ktx/packages/cli/test/context/ingest/local-embedding-provider.integration.test.ts

163 lines
5.5 KiB
TypeScript
Raw Permalink Normal View History

2026-05-10 23:12:26 +02:00
import { mkdir, mkdtemp, rm, writeFile } from 'node:fs/promises';
import { tmpdir } from 'node:os';
import { join } from 'node:path';
import { afterEach, beforeEach, describe, expect, it } from 'vitest';
test: split cli tests from source tree (#216) * feat(cli): define full warehouse dialect contract * test(cli): keep dialect edge tests focused * fix(cli): stabilize dialect contract foundation * refactor(connectors): own read-only query preparation * refactor(connectors): resolve dialects through registry * refactor(connectors): keep concrete dialect classes internal * chore(workspace): enforce dialect import boundary * refactor(cli): resolve relationship dialect at scan boundary * refactor(cli): use dialect display parsing for entity details * refactor(cli): use dialect display parsing for warehouse catalog * refactor(cli): use dialect SQL in relationship workflows * test(cli): verify solid dialect scan workflow closure * test: split cli tests from source tree * refactor(cli): standardize BigQuery scope listing * feat(sqlite): implement connector scope listing * test(connectors): cover required table listing * feat(cli): add warehouse driver registry * refactor(setup): route scope discovery through driver registry * refactor(cli): route local query execution through driver registry * refactor(historic-sql): route dialect support through driver registry * refactor(cli): test warehouse connections through driver registry * fix(cli): close driver registry type export gaps * Improve setup daemon diagnostics * refactor(setup): centralize rail-prefixed diagnostics + query-history fallback Extract errorMessage, writePrefixedLines, and flushPrefixedBufferedCommandOutput into clack.ts so the setup wizard, managed daemons, and embedding/agent steps share one rail-formatted writer. setup-databases.ts also adds a "disable query history and retry" option when the schema-context build fails and query history is the likely culprit, surfaced via a new failed-query-history-unavailable status. * fix(cli): carry catalog through the picker so BigQuery/Snowflake/SQL Server scope filters match The setup picker's KtxTableListEntry was a 2-level { schema, name }, so qualifiedTableId always wrote db.name into enabled_tables. When BigQuery, Snowflake, or SQL Server later ran fast ingest, their introspect step filtered the scope set with scopedTableNames(scope, { catalog: projectId|database, db }) — catalog was non-null on the introspect side but null in the scope refs, so every entry was rejected, the live-database adapter staged zero table files, and detect() failed with 'Adapter "live-database" did not recognize fetched source output'. Align the picker boundary with the canonical 3-level KtxTableRef: - Add catalog: string | null to KtxTableListEntry. - BigQuery/Snowflake/SQL Server listTables populate catalog from the resolved projectId / database; Postgres/MySQL/ClickHouse/SQLite set null. - qualifiedTableId emits catalog.schema.name when catalog is non-null (resolveEnabledTables already accepts the 3-part shape) and schemasFromEnabledTables now goes through parseDottedTableEntry so it recovers the schema correctly from both 2-part and 3-part entries. - Export parseDottedTableEntry from enabled-tables.ts (@internal) for picker reuse. Update listTables expectations in all seven connector tests and the setup / picker test fixtures. Add a picker regression test that covers the catalog-bearing round-trip (save + refine). * fix(cli): allow debug telemetry under opt-out env
2026-05-26 08:49:05 +02:00
import type { KtxEmbeddingPort } from '../../../src/context/core/embedding.js';
import { CandidateDedupService } from '../../../src/context/ingest/context-candidates/candidate-dedup.service.js';
import { ContextEvidenceIndexService } from '../../../src/context/ingest/context-evidence/context-evidence-index.service.js';
import { SqliteContextEvidenceStore } from '../../../src/context/ingest/context-evidence/sqlite-context-evidence-store.js';
import type { DiffSet } from '../../../src/context/ingest/types.js';
2026-05-10 23:12:26 +02:00
describe('local ingest embedding providers with SQLite ingest stores', () => {
let tempDir: string;
let dbPath: string;
let stagedDir: string;
beforeEach(async () => {
2026-05-10 23:51:24 +02:00
tempDir = await mkdtemp(join(tmpdir(), 'ktx-local-ingest-embedding-'));
dbPath = join(tempDir, '.ktx', 'db.sqlite');
2026-05-10 23:12:26 +02:00
stagedDir = join(tempDir, 'staged');
await mkdir(join(stagedDir, 'pages', 'revenue'), { recursive: true });
await writeFile(
join(stagedDir, 'pages', 'revenue', 'metadata.json'),
`${JSON.stringify({
objectType: 'page',
id: 'page-revenue',
title: 'Revenue Policy',
path: 'Revenue Policy',
url: 'https://notion.test/revenue',
parentId: null,
lastEditedAt: '2026-04-30T12:00:00.000Z',
properties: {},
})}\n`,
'utf8',
);
await writeFile(
join(stagedDir, 'pages', 'revenue', 'page.md'),
['# Approval', '', 'Owner approval is required before enterprise discounts are granted.', ''].join('\n'),
'utf8',
);
});
afterEach(async () => {
await rm(tempDir, { recursive: true, force: true });
});
function embeddings(): KtxEmbeddingPort {
return {
maxBatchSize: 4,
async computeEmbedding() {
return [1, 0, 0];
},
async computeEmbeddingsBulk(texts) {
return texts.map(() => [1, 0, 0]);
},
};
2026-05-10 23:12:26 +02:00
}
it('indexes and searches context evidence using a package-owned local embedding provider', async () => {
const store = new SqliteContextEvidenceStore({ dbPath });
const embeddingPort = embeddings();
const indexer = new ContextEvidenceIndexService({ store, embeddings: embeddingPort });
const diffSet: DiffSet = {
added: ['pages/revenue/metadata.json', 'pages/revenue/page.md'],
modified: [],
deleted: [],
unchanged: [],
};
const summary = await indexer.indexStagedDir({
stagedDir,
runId: 'run-1',
connectionId: 'docs',
sourceKey: 'notion',
syncId: 'sync-1',
diffSet,
currentHashes: new Map([
['pages/revenue/metadata.json', 'metadata-hash'],
['pages/revenue/page.md', 'page-hash'],
]),
});
expect(summary).toMatchObject({
documentsIndexed: 1,
embeddingFailures: 0,
});
expect(summary.chunksIndexed).toBeGreaterThan(0);
const queryText = [
'Revenue Policy',
'Revenue Policy',
'Approval',
'Owner approval is required before enterprise discounts are granted.',
].join('\n');
const queryEmbedding = await embeddingPort.computeEmbedding(queryText);
const results = await store.searchRRF({
connectionId: 'docs',
sourceKey: 'notion',
queryEmbedding,
queryText,
limit: 5,
includeDeleted: false,
currentRunId: 'run-1',
});
expect(results[0]?.title).toBe('Revenue Policy');
expect(results[0]?.stableCitationKey).toContain('notion:page-revenue');
expect(results[0]).toMatchObject({
matchReasons: expect.arrayContaining(['semantic']),
lanes: expect.arrayContaining([
expect.objectContaining({ lane: 'semantic', status: 'available' }),
expect.objectContaining({ lane: 'lexical', status: 'available' }),
expect.objectContaining({ lane: 'token', status: 'available' }),
]),
});
});
it('deduplicates candidates using package-owned local embeddings and SQLite persistence', async () => {
const store = new SqliteContextEvidenceStore({ dbPath });
const embeddingPort = embeddings();
const candidateBase = {
runId: 'run-1',
connectionId: 'docs',
sourceKey: 'notion',
topic: 'Enterprise discount approval',
assertion: 'Owner approval is required before enterprise discounts are granted.',
rationale: 'The source policy states that approval is required.',
evidenceChunkIds: [],
evidenceRefs: [],
suggestedPageKey: 'revenue-policy',
actionHint: 'create' as const,
durabilityScore: 3,
authorityScore: 3,
reuseScore: 3,
noveltyScore: 2,
riskScore: 0,
promotionScore: 11,
status: 'pending' as const,
rejectionReason: null,
lane: 'full' as const,
embedding: null,
};
await store.insertCandidate({ ...candidateBase, candidateKey: 'discount-policy-a' });
await store.insertCandidate({ ...candidateBase, candidateKey: 'discount-policy-b' });
const result = await new CandidateDedupService({
store,
embeddings: embeddingPort,
settings: {
enabled: true,
topicSimilarityThreshold: -1,
scoreAggregation: 'max',
},
}).deduplicateRun('run-1');
expect(result.enabled).toBe(true);
expect(result.embeddingFailures).toBe(0);
expect(result.candidatesIn).toBe(2);
expect(result.clustersOut).toBe(1);
expect(result.mergedCount).toBe(1);
});
});