memind/memory-v2-pgvector.test.mjs

import assert from 'node:assert/strict';
import test from 'node:test';
import { createMemoryV2 } from './memory-v2.mjs';
import {
  createPgvectorMemoryBackend,
  pgvectorMemoryBackendInternals,
} from './memory-v2-pgvector.mjs';

test('pgvector backend is disabled by default and does not query storage', async () => {
  let queried = false;
  const backend = createPgvectorMemoryBackend({
    pool: {
      async query() {
        queried = true;
        return { rows: [] };
      },
    },
  });

  assert.equal(backend.isAvailable(), false);
  assert.deepEqual(await backend.resolve({
    userId: 'user-1',
    embedding: [0.1, 0.2],
  }), {
    memories: [],
    semanticMemories: [],
  });
  assert.equal(queried, false);
});

test('pgvector backend returns empty semantic result when embedding is unavailable', async () => {
  let queried = false;
  const backend = createPgvectorMemoryBackend({
    enabled: true,
    pool: {
      async query() {
        queried = true;
        return { rows: [] };
      },
    },
  });

  const result = await backend.resolve({
    userId: 'user-1',
    query: 'memory-chain',
  });

  assert.deepEqual(result, {
    memories: [],
    semanticMemories: [],
  });
  assert.equal(queried, false);
});

test('pgvector backend performs parameterized vector lookup when explicitly enabled', async () => {
  const queries = [];
  const backend = createPgvectorMemoryBackend({
    enabled: true,
    tableName: 'memory_embeddings',
    pool: {
      async query(sql, params) {
        queries.push({ sql, params });
        return {
          rows: [
            {
              id: 7,
              content: '用户关注 Memory V2 的 facade 边界',
              type: 'fact',
              score: '0.87',
              created_at: '2026-07-02T00:00:00.000Z',
            },
          ],
        };
      },
    },
    embedQuery: async (query) => {
      assert.equal(query, 'memory-chain');
      return [0.25, 0.5, 0.75];
    },
  });

  assert.equal(backend.isAvailable(), true);
  const result = await backend.resolve({
    userId: 'user-1',
    query: 'memory-chain',
    limit: 5,
  });

  assert.equal(queries.length, 2);
  assert.match(queries[0].sql, /FROM memory_embeddings/);
  assert.match(queries[0].sql, /WITH top_score/);
  assert.match(queries[0].sql, /vector_candidates/);
  assert.match(queries[0].sql, /recent_candidates/);
  assert.deepEqual(queries[0].params, ['user-1', '[0.25,0.5,0.75]', 50, 0.15, 200]);
  assert.match(queries[1].sql, /ILIKE/);
  assert.deepEqual(result.semanticMemories, ['用户关注 Memory V2 的 facade 边界']);
  assert.deepEqual(result.memories, [
    {
      id: '7',
      label: 'fact',
      text: '用户关注 Memory V2 的 facade 边界',
      score: 0.87,
      createdAt: '2026-07-02T00:00:00.000Z',
      updatedAt: '2026-07-02T00:00:00.000Z',
    },
  ]);
});

test('pgvector hybrid ranking recovers a recent Chinese memory missed by vector top-k', async () => {
  const marker = 'MEM-RECALL-NEW';
  const backend = createPgvectorMemoryBackend({
    enabled: true,
    pool: {
      async query() {
        return {
          rows: [
            {
              id: 1,
              content: '用户以前关注贵州旅游攻略',
              type: 'interest',
              score: 0.91,
              created_at: '2026-06-01T00:00:00.000Z',
              updated_at: '2026-06-01T00:00:00.000Z',
            },
            {
              id: 2,
              content: `用户的记忆召回灰度测试代号是 ${marker}`,
              type: 'fact',
              score: -0.49,
              created_at: '2026-07-22T00:00:00.000Z',
              updated_at: '2026-07-22T00:00:00.000Z',
            },
            {
              id: 3,
              content: '用户偏好简洁回答',
              type: 'preference',
              score: 0.3,
              created_at: '2026-07-21T00:00:00.000Z',
              updated_at: '2026-07-21T00:00:00.000Z',
            },
          ],
        };
      },
    },
    embedQuery: async () => [0.25, 0.5, 0.75],
  });

  const result = await backend.resolve({
    userId: 'user-1',
    query: '我之前让你记住的记忆召回灰度测试代号是什么？请只回答完整代号。',
    limit: 3,
  });

  assert.match(result.memories[0].text, new RegExp(marker));
});

test('recallRankingWeights boosts vector only when semantic spread is visible', () => {
  const { recallRankingWeights } = pgvectorMemoryBackendInternals;
  const flat = recallRankingWeights('Why did Sarah close the curtains?', [
    { vectorScore: 0.41 },
    { vectorScore: 0.39 },
    { vectorScore: 0.38 },
  ]);
  assert.deepEqual(flat, { lexical: 1, vector: 1 });

  const semantic = recallRankingWeights('Why did Sarah close the curtains?', [
    { vectorScore: 0.82 },
    { vectorScore: 0.55 },
    { vectorScore: 0.41 },
  ]);
  assert.deepEqual(semantic, { lexical: 0.45, vector: 1.55 });
});

test('pgvector RRF hybrid ranking promotes semantic vector match over topical noise', () => {
  const ranked = pgvectorMemoryBackendInternals.rankHybridCandidates([
    {
      id: 'noise',
      content: 'The curtains the curtains the curtains were recently updated in the living room',
      score: 0.42,
    },
    {
      id: 'gold',
      content: 'Sarah closed the smart curtains to reduce pollen entry',
      score: 0.86,
    },
  ], 'Why did Sarah close the curtains?', 1);
  assert.equal(ranked[0].id, 'gold');
});

test('pgvector keyword-only rows do not pollute vector RRF ranks', () => {
  const ranked = pgvectorMemoryBackendInternals.rankHybridCandidates([
    {
      id: 'keyword-noise',
      content: 'The curtains the curtains the curtains were recently updated in the living room',
      score: null,
    },
    {
      id: 'vector-gold',
      content: 'Sarah closed the smart curtains to reduce pollen entry',
      score: 0.86,
    },
  ], 'Why did Sarah close the curtains?', 1);
  assert.equal(ranked[0].id, 'vector-gold');
});

test('pgvector semantic spread attenuates zero-overlap vector-only noise', () => {
  const ranked = pgvectorMemoryBackendInternals.rankHybridCandidates([
    {
      id: 'vector-noise',
      content: 'Ambient living room humidity sensor calibration report for May',
      score: 0.91,
    },
    {
      id: 'weak-overlap-gold',
      content: 'David reported his back felt sore after the morning stretch routine',
      score: 0.68,
    },
  ], "How was David's back today?", 1);
  assert.equal(ranked[0].id, 'weak-overlap-gold');
});

test('pgvector hybrid ranking keeps vector order when query has no lexical overlap', () => {
  const ranked = pgvectorMemoryBackendInternals.rankHybridCandidates([
    { id: 1, content: 'alpha', score: 0.2 },
    { id: 2, content: 'beta', score: 0.8 },
  ], '完全无关的中文查询', 2);
  assert.equal(ranked[0].id, '2');
  assert.equal(ranked[1].id, '1');
});

test('pgvector keyword fallback ranks ILIKE matches by lexical coverage, not recency', async () => {
  const backend = createPgvectorMemoryBackend({
    enabled: true,
    pool: {
      async query(sql) {
        if (String(sql).includes('ILIKE')) {
          assert.doesNotMatch(String(sql), /ORDER BY updated_at DESC/i);
          return {
            rows: [
              {
                id: 902,
                content: 'Recent but weak match for curtains only',
                type: 'noise',
                score: 1,
                created_at: '2026-09-01T00:00:00.000Z',
                updated_at: '2026-09-01T00:00:00.000Z',
              },
              {
                id: 901,
                content: 'Sarah closed the smart curtains to reduce pollen entry',
                type: 'fact',
                score: 1,
                created_at: '2026-05-01T00:00:00.000Z',
                updated_at: '2026-05-01T00:00:00.000Z',
              },
            ],
          };
        }
        return { rows: [] };
      },
    },
    embedQuery: async () => [0.25, 0.5, 0.75],
  });

  const result = await backend.resolve({
    userId: 'user-1',
    query: 'Why did Sarah close the curtains?',
    limit: 1,
  });

  assert.match(result.memories[0].text, /Sarah closed the smart curtains/);
});

test('pgvector keyword fallback retrieves topic memories missed by vector top-k', async () => {
  const queries = [];
  const backend = createPgvectorMemoryBackend({
    enabled: true,
    pool: {
      async query(sql, params) {
        queries.push({ sql, params });
        if (String(sql).includes('ILIKE')) {
          return {
            rows: [{
              id: 901,
              content: '用户对日本战国人物德川家康感兴趣，并希望继续深入讨论',
              type: 'interest',
              score: 1,
              created_at: '2026-08-01T13:57:00.000Z',
              updated_at: '2026-08-01T13:57:00.000Z',
            }],
          };
        }
        return {
          rows: [{
            id: 1,
            content: '用户以后只要说“帮我搜索今天国际国内热门新闻和小知识，做成页面”',
            type: 'preference',
            score: 0.91,
            created_at: '2026-07-31T00:00:00.000Z',
            updated_at: '2026-07-31T00:00:00.000Z',
          }],
        };
      },
    },
    embedQuery: async () => [0.25, 0.5, 0.75],
  });

  const result = await backend.resolve({
    userId: 'user-tang',
    query: '我们继续聊聊德川家康',
    limit: 2,
  });

  assert.match(result.memories[0].text, /德川家康/);
  assert.equal(queries.some((entry) => String(entry.sql).includes('ILIKE')), true);
});

test('extractKeywordTerms keeps topic phrases and drops recall boilerplate', () => {
  const terms = pgvectorMemoryBackendInternals.extractKeywordTerms('我们之前有聊过，你记得吗');
  assert.equal(terms.includes('记得'), false);
  assert.equal(terms.includes('我们'), false);
  const topicTerms = pgvectorMemoryBackendInternals.extractKeywordTerms('我们继续聊聊德川家康');
  assert.equal(topicTerms.some((term) => term.includes('德川')), true);
});

test('latinWordQueryCoverage ranks topical English content over bigram noise', () => {
  const { latinWordQueryCoverage, queryScriptProfile } = pgvectorMemoryBackendInternals;
  const query = 'Why did Sarah close the curtains?';
  assert.equal(queryScriptProfile(query), 'latin');
  const gold = 'Sarah closed the smart curtains to reduce pollen entry';
  const noise = 'The curtains the curtains the curtains were recently updated';
  assert.ok(
    latinWordQueryCoverage(query, gold) > latinWordQueryCoverage(query, noise),
  );
});

test('extractKeywordTerms drops English recall boilerplate', () => {
  const terms = pgvectorMemoryBackendInternals.extractKeywordTerms('Why did Sarah close the curtains?');
  assert.equal(terms.includes('why'), false);
  assert.equal(terms.includes('did'), false);
  assert.equal(terms.includes('the'), false);
  assert.equal(terms.includes('sarah'), true);
  assert.equal(terms.includes('curtains'), true);
});

test('extractKeywordTerms prioritizes proper nouns over generic English terms', () => {
  const terms = pgvectorMemoryBackendInternals.extractKeywordTerms(
    'Can you piece together what happened with Ethan from soccer practice until he got home?',
  );
  assert.equal(terms[0], 'ethan');
  assert.equal(terms.includes('together'), false);
  assert.equal(terms.includes('happened'), false);
  assert.equal(terms.includes('soccer'), true);
});

test('selectKeywordCandidateRows keeps priority-term gold under broad OR truncation', () => {
  const { selectKeywordCandidateRows, extractKeywordTerms } = pgvectorMemoryBackendInternals;
  const query = 'Why did David promise Ethan extra LEGO time on weekends?';
  const terms = extractKeywordTerms(query);
  const goldId = 'gold-lego';
  const rows = [
    { id: goldId, content: 'David promised Ethan extra LEGO time on weekends during recovery' },
    ...Array.from({ length: 900 }, (_entry, index) => ({
      id: `noise-${index}`,
      content: `David mentioned schedule item ${index} for the household calendar update`,
    })),
  ];
  const selected = selectKeywordCandidateRows(rows, query, terms, { returnLimit: 50 });
  assert.ok(selected.some((row) => row.id === goldId));
});

test('selectVectorCandidateRows expands margin band beyond fixed top-N', () => {
  const { selectVectorCandidateRows } = pgvectorMemoryBackendInternals;
  const scored = [
    { row: { id: 'top' }, score: 0.9 },
    ...Array.from({ length: 120 }, (_entry, index) => ({
      row: { id: `filler-${index}` },
      score: 0.85 - index * 0.001,
    })),
    { row: { id: 'near-gold' }, score: 0.79 },
  ];
  const selected = selectVectorCandidateRows(scored, { baseLimit: 100, margin: 0.12, expandCap: 200 });
  const ids = selected.map((entry) => entry.row.id);
  assert.ok(ids.includes('near-gold'));
  assert.ok(!ids.includes('filler-119') || ids.includes('near-gold'));
});

test('pgvector backend validates table names before building SQL', () => {
  assert.throws(
    () => createPgvectorMemoryBackend({ tableName: 'memory_embeddings;DROP TABLE users' }),
    /Invalid pgvector table name/,
  );
});

test('Memory V2 can expose pgvector as unavailable plugin without selecting it', async () => {
  const memory = createMemoryV2({
    logger: { warn() {} },
    backends: [
      createPgvectorMemoryBackend({ enabled: false }),
      {
        name: 'legacy-conversation-memory',
        async resolve() {
          return { memories: [{ label: 'fact', text: 'legacy survives' }] };
        },
      },
    ],
    env: {
      MEMORY_ENABLED: '1',
      MEMORY_BACKEND: 'pgvector',
      MEMORY_VECTOR_ENABLED: '1',
    },
  });

  const status = memory.getStatus();
  const result = await memory.resolve({ userId: 'user-1', query: 'memory-chain' });

  assert.equal(status.vectorEnabled, true);
  assert.equal(status.selectedBackend, 'legacy-conversation-memory');
  assert.equal(status.backends.find((item) => item.name === 'pgvector')?.available, false);
  assert.deepEqual(result.memories, [{ label: 'fact', text: 'legacy survives' }]);
});