@chainlesschain/personal-data-hub 0.4.29 → 0.4.31

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (199) hide show
  1. package/lib/forensics/qq-nt-collect.js +190 -0
  2. package/lib/prompt-builder.js +15 -1
  3. package/package.json +8 -3
  4. package/__tests__/adapter-guide.test.js +0 -47
  5. package/__tests__/adapter-spec.test.js +0 -78
  6. package/__tests__/adapters/ai-chat-cookie-capture-spec.test.js +0 -211
  7. package/__tests__/adapters/ai-chat-health-checker.test.js +0 -262
  8. package/__tests__/adapters/ai-chat-history.test.js +0 -396
  9. package/__tests__/adapters/ai-chat-http-client.test.js +0 -242
  10. package/__tests__/adapters/ai-chat-vendors.test.js +0 -874
  11. package/__tests__/adapters/alipay-bill-adapter.test.js +0 -538
  12. package/__tests__/adapters/apple-health.test.js +0 -95
  13. package/__tests__/adapters/bank-family.test.js +0 -125
  14. package/__tests__/adapters/biz-tianyancha.test.js +0 -159
  15. package/__tests__/adapters/browser-history-chrome.test.js +0 -377
  16. package/__tests__/adapters/browser-history-edge.test.js +0 -159
  17. package/__tests__/adapters/car-mercedesme.test.js +0 -74
  18. package/__tests__/adapters/doc-baidu-netdisk.test.js +0 -102
  19. package/__tests__/adapters/doc-camscanner.test.js +0 -147
  20. package/__tests__/adapters/doc-platforms.test.js +0 -177
  21. package/__tests__/adapters/edu-huawei-learning-live.test.js +0 -198
  22. package/__tests__/adapters/edu-zuoyebang-live.test.js +0 -226
  23. package/__tests__/adapters/email-adapter-snapshot.test.js +0 -237
  24. package/__tests__/adapters/email-adapter.test.js +0 -742
  25. package/__tests__/adapters/email-classifier.test.js +0 -347
  26. package/__tests__/adapters/email-imap-session.test.js +0 -334
  27. package/__tests__/adapters/email-parser.test.js +0 -244
  28. package/__tests__/adapters/email-pdf-extractor.test.js +0 -529
  29. package/__tests__/adapters/email-providers.test.js +0 -84
  30. package/__tests__/adapters/email-retry-progress.test.js +0 -294
  31. package/__tests__/adapters/email-templates.test.js +0 -822
  32. package/__tests__/adapters/family-23-collectors-scaffold.test.js +0 -182
  33. package/__tests__/adapters/finance-alipay-live.test.js +0 -258
  34. package/__tests__/adapters/finance-dcep.test.js +0 -74
  35. package/__tests__/adapters/fitness-joyrun.test.js +0 -82
  36. package/__tests__/adapters/game-genshin-live.test.js +0 -238
  37. package/__tests__/adapters/game-genshin-scaffold.test.js +0 -108
  38. package/__tests__/adapters/game-honor-of-kings-live.test.js +0 -230
  39. package/__tests__/adapters/git-activity.test.js +0 -222
  40. package/__tests__/adapters/gov-12123.test.js +0 -103
  41. package/__tests__/adapters/gov-ixiamen.test.js +0 -150
  42. package/__tests__/adapters/gov-tax.test.js +0 -135
  43. package/__tests__/adapters/health-meiyou.test.js +0 -125
  44. package/__tests__/adapters/local-files.test.js +0 -264
  45. package/__tests__/adapters/local-im-pc.test.js +0 -154
  46. package/__tests__/adapters/messaging-whatsapp.test.js +0 -289
  47. package/__tests__/adapters/music-kugou.test.js +0 -187
  48. package/__tests__/adapters/music-qq.test.js +0 -112
  49. package/__tests__/adapters/netease-music-live.test.js +0 -244
  50. package/__tests__/adapters/netease-music.test.js +0 -74
  51. package/__tests__/adapters/pc-local-discovery.test.js +0 -141
  52. package/__tests__/adapters/qq-pc-direct-read.test.js +0 -227
  53. package/__tests__/adapters/reading-family.test.js +0 -108
  54. package/__tests__/adapters/recruit-boss.test.js +0 -180
  55. package/__tests__/adapters/shell-history.test.js +0 -180
  56. package/__tests__/adapters/shopping-base.test.js +0 -179
  57. package/__tests__/adapters/shopping-dianping.test.js +0 -239
  58. package/__tests__/adapters/social-bilibili-adb-api-client.test.js +0 -721
  59. package/__tests__/adapters/social-bilibili-adb-chromium-cookies-reader.test.js +0 -346
  60. package/__tests__/adapters/social-bilibili-adb-collector.test.js +0 -284
  61. package/__tests__/adapters/social-bilibili-adb-cookies-extension.test.js +0 -343
  62. package/__tests__/adapters/social-bilibili-adb-snapshot-builder.test.js +0 -296
  63. package/__tests__/adapters/social-csdn.test.js +0 -175
  64. package/__tests__/adapters/social-dongchedi.test.js +0 -165
  65. package/__tests__/adapters/social-douyin-adb-aweme-detail.test.js +0 -165
  66. package/__tests__/adapters/social-douyin-adb-collector.test.js +0 -254
  67. package/__tests__/adapters/social-douyin-adb-db-extension.test.js +0 -114
  68. package/__tests__/adapters/social-douyin-adb-im-db-parser.test.js +0 -304
  69. package/__tests__/adapters/social-douyin-adb-snapshot-builder.test.js +0 -216
  70. package/__tests__/adapters/social-douyin-adb-usage-profile.test.js +0 -229
  71. package/__tests__/adapters/social-douyin-adb-watch-history.test.js +0 -269
  72. package/__tests__/adapters/social-kuaishou-adb-api-client.test.js +0 -496
  73. package/__tests__/adapters/social-kuaishou-adb-collector.test.js +0 -276
  74. package/__tests__/adapters/social-kuaishou-adb-cookies-extension.test.js +0 -152
  75. package/__tests__/adapters/social-kuaishou-adb-snapshot-builder.test.js +0 -178
  76. package/__tests__/adapters/social-toutiao-adb-account-reader.test.js +0 -135
  77. package/__tests__/adapters/social-toutiao-adb-api-client.test.js +0 -626
  78. package/__tests__/adapters/social-toutiao-adb-article.test.js +0 -155
  79. package/__tests__/adapters/social-toutiao-adb-collector.test.js +0 -378
  80. package/__tests__/adapters/social-toutiao-adb-cookies-extension.test.js +0 -193
  81. package/__tests__/adapters/social-toutiao-adb-snapshot-builder.test.js +0 -196
  82. package/__tests__/adapters/social-toutiao-kuaishou-scaffold.test.js +0 -311
  83. package/__tests__/adapters/social-weibo-adb-api-client.test.js +0 -362
  84. package/__tests__/adapters/social-weibo-adb-collector.test.js +0 -201
  85. package/__tests__/adapters/social-weibo-adb-cookies-extension.test.js +0 -167
  86. package/__tests__/adapters/social-weibo-adb-snapshot-builder.test.js +0 -189
  87. package/__tests__/adapters/social-xiaohongshu-adb-api-client.test.js +0 -431
  88. package/__tests__/adapters/social-xiaohongshu-adb-collector.test.js +0 -207
  89. package/__tests__/adapters/social-xiaohongshu-adb-cookies-extension.test.js +0 -0
  90. package/__tests__/adapters/social-xiaohongshu-adb-sign-provider-injection.test.js +0 -351
  91. package/__tests__/adapters/social-xiaohongshu-adb-sign.test.js +0 -130
  92. package/__tests__/adapters/social-xiaohongshu-adb-snapshot-builder.test.js +0 -200
  93. package/__tests__/adapters/social-zhihu.test.js +0 -246
  94. package/__tests__/adapters/system-data-adapter.test.js +0 -443
  95. package/__tests__/adapters/system-data-android-ingest.test.js +0 -144
  96. package/__tests__/adapters/system-data-android.test.js +0 -519
  97. package/__tests__/adapters/system-data-disclosure.test.js +0 -153
  98. package/__tests__/adapters/travel-12306.test.js +0 -512
  99. package/__tests__/adapters/travel-amap.test.js +0 -219
  100. package/__tests__/adapters/travel-baidu-map.test.js +0 -305
  101. package/__tests__/adapters/travel-base.test.js +0 -205
  102. package/__tests__/adapters/travel-ctrip.test.js +0 -377
  103. package/__tests__/adapters/travel-didi-consumer.test.js +0 -66
  104. package/__tests__/adapters/travel-didi.test.js +0 -204
  105. package/__tests__/adapters/travel-tencent-map.test.js +0 -207
  106. package/__tests__/adapters/travel-tongcheng.test.js +0 -289
  107. package/__tests__/adapters/video-platforms.test.js +0 -152
  108. package/__tests__/adapters/video-xigua.test.js +0 -106
  109. package/__tests__/adapters/vscode.test.js +0 -299
  110. package/__tests__/adapters/wechat-bootstrap.test.js +0 -240
  111. package/__tests__/adapters/wechat-env-probe.test.js +0 -162
  112. package/__tests__/adapters/wechat-frida-agent.test.js +0 -322
  113. package/__tests__/adapters/wechat-frida-integration.test.js +0 -149
  114. package/__tests__/adapters/wechat-frida-key-provider.test.js +0 -188
  115. package/__tests__/adapters/wechat-md5-key-provider.test.js +0 -101
  116. package/__tests__/adapters/wechat-pc-direct-read.test.js +0 -365
  117. package/__tests__/adapters/wechat-pc-group-topic.test.js +0 -63
  118. package/__tests__/adapters/wechat-pc-v4-sidecar.test.js +0 -72
  119. package/__tests__/adapters/weread.test.js +0 -123
  120. package/__tests__/adapters/wework-pc.test.js +0 -124
  121. package/__tests__/adapters/win-recent.test.js +0 -192
  122. package/__tests__/analysis-skills.test.js +0 -754
  123. package/__tests__/analysis.test.js +0 -1845
  124. package/__tests__/audio-ximalaya-snapshot.test.js +0 -279
  125. package/__tests__/batch.test.js +0 -133
  126. package/__tests__/bridges-cc-kg.test.js +0 -231
  127. package/__tests__/bridges-cc-llm.test.js +0 -191
  128. package/__tests__/bridges-cc-rag.test.js +0 -162
  129. package/__tests__/categories.test.js +0 -92
  130. package/__tests__/e2e/ai-chat-cross-source-journey.test.js +0 -213
  131. package/__tests__/e2e/full-user-journey.test.js +0 -188
  132. package/__tests__/e2e/local-data-adapters-cli.e2e.test.js +0 -146
  133. package/__tests__/entity-resolver-ingest-hook.test.js +0 -177
  134. package/__tests__/entity-resolver-stages.test.js +0 -411
  135. package/__tests__/entity-resolver-vault.test.js +0 -249
  136. package/__tests__/entity-resolver.test.js +0 -526
  137. package/__tests__/fitness-keep-snapshot.test.js +0 -224
  138. package/__tests__/fixtures/entity-resolver-200-mock.json +0 -96
  139. package/__tests__/ids.test.js +0 -45
  140. package/__tests__/integration/ai-chat-history-registry.test.js +0 -228
  141. package/__tests__/integration/aichat-wizard-end-to-end.test.js +0 -282
  142. package/__tests__/integration/cross-adapter-pipelines.test.js +0 -396
  143. package/__tests__/integration/local-data-adapters-pipeline.test.js +0 -373
  144. package/__tests__/integration/social-bilibili-pipeline.test.js +0 -261
  145. package/__tests__/integration/wechat-bootstrap-end-to-end.test.js +0 -390
  146. package/__tests__/key-providers.test.js +0 -126
  147. package/__tests__/kg-derive.test.js +0 -219
  148. package/__tests__/llm-client.test.js +0 -122
  149. package/__tests__/longtail-adapters.test.js +0 -281
  150. package/__tests__/messaging-qq-snapshot.test.js +0 -294
  151. package/__tests__/mobile-extractor-encrypted.test.js +0 -460
  152. package/__tests__/mobile-extractor.test.js +0 -288
  153. package/__tests__/mock-adapter.test.js +0 -93
  154. package/__tests__/prompt-builder.test.js +0 -249
  155. package/__tests__/query-parser.test.js +0 -365
  156. package/__tests__/rag-derive.test.js +0 -169
  157. package/__tests__/registry-readiness.test.js +0 -292
  158. package/__tests__/registry.test.js +0 -420
  159. package/__tests__/salvage-ingest.test.js +0 -97
  160. package/__tests__/schemas.test.js +0 -331
  161. package/__tests__/shopping-adapters.test.js +0 -392
  162. package/__tests__/shopping-eleme-snapshot.test.js +0 -454
  163. package/__tests__/shopping-pinduoduo-snapshot.test.js +0 -484
  164. package/__tests__/shopping-snapshot.test.js +0 -438
  165. package/__tests__/shopping-vipshop-snapshot.test.js +0 -425
  166. package/__tests__/shopping-xianyu-snapshot.test.js +0 -451
  167. package/__tests__/sidecar-contacts-cross-validate.test.js +0 -186
  168. package/__tests__/sidecar-supervisor.test.js +0 -128
  169. package/__tests__/sign-providers.test.js +0 -62
  170. package/__tests__/social-adapters.test.js +0 -280
  171. package/__tests__/social-bilibili-snapshot.test.js +0 -278
  172. package/__tests__/social-douban-snapshot.test.js +0 -351
  173. package/__tests__/social-douyin-im-direct-read.test.js +0 -377
  174. package/__tests__/social-douyin-salvage-collector.test.js +0 -98
  175. package/__tests__/social-douyin-salvage-mapper.test.js +0 -90
  176. package/__tests__/social-douyin-snapshot.test.js +0 -256
  177. package/__tests__/social-kuaishou-snapshot.test.js +0 -362
  178. package/__tests__/social-toutiao-snapshot.test.js +0 -366
  179. package/__tests__/social-weibo-snapshot.test.js +0 -234
  180. package/__tests__/social-weibo-sqlite-device.test.js +0 -174
  181. package/__tests__/social-xiaohongshu-snapshot.test.js +0 -232
  182. package/__tests__/sqlite-leaf-salvage.test.js +0 -97
  183. package/__tests__/travel-adapters.test.js +0 -483
  184. package/__tests__/travel-maps-snapshot.test.js +0 -426
  185. package/__tests__/vault-driver-error.test.js +0 -74
  186. package/__tests__/vault-search-helpers.test.js +0 -104
  187. package/__tests__/vault-search.test.js +0 -423
  188. package/__tests__/vault.test.js +0 -767
  189. package/__tests__/wechat-adapter.test.js +0 -594
  190. package/__tests__/whatsapp-adapter.test.js +0 -138
  191. package/scripts/_make-fixture-all.js +0 -126
  192. package/scripts/_make-fixture-contacts.js +0 -84
  193. package/scripts/evaluate-entity-resolver.js +0 -213
  194. package/scripts/run-native-tests-sandbox.sh +0 -55
  195. package/scripts/smoke-phase-5-5.js +0 -196
  196. package/scripts/smoke-phase-5-7.js +0 -181
  197. package/scripts/smoke-system-data-contacts.js +0 -309
  198. package/scripts/smoke-system-data.js +0 -312
  199. package/vitest.config.js +0 -88
@@ -1,411 +0,0 @@
1
- "use strict";
2
-
3
- import { describe, it, expect, vi } from "vitest";
4
-
5
- const {
6
- EntityResolverEmbeddingStage,
7
- entityResolverCosineSimilarity: cosineSimilarity,
8
- EntityResolverLLMStage,
9
- parseEntityResolverLLMResponse: parseLLMResponse,
10
- EntityResolverWorker,
11
- EntityResolver,
12
- } = require("../lib/entity-resolver");
13
- const { LocalVault } = require("../lib/vault");
14
- const { generateKeyHex } = require("../lib/key-providers");
15
-
16
- const path = require("node:path");
17
- const fs = require("node:fs");
18
- const os = require("node:os");
19
-
20
- // ─── cosineSimilarity ────────────────────────────────────────────────────
21
-
22
- describe("cosineSimilarity", () => {
23
- it("identical vectors → 1", () => {
24
- expect(cosineSimilarity([1, 2, 3], [1, 2, 3])).toBeCloseTo(1, 5);
25
- });
26
-
27
- it("orthogonal vectors → 0", () => {
28
- expect(cosineSimilarity([1, 0], [0, 1])).toBeCloseTo(0, 5);
29
- });
30
-
31
- it("opposite vectors → 0 (clamped)", () => {
32
- // Mathematically -1, but we clamp to [0,1]
33
- expect(cosineSimilarity([1, 0], [-1, 0])).toBe(0);
34
- });
35
-
36
- it("empty inputs → 0", () => {
37
- expect(cosineSimilarity([], [])).toBe(0);
38
- expect(cosineSimilarity(null, [1, 2])).toBe(0);
39
- });
40
-
41
- it("zero-norm vectors → 0", () => {
42
- expect(cosineSimilarity([0, 0, 0], [1, 2, 3])).toBe(0);
43
- });
44
-
45
- it("Float32Array works equivalently to plain array", () => {
46
- const a = new Float32Array([1, 2, 3]);
47
- const b = [1, 2, 3];
48
- expect(cosineSimilarity(a, b)).toBeCloseTo(1, 5);
49
- });
50
- });
51
-
52
- // ─── EmbeddingStage with injected embedFn ────────────────────────────────
53
-
54
- describe("EntityResolverEmbeddingStage", () => {
55
- const personA = {
56
- id: "p-a",
57
- type: "person",
58
- names: ["张三", "Zhang"],
59
- identifiers: { email: ["a@x.com"], phone: ["13800001111"] },
60
- };
61
- const personB = {
62
- id: "p-b",
63
- type: "person",
64
- names: ["张三"],
65
- identifiers: { phone: ["13800001111"] },
66
- };
67
-
68
- it("constructor accepts empty opts (defaults Ollama)", () => {
69
- const stage = new EntityResolverEmbeddingStage({});
70
- expect(stage._ollamaUrl).toContain("11434");
71
- expect(stage._model).toBe("nomic-embed-text");
72
- });
73
-
74
- it("constructor rejects non-object opts", () => {
75
- expect(() => new EntityResolverEmbeddingStage(null)).toThrow();
76
- expect(() => new EntityResolverEmbeddingStage("string")).toThrow();
77
- });
78
-
79
- it("compare returns sim + profile", async () => {
80
- const stage = new EntityResolverEmbeddingStage({
81
- embedFn: async (text) => {
82
- // Deterministic stub: hash → vec
83
- return text.length === text.length ? [1, 2, 3] : [3, 2, 1];
84
- },
85
- });
86
- const r = await stage.compare(personA, personB);
87
- expect(r.sim).toBeCloseTo(1, 5);
88
- expect(r.profileA).toContain("张三");
89
- expect(r.profileB).toContain("张三");
90
- });
91
-
92
- it("buildProfile includes name/aliases/identifiers", () => {
93
- const stage = new EntityResolverEmbeddingStage({ embedFn: async () => [1] });
94
- const p = stage.buildProfile(personA);
95
- expect(p).toContain("张三");
96
- expect(p).toContain("Zhang");
97
- expect(p).toContain("email:a@x.com");
98
- expect(p).toContain("phone:13800001111");
99
- });
100
-
101
- it("caches embeddings per person id (1 embedFn call for repeated person)", async () => {
102
- let callCount = 0;
103
- const stage = new EntityResolverEmbeddingStage({
104
- embedFn: async () => { callCount += 1; return [1, 2, 3]; },
105
- });
106
- await stage.compare(personA, personB);
107
- await stage.compare(personA, personB);
108
- expect(callCount).toBe(2); // first compare embeds both; second uses cache
109
- });
110
-
111
- it("cache evicts FIFO when over cacheMaxSize", async () => {
112
- const stage = new EntityResolverEmbeddingStage({
113
- embedFn: async () => [1, 2, 3],
114
- cacheMaxSize: 2,
115
- });
116
- await stage.compare({ id: "p-1", names: ["a"], identifiers: {} }, { id: "p-2", names: ["b"], identifiers: {} });
117
- await stage.compare({ id: "p-3", names: ["c"], identifiers: {} }, { id: "p-4", names: ["d"], identifiers: {} });
118
- expect(stage._cache.size).toBeLessThanOrEqual(2);
119
- });
120
-
121
- it("asStageFn returns bound function", async () => {
122
- const stage = new EntityResolverEmbeddingStage({
123
- embedFn: async () => [1, 0, 0],
124
- });
125
- const fn = stage.asStageFn();
126
- const r = await fn(personA, personB);
127
- expect(r.sim).toBeDefined();
128
- });
129
-
130
- it("rejects non-array embedFn output", async () => {
131
- const stage = new EntityResolverEmbeddingStage({
132
- embedFn: async () => "not an array",
133
- });
134
- await expect(stage.compare(personA, personB)).rejects.toThrow(/Array/);
135
- });
136
- });
137
-
138
- // ─── parseLLMResponse ────────────────────────────────────────────────────
139
-
140
- describe("parseLLMResponse", () => {
141
- it("parses strict JSON", () => {
142
- const r = parseLLMResponse('{"same":true,"confidence":0.9,"reason":"phone match"}');
143
- expect(r.same).toBe(true);
144
- expect(r.confidence).toBe(0.9);
145
- });
146
-
147
- it("parses fenced JSON", () => {
148
- const r = parseLLMResponse('```json\n{"same":false,"confidence":0.8,"reason":"different ids"}\n```');
149
- expect(r.same).toBe(false);
150
- });
151
-
152
- it("parses fenced JSON without language tag", () => {
153
- const r = parseLLMResponse('```\n{"same":null,"confidence":0.5,"reason":"unclear"}\n```');
154
- expect(r.same).toBeNull();
155
- });
156
-
157
- it("falls back to regex extraction from messy preamble", () => {
158
- const text = 'Here is my analysis:\n{"same":true,"confidence":0.85,"reason":"shared phone"}\nLet me know if you have questions.';
159
- const r = parseLLMResponse(text);
160
- expect(r.same).toBe(true);
161
- });
162
-
163
- it("returns null on totally garbage input", () => {
164
- expect(parseLLMResponse("just some prose, no json")).toBeNull();
165
- expect(parseLLMResponse("")).toBeNull();
166
- expect(parseLLMResponse(null)).toBeNull();
167
- });
168
-
169
- it("ignores JSON objects without a 'same' field", () => {
170
- expect(parseLLMResponse('{"foo":"bar"}')).toBeNull();
171
- });
172
- });
173
-
174
- // ─── LLMStage ────────────────────────────────────────────────────────────
175
-
176
- describe("EntityResolverLLMStage", () => {
177
- const a = { id: "p-a", names: ["张三"], identifiers: { phone: ["13800001111"] }, source: { adapter: "email" } };
178
- const b = { id: "p-b", names: ["张三"], identifiers: { phone: ["13800001111"] }, source: { adapter: "alipay" } };
179
-
180
- it("constructor requires llm with .chat()", () => {
181
- expect(() => new EntityResolverLLMStage()).toThrow();
182
- expect(() => new EntityResolverLLMStage({})).toThrow();
183
- expect(() => new EntityResolverLLMStage({ llm: {} })).toThrow();
184
- });
185
-
186
- it("arbitrate maps same:true → verdict yes", async () => {
187
- const stage = new EntityResolverLLMStage({
188
- llm: {
189
- isLocal: true,
190
- chat: async () => ({ text: '{"same":true,"confidence":0.92,"reason":"phone match"}' }),
191
- },
192
- });
193
- const r = await stage.arbitrate(a, b);
194
- expect(r.verdict).toBe("yes");
195
- expect(r.confidence).toBe(0.92);
196
- });
197
-
198
- it("arbitrate maps same:false → verdict no", async () => {
199
- const stage = new EntityResolverLLMStage({
200
- llm: {
201
- isLocal: true,
202
- chat: async () => ({ text: '{"same":false,"confidence":0.8,"reason":"different ids"}' }),
203
- },
204
- });
205
- const r = await stage.arbitrate(a, b);
206
- expect(r.verdict).toBe("no");
207
- });
208
-
209
- it("arbitrate maps same:null → verdict maybe", async () => {
210
- const stage = new EntityResolverLLMStage({
211
- llm: {
212
- isLocal: true,
213
- chat: async () => ({ text: '{"same":null,"confidence":0.5,"reason":"insufficient evidence"}' }),
214
- },
215
- });
216
- const r = await stage.arbitrate(a, b);
217
- expect(r.verdict).toBe("maybe");
218
- });
219
-
220
- it("unparseable LLM response → maybe + reason", async () => {
221
- const stage = new EntityResolverLLMStage({
222
- llm: {
223
- isLocal: true,
224
- chat: async () => ({ text: 'just some prose' }),
225
- },
226
- });
227
- const r = await stage.arbitrate(a, b);
228
- expect(r.verdict).toBe("maybe");
229
- expect(r.confidence).toBe(0);
230
- expect(r.reason).toMatch(/not parseable/);
231
- });
232
-
233
- it("non-local LLM without acceptNonLocal → refuses + returns maybe", async () => {
234
- const stage = new EntityResolverLLMStage({
235
- llm: {
236
- isLocal: false,
237
- chat: async () => { throw new Error("should not be called"); },
238
- },
239
- });
240
- const r = await stage.arbitrate(a, b);
241
- expect(r.verdict).toBe("maybe");
242
- expect(r.reason).toMatch(/non-local/);
243
- });
244
-
245
- it("non-local LLM with acceptNonLocal:true → allowed", async () => {
246
- const stage = new EntityResolverLLMStage({
247
- llm: {
248
- isLocal: false,
249
- chat: async () => ({ text: '{"same":true,"confidence":0.9,"reason":"ok"}' }),
250
- },
251
- acceptNonLocal: true,
252
- });
253
- const r = await stage.arbitrate(a, b);
254
- expect(r.verdict).toBe("yes");
255
- });
256
-
257
- it("LLM throwing propagates (caller handles retry)", async () => {
258
- const stage = new EntityResolverLLMStage({
259
- llm: {
260
- isLocal: true,
261
- chat: async () => { throw new Error("ollama down"); },
262
- },
263
- });
264
- await expect(stage.arbitrate(a, b)).rejects.toThrow(/ollama/);
265
- });
266
-
267
- it("confidence outside [0,1] gets clamped", async () => {
268
- const stage = new EntityResolverLLMStage({
269
- llm: {
270
- isLocal: true,
271
- chat: async () => ({ text: '{"same":true,"confidence":42}' }),
272
- },
273
- });
274
- const r = await stage.arbitrate(a, b);
275
- expect(r.confidence).toBe(1);
276
- });
277
- });
278
-
279
- // ─── EntityResolverWorker ───────────────────────────────────────────────
280
-
281
- function makeMockResolver(initialQueue, drainResults) {
282
- let i = 0;
283
- const queue = [...initialQueue];
284
- return {
285
- drain: async ({ limit }) => {
286
- if (queue.length === 0) {
287
- return { processed: 0, same: 0, different: 0, review: 0, error: 0, skipped: 0 };
288
- }
289
- const result = drainResults[i] || {
290
- processed: Math.min(limit, queue.length),
291
- same: 0, different: 0, review: 0, error: 0, skipped: 0,
292
- };
293
- queue.splice(0, result.processed);
294
- i += 1;
295
- return result;
296
- },
297
- _queueLength: () => queue.length,
298
- };
299
- }
300
-
301
- describe("EntityResolverWorker", () => {
302
- it("constructor requires resolver", () => {
303
- expect(() => new EntityResolverWorker()).toThrow();
304
- expect(() => new EntityResolverWorker({})).toThrow(/resolver/);
305
- });
306
-
307
- it("tick processes one batch and updates stats", async () => {
308
- const resolver = makeMockResolver(["p1", "p2", "p3"], [
309
- { processed: 3, same: 1, different: 1, review: 1, error: 0, skipped: 0 },
310
- ]);
311
- const worker = new EntityResolverWorker({ resolver });
312
- const r = await worker.tick();
313
- expect(r.processed).toBe(3);
314
- const s = worker.stats();
315
- expect(s.batchesProcessed).toBe(1);
316
- expect(s.itemsProcessed).toBe(3);
317
- expect(s.same).toBe(1);
318
- expect(s.review).toBe(1);
319
- });
320
-
321
- it("start + stop cycle works without leaks", async () => {
322
- const resolver = makeMockResolver([], []);
323
- const worker = new EntityResolverWorker({ resolver, idleIntervalMs: 100 });
324
- expect(worker.isRunning()).toBe(false);
325
- worker.start();
326
- expect(worker.isRunning()).toBe(true);
327
- // Give the loop one tick
328
- await new Promise((resolve) => setTimeout(resolve, 250));
329
- await worker.stop();
330
- expect(worker.isRunning()).toBe(false);
331
- });
332
-
333
- it("processes queue then sleeps; onProgress fires per batch", async () => {
334
- const resolver = makeMockResolver(["p1", "p2"], [
335
- { processed: 2, same: 2, different: 0, review: 0, error: 0, skipped: 0 },
336
- ]);
337
- const events = [];
338
- const worker = new EntityResolverWorker({
339
- resolver,
340
- idleIntervalMs: 5000, // long so we observe just the first batch
341
- batchSpacingMs: 1,
342
- onProgress: (e) => events.push(e),
343
- });
344
- worker.start();
345
- // Give time for one batch + post-batch tiny delay
346
- await new Promise((resolve) => setTimeout(resolve, 200));
347
- await worker.stop();
348
- expect(events.length).toBeGreaterThanOrEqual(1);
349
- expect(events[0].batch.same).toBe(2);
350
- });
351
-
352
- it("onProgress listener errors do NOT break the loop", async () => {
353
- const resolver = makeMockResolver(["p1"], [
354
- { processed: 1, same: 0, different: 0, review: 0, error: 0, skipped: 1 },
355
- ]);
356
- const worker = new EntityResolverWorker({
357
- resolver,
358
- idleIntervalMs: 5000,
359
- onProgress: () => { throw new Error("listener boom"); },
360
- });
361
- worker.start();
362
- await new Promise((resolve) => setTimeout(resolve, 200));
363
- await worker.stop();
364
- expect(worker.stats().batchesProcessed).toBeGreaterThanOrEqual(1);
365
- });
366
- });
367
-
368
- // ─── End-to-end: real EntityResolver + real EmbeddingStage stub + real Worker ──
369
-
370
- describe("EntityResolver + EmbeddingStage + LLMStage + Worker integration", () => {
371
- it("queued uncertain pair → embedding high sim → auto-merged", async () => {
372
- const dir = fs.mkdtempSync(path.join(os.tmpdir(), "er-int-"));
373
- const vault = new LocalVault({ path: path.join(dir, "v.db"), key: generateKeyHex() });
374
- vault.open();
375
- try {
376
- // 2 persons sharing only name (rule → uncertain)
377
- const baseSource = { adapter: "test", adapterVersion: "0.1", originalId: "x", capturedAt: Date.now(), capturedBy: "api" };
378
- vault.putPerson({
379
- id: "p-a", type: "person", subtype: "contact",
380
- names: ["张三"], identifiers: {}, ingestedAt: Date.now(),
381
- source: { ...baseSource, adapter: "email", originalId: "1" },
382
- });
383
- vault.putPerson({
384
- id: "p-b", type: "person", subtype: "contact",
385
- names: ["张三"], identifiers: {}, ingestedAt: Date.now(),
386
- source: { ...baseSource, adapter: "alipay", originalId: "2" },
387
- });
388
-
389
- const embedStage = new EntityResolverEmbeddingStage({
390
- embedFn: async () => [1, 0, 0], // deterministic identical vec → sim=1
391
- });
392
- const llmStage = new EntityResolverLLMStage({
393
- llm: { isLocal: true, chat: async () => ({ text: '{"same":true,"confidence":0.9,"reason":"name"}' }) },
394
- });
395
- const resolver = new EntityResolver({
396
- vault,
397
- embeddingStage: embedStage.asStageFn(),
398
- llmStage: llmStage.asStageFn(),
399
- });
400
-
401
- vault.enqueueResolve("p-b");
402
- const worker = new EntityResolverWorker({ resolver, batchSize: 10 });
403
- const r = await worker.tick();
404
- expect(r.same).toBe(1);
405
- expect(vault.getMergeGroupMembers("p-a").sort()).toEqual(["p-a", "p-b"]);
406
- } finally {
407
- try { vault.close(); } catch (_e) {}
408
- try { fs.rmSync(dir, { recursive: true, force: true }); } catch (_e) {}
409
- }
410
- });
411
- });
@@ -1,249 +0,0 @@
1
- "use strict";
2
-
3
- import { describe, it, expect, beforeEach, afterEach } from "vitest";
4
-
5
- const path = require("node:path");
6
- const fs = require("node:fs");
7
- const os = require("node:os");
8
- const { LocalVault } = require("../lib/vault");
9
- const { generateKeyHex } = require("../lib/key-providers");
10
- const { TARGET_VERSION } = require("../lib/migrations");
11
-
12
- // Helper to spin up a fresh vault each test
13
- function makeVault() {
14
- const dir = fs.mkdtempSync(path.join(os.tmpdir(), "hub-er-test-"));
15
- const dbPath = path.join(dir, "vault.db");
16
- const key = generateKeyHex();
17
- const vault = new LocalVault({ path: dbPath, key });
18
- vault.open();
19
- return { vault, dir };
20
- }
21
-
22
- function cleanup(vault, dir) {
23
- try { vault.close(); } catch (_e) {}
24
- try { fs.rmSync(dir, { recursive: true, force: true }); } catch (_e) {}
25
- }
26
-
27
- // ─── Migration v2 — Phase 8 tables exist ─────────────────────────────────
28
-
29
- describe("Phase 8 migration v2 — EntityResolver tables", () => {
30
- let vault, dir;
31
- beforeEach(() => { ({ vault, dir } = makeVault()); });
32
- afterEach(() => cleanup(vault, dir));
33
-
34
- it("schemaVersion is current after open()", () => {
35
- expect(vault.schemaVersion()).toBe(TARGET_VERSION);
36
- // Phase 8 ER tables landed in v2; subsequent migrations must not regress.
37
- expect(TARGET_VERSION).toBeGreaterThanOrEqual(2);
38
- });
39
-
40
- it("all 5 new tables exist + are queryable", () => {
41
- const s = vault.stats();
42
- expect(s.mergeGroups).toBe(0);
43
- expect(s.mergeMembers).toBe(0);
44
- expect(s.resolveQueue).toBe(0);
45
- expect(s.reviewQueue).toBe(0);
46
- expect(s.resolveDecisions).toBe(0);
47
- });
48
-
49
- });
50
-
51
- // ─── resolve_queue ───────────────────────────────────────────────────────
52
-
53
- describe("Phase 8 — resolve_queue helpers", () => {
54
- let vault, dir;
55
- beforeEach(() => { ({ vault, dir } = makeVault()); });
56
- afterEach(() => cleanup(vault, dir));
57
-
58
- it("enqueueResolve creates a pending row", () => {
59
- const id = vault.enqueueResolve("person-a");
60
- expect(typeof id === "number" || typeof id === "bigint").toBe(true);
61
- expect(vault.resolveQueueStats().pending).toBe(1);
62
- });
63
-
64
- it("enqueueResolve is idempotent for pending rows", () => {
65
- const a = vault.enqueueResolve("person-a");
66
- const b = vault.enqueueResolve("person-a");
67
- expect(a).toBe(b);
68
- expect(vault.resolveQueueStats().pending).toBe(1);
69
- });
70
-
71
- it("claimResolveBatch marks pending → in-progress + increments attempts", () => {
72
- vault.enqueueResolve("person-a");
73
- vault.enqueueResolve("person-b");
74
- const batch = vault.claimResolveBatch(50);
75
- expect(batch).toHaveLength(2);
76
- expect(batch[0].person_id).toBe("person-a");
77
- expect(batch[0].attempts).toBe(0); // before increment
78
- const stats = vault.resolveQueueStats();
79
- expect(stats.pending).toBe(0);
80
- expect(stats["in-progress"]).toBe(2);
81
- });
82
-
83
- it("completeResolve flips status to done", () => {
84
- vault.enqueueResolve("person-a");
85
- const [row] = vault.claimResolveBatch(10);
86
- vault.completeResolve(row.id);
87
- expect(vault.resolveQueueStats().done).toBe(1);
88
- });
89
-
90
- it("errorResolve re-pends until 3 attempts then marks error", () => {
91
- vault.enqueueResolve("person-a");
92
- let row = vault.claimResolveBatch(10)[0];
93
- vault.errorResolve(row.id, "fail 1");
94
- expect(vault.resolveQueueStats().pending).toBe(1);
95
- row = vault.claimResolveBatch(10)[0];
96
- vault.errorResolve(row.id, "fail 2");
97
- expect(vault.resolveQueueStats().pending).toBe(1);
98
- row = vault.claimResolveBatch(10)[0];
99
- vault.errorResolve(row.id, "fail 3");
100
- expect(vault.resolveQueueStats().error).toBe(1);
101
- expect(vault.resolveQueueStats().pending).toBe(0);
102
- });
103
- });
104
-
105
- // ─── resolve_decisions ───────────────────────────────────────────────────
106
-
107
- describe("Phase 8 — resolve_decisions", () => {
108
- let vault, dir;
109
- beforeEach(() => { ({ vault, dir } = makeVault()); });
110
- afterEach(() => cleanup(vault, dir));
111
-
112
- it("recordResolveDecision lex-orders ids", () => {
113
- vault.recordResolveDecision({
114
- aId: "z-person", bId: "a-person",
115
- verdict: "same", confidence: 0.9, decidedBy: "rule", reason: "test",
116
- });
117
- // Looking up either order returns the same row
118
- const r1 = vault.getResolveDecision("a-person", "z-person");
119
- const r2 = vault.getResolveDecision("z-person", "a-person");
120
- expect(r1).toBeDefined();
121
- expect(r2).toBeDefined();
122
- expect(r1.a_person_id).toBe("a-person");
123
- expect(r1.b_person_id).toBe("z-person");
124
- expect(r1.verdict).toBe("same");
125
- });
126
-
127
- it("recordResolveDecision upserts on conflict", () => {
128
- vault.recordResolveDecision({ aId: "p-a", bId: "p-b", verdict: "same", confidence: 0.7, decidedBy: "embedding" });
129
- vault.recordResolveDecision({ aId: "p-a", bId: "p-b", verdict: "different", confidence: 0.95, decidedBy: "user" });
130
- const r = vault.getResolveDecision("p-a", "p-b");
131
- expect(r.verdict).toBe("different");
132
- expect(r.decided_by).toBe("user");
133
- expect(r.confidence).toBe(0.95);
134
- });
135
- });
136
-
137
- // ─── mergePair / unmergePerson / merge group closure ─────────────────────
138
-
139
- describe("Phase 8 — merge_groups + merge_members", () => {
140
- let vault, dir;
141
- beforeEach(() => { ({ vault, dir } = makeVault()); });
142
- afterEach(() => cleanup(vault, dir));
143
-
144
- it("mergePair creates a new group with 2 members", () => {
145
- const groupId = vault.mergePair({ aId: "p-a", bId: "p-b" });
146
- expect(typeof groupId).toBe("string");
147
- const members = vault.getMergeGroupMembers("p-a");
148
- expect(members.sort()).toEqual(["p-a", "p-b"]);
149
- });
150
-
151
- it("mergePair adds to existing group when one side already in it", () => {
152
- vault.mergePair({ aId: "p-a", bId: "p-b" });
153
- vault.mergePair({ aId: "p-b", bId: "p-c" });
154
- const members = vault.getMergeGroupMembers("p-a");
155
- expect(members.sort()).toEqual(["p-a", "p-b", "p-c"]);
156
- });
157
-
158
- it("mergePair fuses two existing groups", () => {
159
- vault.mergePair({ aId: "p-a", bId: "p-b" });
160
- vault.mergePair({ aId: "p-c", bId: "p-d" });
161
- // a-b group and c-d group exist independently
162
- expect(vault.getMergeGroupMembers("p-a").sort()).toEqual(["p-a", "p-b"]);
163
- expect(vault.getMergeGroupMembers("p-c").sort()).toEqual(["p-c", "p-d"]);
164
- // Fuse them
165
- vault.mergePair({ aId: "p-b", bId: "p-c" });
166
- const all = vault.getMergeGroupMembers("p-a").sort();
167
- expect(all).toEqual(["p-a", "p-b", "p-c", "p-d"]);
168
- // Verify there's only ONE group remaining
169
- const s = vault.stats();
170
- expect(s.mergeGroups).toBe(1);
171
- expect(s.mergeMembers).toBe(4);
172
- });
173
-
174
- it("mergePair on already-same-group is a no-op", () => {
175
- vault.mergePair({ aId: "p-a", bId: "p-b" });
176
- const groupBefore = vault.getMergeGroupMembers("p-a").length;
177
- vault.mergePair({ aId: "p-a", bId: "p-b" });
178
- const groupAfter = vault.getMergeGroupMembers("p-a").length;
179
- expect(groupAfter).toBe(groupBefore);
180
- });
181
-
182
- it("getMergeGroupMembers returns [personId] for unmerged person", () => {
183
- const members = vault.getMergeGroupMembers("loner-1");
184
- expect(members).toEqual(["loner-1"]);
185
- });
186
-
187
- it("unmergePerson removes from group; group of 2 → group dissolves", () => {
188
- vault.mergePair({ aId: "p-a", bId: "p-b" });
189
- const r = vault.unmergePerson("p-a");
190
- expect(r.ok).toBe(true);
191
- expect(vault.getMergeGroupMembers("p-a")).toEqual(["p-a"]);
192
- expect(vault.getMergeGroupMembers("p-b")).toEqual(["p-b"]);
193
- expect(vault.stats().mergeGroups).toBe(0);
194
- });
195
-
196
- it("unmergePerson shrinks group when 3+ members", () => {
197
- vault.mergePair({ aId: "p-a", bId: "p-b" });
198
- vault.mergePair({ aId: "p-b", bId: "p-c" });
199
- vault.unmergePerson("p-c");
200
- expect(vault.getMergeGroupMembers("p-a").sort()).toEqual(["p-a", "p-b"]);
201
- expect(vault.getMergeGroupMembers("p-c")).toEqual(["p-c"]);
202
- expect(vault.stats().mergeGroups).toBe(1);
203
- });
204
-
205
- it("unmergePerson on non-member returns ok:false", () => {
206
- const r = vault.unmergePerson("never-merged");
207
- expect(r.ok).toBe(false);
208
- });
209
- });
210
-
211
- // ─── review_queue ────────────────────────────────────────────────────────
212
-
213
- describe("Phase 8 — review_queue", () => {
214
- let vault, dir;
215
- beforeEach(() => { ({ vault, dir } = makeVault()); });
216
- afterEach(() => cleanup(vault, dir));
217
-
218
- it("enqueueReview lex-orders + returns id", () => {
219
- const id = vault.enqueueReview({
220
- aId: "z-p", bId: "a-p",
221
- embedSim: 0.72, llmVerdict: "maybe", llmReason: "names similar",
222
- llmConfidence: 0.5,
223
- });
224
- expect(id).toBeDefined();
225
- const queue = vault.listReviewQueue();
226
- expect(queue).toHaveLength(1);
227
- expect(queue[0].a_person_id).toBe("a-p");
228
- expect(queue[0].b_person_id).toBe("z-p");
229
- expect(queue[0].embed_sim).toBe(0.72);
230
- });
231
-
232
- it("listReviewQueue only returns pending rows (reviewed_at IS NULL)", () => {
233
- const id1 = vault.enqueueReview({ aId: "p1", bId: "p2", embedSim: 0.7 });
234
- const id2 = vault.enqueueReview({ aId: "p3", bId: "p4", embedSim: 0.75 });
235
- vault.recordReviewDecision({ reviewId: id1, decision: "same" });
236
- const pending = vault.listReviewQueue();
237
- expect(pending).toHaveLength(1);
238
- expect(pending[0].id).toBe(id2);
239
- });
240
-
241
- it("recordReviewDecision rejects invalid decision", () => {
242
- const id = vault.enqueueReview({ aId: "p1", bId: "p2" });
243
- expect(() => vault.recordReviewDecision({ reviewId: id, decision: "garbage" })).toThrow();
244
- });
245
-
246
- it("recordReviewDecision rejects non-existent review id", () => {
247
- expect(() => vault.recordReviewDecision({ reviewId: 9999, decision: "same" })).toThrow();
248
- });
249
- });