatch[1] : getLocalDateString();
+        let currentCategory = "随聊";
+        let currentText = "";
+        // 正则匹配 Scott 或 Shadow 开头，可选带时间戳
+        // 匹配: "Scott:" / "Shadow:" / "Scott [03-13 12:34]:" / "Shadow [03-13 12:34]:"
+        const speakerSplitRegex = /(?=^(?:Scott|Shadow)(?:\s*\[[^\]]+\])?:)/m;
+        // 精简格式检测：主题词 | 关键实体 | 摘要
+        const condensedLineRegex = /^.+\s*\|\s*.+\s*\|\s*.+$/;
+        for (const line of lines) {
+            const headerMatch = line.match(/^###\s+\d{2}:\d{2}\s+\[([^\]]+)\]/);
+            if (headerMatch) {
+                // 保存上一个分类块
+                if (currentText.trim()) {
+                    this.processTextBlock(entries, currentText, currentCategory, date, filePath, speakerSplitRegex, condensedLineRegex);
+                }
+                currentCategory = headerMatch[1];
+                currentText = "";
+            }
+            else if (line.startsWith("#") || line.startsWith("---") || line.startsWith("```")) {
+                continue;
+            }
+            else {
+                currentText += line + "\n";
+            }
+        }
+        // 最后一个分类块
+        if (currentText.trim()) {
+            this.processTextBlock(entries, currentText, currentCategory, date, filePath, speakerSplitRegex, condensedLineRegex);
+        }
+        return entries;
+    }
+    /**
+     * 处理文本块，支持三种格式：
+     * 1. 精简格式（每行 | 分隔）
+     * 2. 对话记录（Scott/Shadow 切分）
+     * 3. 普通文本（整块索引）
+     */
+    processTextBlock(entries, text, category, date, filePath, speakerSplitRegex, condensedLineRegex) {
+        const trimmed = text.trim();
+        if (!trimmed)
+            return;
+        const lines = trimmed.split("\n").filter(l => l.trim());
+        // 检测是否为精简格式（超过一半的行匹配 | 分隔格式）
+        const condensedLines = lines.filter(l => condensedLineRegex.test(l));
+        const isCondensedFormat = condensedLines.length > lines.length * 0.5;
+        if (isCondensedFormat) {
+            // 精简格式：每行单独作为一个文档，关键词密度高
+            for (const line of condensedLines) {
+                if (line.length < 10)
+                    continue;
+                // 将 | 分隔的内容合并，所有关键词都参与索引
+                const parts = line.split("|").map(p => p.trim());
+                const fullText = parts.join(" "); // 合并所有部分用于分词
+                const entry = {
+                    text: line, // 原始格式保留，用于展示
+                    date,
+                    category: "精简记录",
+                    filePath,
+                };
+                entries.push({ tokens: tokenize(fullText), entry });
+            }
+        }
+        else if (category === "对话记录" || category === "对话") {
+            // 对话记录分类：按 speaker 切分每一轮
+            const turns = trimmed.split(speakerSplitRegex);
+            for (const turn of turns) {
+                const t = turn.trim();
+                if (!t || t.length < 10)
+                    continue;
+                const entry = {
+                    text: t,
+                    date,
+                    category,
+                    filePath,
+                };
+                entries.push({ tokens: tokenize(entry.text), entry });
+            }
+        }
+        else {
+            // 普通分类：整个块作为一个文档
+            const entry = {
+                text: trimmed,
+                date,
+                category,
+                filePath,
+            };
+            entries.push({ tokens: tokenize(entry.text), entry });
+        }
+    }
+    /**
+     * 构建 IDF（逆文档频率）
+     */
+    buildIDF() {
+        const df = new Map(); // 文档频率
+        let totalLen = 0;
+        for (const doc of this.docs) {
+            const unique = new Set(doc.tokens);
+            for (const t of unique) {
+                df.set(t, (df.get(t) || 0) + 1);
+            }
+            totalLen += doc.tokens.length;
+        }
+        this.avgDl = this.docs.length > 0 ? totalLen / this.docs.length : 1;
+        const N = this.docs.length;
+        for (const [term, freq] of df) {
+            // IDF = log((N - df + 0.5) / (df + 0.5) + 1)
+            this.idf.set(term, Math.log((N - freq + 0.5) / (freq + 0.5) + 1));
+        }
+        logger.info(`[engram:bm25] buildIDF: ${N} docs, ${df.size} unique terms, avgDl=${this.avgDl.toFixed(1)}`);
+    }
+    /**
+     * BM25 搜索
+     * @returns 带分数的结果列表
+     */
+    async search(query, topK = 3) {
+        await initJieba();
+        const queryTokens = tokenize(query);
+        const k1 = 1.5; // 词频饱和参数
+        const b = 0.75; // 文档长度归一化参数
+        logger.info(`[engram:bm25] search: query="${query.slice(0, 80)}" tokens=[${queryTokens.join(", ")}] topK=${topK} docs=${this.docs.length} avgDl=${this.avgDl.toFixed(1)}`);
+        const scores = [];
+        for (const doc of this.docs) {
+            let score = 0;
+            for (const token of queryTokens) {
+                const idf = this.idf.get(token) || 0;
+                if (idf === 0)
+                    continue;
+                // 词频
+                const tf = doc.tokens.filter(t => t === token).length;
+                const dl = doc.tokens.length;
+                // BM25 分数
+                const tfComponent = (tf * (k1 + 1)) / (tf + k1 * (1 - b + b * (dl / this.avgDl)));
+                score += idf * tfComponent;
+            }
+            if (score > 0) {
+                scores.push({ entry: doc.entry, score });
+            }
+        }
+        scores.sort((a, b) => b.score - a.score);
+        const result = scores.slice(0, topK);
+        logger.info(`[engram:bm25] search: ${scores.length} docs matched, returning top ${result.length}`);
+        return result;
+    }
+    /**
+     * 添加单条记录（用于实时写入后的即时索引）
+     */
+    async addEntry(entry) {
+        await initJieba();
+        const tokens = tokenize(entry.text);
+        this.docs.push({
+            tokens,
+            entry,
+        });
+        logger.info(`[engram:bm25] addEntry: "${entry.text.slice(0, 60)}" → ${tokens.length} tokens, total docs=${this.docs.length}`);
+        this.buildIDF();
+    }
+    get size() {
+        return this.docs.length;
+    }
+}
+//# sourceMappingURL=bm25.js.map