docorbit 0.1.0 → 0.1.1

package/dist/packages/storage/src/repository.d.ts

@@ -0,0 +1,92 @@
+import type { DiscoveredSource, NormalizedPage, DocumentChunk, ChunkCode, SymbolReference, ChunkRelationship, ChunkType, ApiEndpoint, IndexedExample, Pitfall, PitfallKind } from '../../shared/src/index.ts';
+import type { DocOrbitDb } from './db.ts';
+import type { Snapshot, ISourceRepository, IPageRepository, IChunkRepository, IApiRepository, IExampleRepository, IPitfallRepository } from './interfaces.ts';
+import { SourceRepository } from './repositories/source-repository.ts';
+import { PageRepository } from './repositories/page-repository.ts';
+import { ChunkRepository } from './repositories/chunk-repository.ts';
+import { ApiRepository } from './repositories/api-repository.ts';
+import { ExampleRepository } from './repositories/example-repository.ts';
+import { PitfallRepository } from './repositories/pitfall-repository.ts';
+export type { Snapshot } from './interfaces.ts';
+/**
+ * Unified Facade for the DocOrbit database storage layer.
+ * Delegates domain responsibilities to specialized sub-repositories:
+ * - sources: ISourceRepository (sources, snapshots, snapshot-page links)
+ * - pages: IPageRepository (normalized pages, page links, page FTS)
+ * - chunks: IChunkRepository (document chunks, code, symbols, relationships, chunk FTS)
+ * - apis: IApiRepository (API endpoints, OpenAPI schemas, endpoint FTS)
+ * - examples: IExampleRepository (code examples, example FTS)
+ * - pitfalls: IPitfallRepository (pitfalls, gotchas, pitfall FTS)
+ */
+export declare class DocOrbitRepository implements ISourceRepository, IPageRepository, IChunkRepository, IApiRepository, IExampleRepository, IPitfallRepository {
+    readonly sources: SourceRepository;
+    readonly pages: PageRepository;
+    readonly chunks: ChunkRepository;
+    readonly apis: ApiRepository;
+    readonly examples: ExampleRepository;
+    readonly pitfalls: PitfallRepository;
+    readonly db: DocOrbitDb;
+    constructor(db: DocOrbitDb);
+    saveSource(source: DiscoveredSource): string;
+    getSource(id: string): DiscoveredSource | null;
+    getSourceByUrl(url: string): DiscoveredSource | null;
+    listSources(): DiscoveredSource[];
+    createSnapshot(sourceId: string, metadata?: Record<string, unknown>, docVersion?: string): string;
+    saveSnapshot(snapshot: Snapshot): void;
+    getSnapshot(id: string): Snapshot | null;
+    getLatestSnapshot(sourceId: string): Snapshot | null;
+    listSnapshots(sourceId: string): Snapshot[];
+    linkSnapshotPages(snapshotId: string, pageIds: string[]): void;
+    getSnapshotPages(snapshotId: string): NormalizedPage[];
+    savePage(page: NormalizedPage): void;
+    getPage(id: string): NormalizedPage | null;
+    getPageByUrl(normalizedUrl: string): NormalizedPage | null;
+    listPages(limit?: number, offset?: number): NormalizedPage[];
+    countPages(sourceId?: string): number;
+    savePageLinks(pageId: string, targetUrls: string[]): void;
+    getPageLinks(pageId: string): string[];
+    searchPagesFts(query: string, limit?: number): NormalizedPage[];
+    saveChunks(chunks: DocumentChunk[], relationships?: ChunkRelationship[], codeSnippets?: ChunkCode[], symbols?: SymbolReference[]): void;
+    getChunk(id: string): DocumentChunk | null;
+    getChunksByPage(pageId: string): DocumentChunk[];
+    getChunksBySnapshot(snapshotId: string): DocumentChunk[];
+    getChunksByVersion(docVersion: string): DocumentChunk[];
+    getChunkCode(chunkId: string): ChunkCode[];
+    getChunkSymbols(chunkId: string): SymbolReference[];
+    getChunkRelationships(chunkId: string): ChunkRelationship[];
+    countChunks(snapshotId?: string): number;
+    searchChunksFts(query: string, options?: {
+        limit?: number;
+        snapshotId?: string;
+        chunkType?: ChunkType;
+        docVersion?: string;
+    }): Array<{
+        chunk: DocumentChunk;
+        ftsRank: number;
+        symbols: SymbolReference[];
+        codeSnippets: ChunkCode[];
+    }>;
+    saveApiEndpoints(endpoints: ApiEndpoint[]): void;
+    getApiEndpoint(id: string): ApiEndpoint | null;
+    getApiEndpointsBySnapshot(snapshotId: string): ApiEndpoint[];
+    searchApiEndpoints(query: string, options?: {
+        method?: string;
+        docVersion?: string;
+        limit?: number;
+    }): ApiEndpoint[];
+    saveIndexedExamples(examples: IndexedExample[]): void;
+    getIndexedExamplesBySnapshot(snapshotId: string): IndexedExample[];
+    searchIndexedExamples(query: string, options?: {
+        language?: string;
+        framework?: string;
+        docVersion?: string;
+        limit?: number;
+    }): IndexedExample[];
+    savePitfalls(pitfalls: Pitfall[]): void;
+    getPitfallsBySnapshot(snapshotId: string): Pitfall[];
+    searchPitfalls(query: string, options?: {
+        kind?: PitfallKind;
+        docVersion?: string;
+        limit?: number;
+    }): Pitfall[];
+}

package/dist/packages/storage/src/repository.js

@@ -0,0 +1,157 @@
+import { SourceRepository } from "./repositories/source-repository.js";
+import { PageRepository } from "./repositories/page-repository.js";
+import { ChunkRepository } from "./repositories/chunk-repository.js";
+import { ApiRepository } from "./repositories/api-repository.js";
+import { ExampleRepository } from "./repositories/example-repository.js";
+import { PitfallRepository } from "./repositories/pitfall-repository.js";
+/**
+ * Unified Facade for the DocOrbit database storage layer.
+ * Delegates domain responsibilities to specialized sub-repositories:
+ * - sources: ISourceRepository (sources, snapshots, snapshot-page links)
+ * - pages: IPageRepository (normalized pages, page links, page FTS)
+ * - chunks: IChunkRepository (document chunks, code, symbols, relationships, chunk FTS)
+ * - apis: IApiRepository (API endpoints, OpenAPI schemas, endpoint FTS)
+ * - examples: IExampleRepository (code examples, example FTS)
+ * - pitfalls: IPitfallRepository (pitfalls, gotchas, pitfall FTS)
+ */
+export class DocOrbitRepository {
+    sources;
+    pages;
+    chunks;
+    apis;
+    examples;
+    pitfalls;
+    db;
+    constructor(db) {
+        this.db = db;
+        this.pages = new PageRepository(db);
+        this.sources = new SourceRepository(db, row => this.pages.hydratePage(row));
+        this.chunks = new ChunkRepository(db);
+        this.apis = new ApiRepository(db);
+        this.examples = new ExampleRepository(db);
+        this.pitfalls = new PitfallRepository(db);
+    }
+    // --- ISourceRepository Delegations ---
+    saveSource(source) {
+        return this.sources.saveSource(source);
+    }
+    getSource(id) {
+        return this.sources.getSource(id);
+    }
+    getSourceByUrl(url) {
+        return this.sources.getSourceByUrl(url);
+    }
+    listSources() {
+        return this.sources.listSources();
+    }
+    createSnapshot(sourceId, metadata, docVersion) {
+        return this.sources.createSnapshot(sourceId, metadata, docVersion);
+    }
+    saveSnapshot(snapshot) {
+        this.sources.saveSnapshot(snapshot);
+    }
+    getSnapshot(id) {
+        return this.sources.getSnapshot(id);
+    }
+    getLatestSnapshot(sourceId) {
+        return this.sources.getLatestSnapshot(sourceId);
+    }
+    listSnapshots(sourceId) {
+        return this.sources.listSnapshots(sourceId);
+    }
+    linkSnapshotPages(snapshotId, pageIds) {
+        this.sources.linkSnapshotPages(snapshotId, pageIds);
+    }
+    getSnapshotPages(snapshotId) {
+        return this.sources.getSnapshotPages(snapshotId);
+    }
+    // --- IPageRepository Delegations ---
+    savePage(page) {
+        this.pages.savePage(page);
+    }
+    getPage(id) {
+        return this.pages.getPage(id);
+    }
+    getPageByUrl(normalizedUrl) {
+        return this.pages.getPageByUrl(normalizedUrl);
+    }
+    listPages(limit, offset) {
+        return this.pages.listPages(limit, offset);
+    }
+    countPages(sourceId) {
+        return this.pages.countPages(sourceId);
+    }
+    savePageLinks(pageId, targetUrls) {
+        this.pages.savePageLinks(pageId, targetUrls);
+    }
+    getPageLinks(pageId) {
+        return this.pages.getPageLinks(pageId);
+    }
+    searchPagesFts(query, limit = 10) {
+        return this.pages.searchPagesFts(query, limit);
+    }
+    // --- IChunkRepository Delegations ---
+    saveChunks(chunks, relationships = [], codeSnippets = [], symbols = []) {
+        this.chunks.saveChunks(chunks, relationships, codeSnippets, symbols);
+    }
+    getChunk(id) {
+        return this.chunks.getChunk(id);
+    }
+    getChunksByPage(pageId) {
+        return this.chunks.getChunksByPage(pageId);
+    }
+    getChunksBySnapshot(snapshotId) {
+        return this.chunks.getChunksBySnapshot(snapshotId);
+    }
+    getChunksByVersion(docVersion) {
+        return this.chunks.getChunksByVersion(docVersion);
+    }
+    getChunkCode(chunkId) {
+        return this.chunks.getChunkCode(chunkId);
+    }
+    getChunkSymbols(chunkId) {
+        return this.chunks.getChunkSymbols(chunkId);
+    }
+    getChunkRelationships(chunkId) {
+        return this.chunks.getChunkRelationships(chunkId);
+    }
+    countChunks(snapshotId) {
+        return this.chunks.countChunks(snapshotId);
+    }
+    searchChunksFts(query, options = {}) {
+        return this.chunks.searchChunksFts(query, options);
+    }
+    // --- IApiRepository Delegations ---
+    saveApiEndpoints(endpoints) {
+        this.apis.saveApiEndpoints(endpoints);
+    }
+    getApiEndpoint(id) {
+        return this.apis.getApiEndpoint(id);
+    }
+    getApiEndpointsBySnapshot(snapshotId) {
+        return this.apis.getApiEndpointsBySnapshot(snapshotId);
+    }
+    searchApiEndpoints(query, options = {}) {
+        return this.apis.searchApiEndpoints(query, options);
+    }
+    // --- IExampleRepository Delegations ---
+    saveIndexedExamples(examples) {
+        this.examples.saveIndexedExamples(examples);
+    }
+    getIndexedExamplesBySnapshot(snapshotId) {
+        return this.examples.getIndexedExamplesBySnapshot(snapshotId);
+    }
+    searchIndexedExamples(query, options = {}) {
+        return this.examples.searchIndexedExamples(query, options);
+    }
+    // --- IPitfallRepository Delegations ---
+    savePitfalls(pitfalls) {
+        this.pitfalls.savePitfalls(pitfalls);
+    }
+    getPitfallsBySnapshot(snapshotId) {
+        return this.pitfalls.getPitfallsBySnapshot(snapshotId);
+    }
+    searchPitfalls(query, options = {}) {
+        return this.pitfalls.searchPitfalls(query, options);
+    }
+}

package/dist/packages/storage/src/schema.d.ts

@@ -0,0 +1,2 @@
+export declare const SCHEMA_SQL = "\nCREATE TABLE IF NOT EXISTS sources (\n  id TEXT PRIMARY KEY,\n  url TEXT UNIQUE NOT NULL,\n  type TEXT NOT NULL,\n  discovered_by TEXT NOT NULL,\n  status TEXT NOT NULL,\n  content_type TEXT,\n  confidence REAL NOT NULL,\n  authority TEXT NOT NULL,\n  machine_readable INTEGER NOT NULL,\n  metadata TEXT,\n  created_at TEXT NOT NULL\n);\n\nCREATE TABLE IF NOT EXISTS pages (\n  id TEXT PRIMARY KEY,\n  source_id TEXT NOT NULL,\n  title TEXT NOT NULL,\n  url TEXT UNIQUE NOT NULL,\n  content TEXT NOT NULL,\n  content_hash TEXT NOT NULL,\n  fetched_at TEXT NOT NULL,\n  raw_bytes INTEGER NOT NULL,\n  estimated_tokens INTEGER NOT NULL,\n  headings_json TEXT,\n  security_annotations_json TEXT,\n  provenance_json TEXT,\n  FOREIGN KEY (source_id) REFERENCES sources(id) ON DELETE CASCADE\n);\n\nCREATE TABLE IF NOT EXISTS links (\n  id TEXT PRIMARY KEY,\n  page_id TEXT NOT NULL,\n  text TEXT NOT NULL,\n  url TEXT NOT NULL,\n  is_external INTEGER NOT NULL,\n  FOREIGN KEY (page_id) REFERENCES pages(id) ON DELETE CASCADE\n);\n\nCREATE TABLE IF NOT EXISTS code_examples (\n  id TEXT PRIMARY KEY,\n  page_id TEXT NOT NULL,\n  language TEXT NOT NULL,\n  code TEXT NOT NULL,\n  caption TEXT,\n  FOREIGN KEY (page_id) REFERENCES pages(id) ON DELETE CASCADE\n);\n\nCREATE TABLE IF NOT EXISTS snapshots (\n  id TEXT PRIMARY KEY,\n  source_id TEXT NOT NULL,\n  snapshot_hash TEXT NOT NULL,\n  page_count INTEGER NOT NULL,\n  captured_at TEXT NOT NULL,\n  doc_version TEXT,\n  metadata TEXT,\n  FOREIGN KEY (source_id) REFERENCES sources(id) ON DELETE CASCADE\n);\n\nCREATE TABLE IF NOT EXISTS snapshot_pages (\n  snapshot_id TEXT NOT NULL,\n  page_id TEXT NOT NULL,\n  content_hash TEXT NOT NULL,\n  PRIMARY KEY (snapshot_id, page_id),\n  FOREIGN KEY (snapshot_id) REFERENCES snapshots(id) ON DELETE CASCADE,\n  FOREIGN KEY (page_id) REFERENCES pages(id) ON DELETE CASCADE\n);\n\nCREATE TABLE IF NOT EXISTS chunks (\n  id TEXT PRIMARY KEY,\n  page_id TEXT NOT NULL,\n  snapshot_id TEXT NOT NULL,\n  title TEXT,\n  section_path_json TEXT NOT NULL,\n  content TEXT NOT NULL,\n  chunk_type TEXT NOT NULL,\n  language TEXT,\n  token_estimate INTEGER NOT NULL,\n  ordinal INTEGER NOT NULL,\n  content_hash TEXT NOT NULL,\n  doc_version TEXT,\n  provenance_json TEXT,\n  created_at TEXT NOT NULL,\n  FOREIGN KEY (page_id) REFERENCES pages(id) ON DELETE CASCADE\n);\n\nCREATE TABLE IF NOT EXISTS chunk_code (\n  id TEXT PRIMARY KEY,\n  chunk_id TEXT NOT NULL,\n  language TEXT,\n  code TEXT NOT NULL,\n  FOREIGN KEY (chunk_id) REFERENCES chunks(id) ON DELETE CASCADE\n);\n\nCREATE TABLE IF NOT EXISTS symbol_references (\n  id TEXT PRIMARY KEY,\n  chunk_id TEXT NOT NULL,\n  name TEXT NOT NULL,\n  kind TEXT NOT NULL,\n  FOREIGN KEY (chunk_id) REFERENCES chunks(id) ON DELETE CASCADE\n);\n\nCREATE TABLE IF NOT EXISTS chunk_relationships (\n  source_chunk_id TEXT NOT NULL,\n  target_chunk_id TEXT NOT NULL,\n  type TEXT NOT NULL,\n  PRIMARY KEY (source_chunk_id, target_chunk_id, type)\n);\n\nCREATE INDEX IF NOT EXISTS idx_pages_source_id ON pages(source_id);\nCREATE INDEX IF NOT EXISTS idx_pages_content_hash ON pages(content_hash);\nCREATE INDEX IF NOT EXISTS idx_links_page_id ON links(page_id);\nCREATE INDEX IF NOT EXISTS idx_code_examples_page_id ON code_examples(page_id);\nCREATE INDEX IF NOT EXISTS idx_snapshots_source_id ON snapshots(source_id);\nCREATE INDEX IF NOT EXISTS idx_snapshots_doc_version ON snapshots(doc_version);\nCREATE INDEX IF NOT EXISTS idx_snapshot_pages_snapshot_id ON snapshot_pages(snapshot_id);\nCREATE INDEX IF NOT EXISTS idx_snapshot_pages_page_id ON snapshot_pages(page_id);\nCREATE INDEX IF NOT EXISTS idx_chunks_page_id ON chunks(page_id);\nCREATE INDEX IF NOT EXISTS idx_chunks_snapshot_id ON chunks(snapshot_id);\nCREATE INDEX IF NOT EXISTS idx_chunks_chunk_type ON chunks(chunk_type);\nCREATE INDEX IF NOT EXISTS idx_chunks_content_hash ON chunks(content_hash);\nCREATE INDEX IF NOT EXISTS idx_chunks_doc_version ON chunks(doc_version);\nCREATE INDEX IF NOT EXISTS idx_chunk_code_chunk_id ON chunk_code(chunk_id);\nCREATE INDEX IF NOT EXISTS idx_symbol_references_chunk_id ON symbol_references(chunk_id);\nCREATE INDEX IF NOT EXISTS idx_symbol_references_name ON symbol_references(name);\nCREATE INDEX IF NOT EXISTS idx_chunk_relationships_source ON chunk_relationships(source_chunk_id);\nCREATE INDEX IF NOT EXISTS idx_chunk_relationships_target ON chunk_relationships(target_chunk_id);\n\nCREATE TABLE IF NOT EXISTS api_endpoints (\n  id TEXT PRIMARY KEY,\n  page_id TEXT NOT NULL,\n  snapshot_id TEXT NOT NULL,\n  method TEXT NOT NULL,\n  path TEXT NOT NULL,\n  summary TEXT,\n  description TEXT,\n  operation_id TEXT,\n  parameters_json TEXT NOT NULL,\n  request_schema_json TEXT,\n  response_schema_json TEXT,\n  auth_json TEXT NOT NULL,\n  errors_json TEXT NOT NULL,\n  pagination_json TEXT,\n  deprecated INTEGER NOT NULL,\n  doc_version TEXT,\n  provenance_json TEXT,\n  created_at TEXT NOT NULL,\n  FOREIGN KEY (page_id) REFERENCES pages(id) ON DELETE CASCADE\n);\n\nCREATE INDEX IF NOT EXISTS idx_api_endpoints_page_id ON api_endpoints(page_id);\nCREATE INDEX IF NOT EXISTS idx_api_endpoints_snapshot_id ON api_endpoints(snapshot_id);\nCREATE INDEX IF NOT EXISTS idx_api_endpoints_method ON api_endpoints(method);\nCREATE INDEX IF NOT EXISTS idx_api_endpoints_path ON api_endpoints(path);\nCREATE INDEX IF NOT EXISTS idx_api_endpoints_doc_version ON api_endpoints(doc_version);\n\nCREATE TABLE IF NOT EXISTS indexed_examples (\n  id TEXT PRIMARY KEY,\n  chunk_id TEXT,\n  page_id TEXT NOT NULL,\n  snapshot_id TEXT NOT NULL,\n  language TEXT NOT NULL,\n  framework TEXT,\n  task TEXT NOT NULL,\n  code TEXT NOT NULL,\n  source_url TEXT NOT NULL,\n  source_authority TEXT NOT NULL,\n  related_api TEXT,\n  related_symbol TEXT,\n  doc_version TEXT,\n  provenance_json TEXT,\n  created_at TEXT NOT NULL,\n  FOREIGN KEY (page_id) REFERENCES pages(id) ON DELETE CASCADE\n);\n\nCREATE INDEX IF NOT EXISTS idx_indexed_examples_page_id ON indexed_examples(page_id);\nCREATE INDEX IF NOT EXISTS idx_indexed_examples_snapshot_id ON indexed_examples(snapshot_id);\nCREATE INDEX IF NOT EXISTS idx_indexed_examples_chunk_id ON indexed_examples(chunk_id);\nCREATE INDEX IF NOT EXISTS idx_indexed_examples_language ON indexed_examples(language);\nCREATE INDEX IF NOT EXISTS idx_indexed_examples_framework ON indexed_examples(framework);\nCREATE INDEX IF NOT EXISTS idx_indexed_examples_doc_version ON indexed_examples(doc_version);\nCREATE INDEX IF NOT EXISTS idx_indexed_examples_related_api ON indexed_examples(related_api);\n\nCREATE TABLE IF NOT EXISTS pitfalls (\n  id TEXT PRIMARY KEY,\n  chunk_id TEXT,\n  page_id TEXT NOT NULL,\n  snapshot_id TEXT NOT NULL,\n  kind TEXT NOT NULL,\n  title TEXT NOT NULL,\n  content TEXT NOT NULL,\n  related_api TEXT,\n  related_symbol TEXT,\n  doc_version TEXT,\n  provenance_json TEXT,\n  created_at TEXT NOT NULL,\n  FOREIGN KEY (page_id) REFERENCES pages(id) ON DELETE CASCADE\n);\n\nCREATE INDEX IF NOT EXISTS idx_pitfalls_page_id ON pitfalls(page_id);\nCREATE INDEX IF NOT EXISTS idx_pitfalls_snapshot_id ON pitfalls(snapshot_id);\nCREATE INDEX IF NOT EXISTS idx_pitfalls_chunk_id ON pitfalls(chunk_id);\nCREATE INDEX IF NOT EXISTS idx_pitfalls_kind ON pitfalls(kind);\nCREATE INDEX IF NOT EXISTS idx_pitfalls_doc_version ON pitfalls(doc_version);\nCREATE INDEX IF NOT EXISTS idx_pitfalls_related_api ON pitfalls(related_api);\n";
+export declare const FTS_SCHEMA_SQL = "\nCREATE VIRTUAL TABLE IF NOT EXISTS pages_fts USING fts5(\n  page_id UNINDEXED,\n  title,\n  content\n);\n\nCREATE VIRTUAL TABLE IF NOT EXISTS chunks_fts USING fts5(\n  chunk_id UNINDEXED,\n  title,\n  section_path,\n  content,\n  symbols\n);\n\nCREATE TRIGGER IF NOT EXISTS trg_chunks_delete AFTER DELETE ON chunks BEGIN\n  DELETE FROM chunks_fts WHERE chunk_id = old.id;\nEND;\n\nCREATE VIRTUAL TABLE IF NOT EXISTS api_endpoints_fts USING fts5(\n  endpoint_id UNINDEXED,\n  method,\n  path,\n  summary,\n  description,\n  parameters\n);\n\nCREATE TRIGGER IF NOT EXISTS trg_api_endpoints_delete AFTER DELETE ON api_endpoints BEGIN\n  DELETE FROM api_endpoints_fts WHERE endpoint_id = old.id;\nEND;\n\nCREATE VIRTUAL TABLE IF NOT EXISTS indexed_examples_fts USING fts5(\n  example_id UNINDEXED,\n  task,\n  language,\n  framework,\n  code,\n  related_api,\n  related_symbol\n);\n\nCREATE TRIGGER IF NOT EXISTS trg_indexed_examples_delete AFTER DELETE ON indexed_examples BEGIN\n  DELETE FROM indexed_examples_fts WHERE example_id = old.id;\nEND;\n\nCREATE VIRTUAL TABLE IF NOT EXISTS pitfalls_fts USING fts5(\n  pitfall_id UNINDEXED,\n  kind,\n  title,\n  content,\n  related_api,\n  related_symbol\n);\n\nCREATE TRIGGER IF NOT EXISTS trg_pitfalls_delete AFTER DELETE ON pitfalls BEGIN\n  DELETE FROM pitfalls_fts WHERE pitfall_id = old.id;\nEND;\n";

package/{packages/storage/src/schema.ts → dist/packages/storage/src/schema.js}

@@ -205,7 +205,6 @@ CREATE INDEX IF NOT EXISTS idx_pitfalls_kind ON pitfalls(kind);
 CREATE INDEX IF NOT EXISTS idx_pitfalls_doc_version ON pitfalls(doc_version);
 CREATE INDEX IF NOT EXISTS idx_pitfalls_related_api ON pitfalls(related_api);
 `;
-
 export const FTS_SCHEMA_SQL = `
 CREATE VIRTUAL TABLE IF NOT EXISTS pages_fts USING fts5(
   page_id UNINDEXED,
@@ -265,5 +264,3 @@ CREATE TRIGGER IF NOT EXISTS trg_pitfalls_delete AFTER DELETE ON pitfalls BEGIN
   DELETE FROM pitfalls_fts WHERE pitfall_id = old.id;
 END;
 `;
-
-

package/dist/packages/storage/src/search-tokens.d.ts

@@ -0,0 +1,6 @@
+export declare const SEARCH_STOPWORDS: Set<string>;
+/**
+ * Extracts substantive tokens from query text, filtering out common stopwords
+ * and short words (< 3 chars), while falling back to short words if query has no substantive words.
+ */
+export declare function extractSearchTokens(text: string): string[];

package/dist/packages/storage/src/search-tokens.js

@@ -0,0 +1,26 @@
+export const SEARCH_STOPWORDS = new Set([
+    'a', 'about', 'above', 'after', 'again', 'against', 'all', 'am', 'an', 'and', 'any', 'are',
+    'as', 'at', 'be', 'because', 'been', 'before', 'being', 'below', 'between', 'both', 'but', 'by',
+    'can', 'did', 'do', 'does', 'doing', 'down', 'during', 'each', 'few', 'for', 'from',
+    'further', 'had', 'has', 'have', 'having', 'he', 'her', 'here', 'hers', 'herself', 'him',
+    'himself', 'his', 'how', 'i', 'if', 'in', 'into', 'is', 'it', 'its', 'itself', 'just',
+    'me', 'more', 'most', 'my', 'myself', 'no', 'nor', 'not', 'now', 'of', 'off', 'on',
+    'once', 'only', 'or', 'other', 'our', 'ours', 'ourselves', 'out', 'over', 'own', 'same',
+    'she', 'should', 'so', 'some', 'such', 'than', 'that', 'the', 'their', 'theirs', 'them',
+    'themselves', 'then', 'there', 'these', 'they', 'this', 'those', 'through', 'to', 'too',
+    'under', 'until', 'up', 'very', 'was', 'we', 'were', 'what', 'when', 'where', 'which',
+    'while', 'who', 'whom', 'why', 'with', 'you', 'your', 'yours', 'yourself', 'yourselves'
+]);
+/**
+ * Extracts substantive tokens from query text, filtering out common stopwords
+ * and short words (< 3 chars), while falling back to short words if query has no substantive words.
+ */
+export function extractSearchTokens(text) {
+    const words = text
+        .replace(/[^\w\s-]/g, ' ')
+        .trim()
+        .split(/\s+/)
+        .filter(w => w.length > 0);
+    const substantive = words.filter(w => w.length >= 3 && !SEARCH_STOPWORDS.has(w.toLowerCase()));
+    return substantive.length > 0 ? substantive : words.filter(w => w.length >= 2);
+}

package/dist/packages/verification/src/diff-engine.d.ts

@@ -0,0 +1,18 @@
+import type { DocDiffResult } from '../../shared/src/index.ts';
+import type { DocOrbitRepository } from '../../storage/src/index.ts';
+export interface DiffOptions {
+    fromVersion?: string;
+    toVersion?: string;
+    fromSnapshotId?: string;
+    toSnapshotId?: string;
+    sourceId?: string;
+}
+export declare class DocDiffEngine {
+    private repo;
+    constructor(repo: DocOrbitRepository);
+    diff(options: DiffOptions): DocDiffResult;
+    private getEndpoints;
+    private getPitfalls;
+    private detectEndpointModifications;
+    private computeContentDiffs;
+}

package/dist/packages/verification/src/diff-engine.js

@@ -0,0 +1,218 @@
+export class DocDiffEngine {
+    repo;
+    constructor(repo) {
+        this.repo = repo;
+    }
+    diff(options) {
+        const fromEndpoints = this.getEndpoints(options.fromSnapshotId, options.fromVersion);
+        const toEndpoints = this.getEndpoints(options.toSnapshotId, options.toVersion);
+        const fromPitfalls = this.getPitfalls(options.fromSnapshotId, options.fromVersion);
+        const toPitfalls = this.getPitfalls(options.toSnapshotId, options.toVersion);
+        // 1. Compute Endpoint Diffs
+        const apiChanges = [];
+        const fromMap = new Map();
+        const toMap = new Map();
+        for (const ep of fromEndpoints) {
+            fromMap.set(`${ep.method.toUpperCase()} ${ep.path}`, ep);
+        }
+        for (const ep of toEndpoints) {
+            toMap.set(`${ep.method.toUpperCase()} ${ep.path}`, ep);
+        }
+        let addedCount = 0;
+        let removedCount = 0;
+        let modifiedCount = 0;
+        let deprecatedCount = 0;
+        // Check additions, deprecations, and modifications in 'to'
+        for (const [key, toEp] of toMap.entries()) {
+            const fromEp = fromMap.get(key);
+            if (!fromEp) {
+                // Newly added endpoint
+                apiChanges.push({
+                    path: toEp.path,
+                    method: toEp.method,
+                    changeType: 'added',
+                    endpoint: toEp,
+                });
+                addedCount++;
+            }
+            else {
+                // Existed previously: check deprecation status
+                if (toEp.deprecated && !fromEp.deprecated) {
+                    apiChanges.push({
+                        path: toEp.path,
+                        method: toEp.method,
+                        changeType: 'deprecated',
+                        endpoint: toEp,
+                        previousEndpoint: fromEp,
+                    });
+                    deprecatedCount++;
+                }
+                // Check semantic parameter and schema modifications
+                const changes = this.detectEndpointModifications(fromEp, toEp);
+                if (changes.length > 0) {
+                    apiChanges.push({
+                        path: toEp.path,
+                        method: toEp.method,
+                        changeType: 'modified',
+                        changes,
+                        endpoint: toEp,
+                        previousEndpoint: fromEp,
+                    });
+                    modifiedCount++;
+                }
+            }
+        }
+        // Check removals from 'from'
+        for (const [key, fromEp] of fromMap.entries()) {
+            if (!toMap.has(key)) {
+                apiChanges.push({
+                    path: fromEp.path,
+                    method: fromEp.method,
+                    changeType: 'removed',
+                    endpoint: fromEp,
+                });
+                removedCount++;
+            }
+        }
+        // 2. Compute Pitfall Diffs
+        const pitfallChanges = [];
+        const fromPitfallTitles = new Set(fromPitfalls.map(p => `${p.kind}:${p.title}`));
+        const toPitfallTitles = new Set(toPitfalls.map(p => `${p.kind}:${p.title}`));
+        for (const toP of toPitfalls) {
+            if (!fromPitfallTitles.has(`${toP.kind}:${toP.title}`)) {
+                pitfallChanges.push({ changeType: 'added', pitfall: toP });
+            }
+        }
+        for (const fromP of fromPitfalls) {
+            if (!toPitfallTitles.has(`${fromP.kind}:${fromP.title}`)) {
+                pitfallChanges.push({ changeType: 'removed', pitfall: fromP });
+            }
+        }
+        // 3. Compute Content/Chunk Diffs (Semantic only, ignoring formatting-only whitespace)
+        const contentChanges = this.computeContentDiffs(options);
+        return {
+            fromVersion: options.fromVersion,
+            toVersion: options.toVersion,
+            fromSnapshotId: options.fromSnapshotId,
+            toSnapshotId: options.toSnapshotId,
+            sourceId: options.sourceId,
+            summary: {
+                endpointsAdded: addedCount,
+                endpointsRemoved: removedCount,
+                endpointsModified: modifiedCount,
+                endpointsDeprecated: deprecatedCount,
+                pitfallsAdded: pitfallChanges.filter(p => p.changeType === 'added').length,
+                pitfallsRemoved: pitfallChanges.filter(p => p.changeType === 'removed').length,
+                contentChanged: contentChanges.length,
+            },
+            apiChanges,
+            pitfallChanges,
+            contentChanges,
+            retrievedAt: new Date().toISOString(),
+        };
+    }
+    getEndpoints(snapshotId, docVersion) {
+        if (snapshotId) {
+            return this.repo.getApiEndpointsBySnapshot(snapshotId);
+        }
+        return this.repo.searchApiEndpoints('', {
+            docVersion,
+            limit: 500,
+        });
+    }
+    getPitfalls(snapshotId, docVersion) {
+        if (snapshotId) {
+            return this.repo.getPitfallsBySnapshot(snapshotId);
+        }
+        return this.repo.searchPitfalls('', {
+            docVersion,
+            limit: 200,
+        });
+    }
+    detectEndpointModifications(fromEp, toEp) {
+        const changes = [];
+        // Compare parameters
+        const fromParams = new Map(fromEp.parameters.map(p => [p.name, p]));
+        const toParams = new Map(toEp.parameters.map(p => [p.name, p]));
+        for (const [name, toP] of toParams.entries()) {
+            const fromP = fromParams.get(name);
+            if (!fromP) {
+                changes.push(`Added parameter "${name}"${toP.required ? ' (required)' : ''}`);
+            }
+            else if (!fromP.required && toP.required) {
+                changes.push(`Parameter "${name}" is now required (was optional)`);
+            }
+            else if (fromP.type !== toP.type && toP.type) {
+                changes.push(`Parameter "${name}" type changed from ${fromP.type || 'any'} to ${toP.type}`);
+            }
+        }
+        for (const name of fromParams.keys()) {
+            if (!toParams.has(name)) {
+                changes.push(`Removed parameter "${name}"`);
+            }
+        }
+        // Compare Request Body required fields
+        const fromReq = new Set(fromEp.requestSchema?.required || []);
+        const toReq = new Set(toEp.requestSchema?.required || []);
+        for (const field of toReq) {
+            if (!fromReq.has(field)) {
+                changes.push(`Request body field "${field}" is now required`);
+            }
+        }
+        for (const field of fromReq) {
+            if (!toReq.has(field)) {
+                changes.push(`Request body field "${field}" is no longer required`);
+            }
+        }
+        return changes;
+    }
+    computeContentDiffs(options) {
+        const fromChunks = options.fromSnapshotId
+            ? this.repo.getChunksBySnapshot(options.fromSnapshotId)
+            : options.fromVersion
+                ? this.repo.getChunksByVersion(options.fromVersion)
+                : [];
+        const toChunks = options.toSnapshotId
+            ? this.repo.getChunksBySnapshot(options.toSnapshotId)
+            : options.toVersion
+                ? this.repo.getChunksByVersion(options.toVersion)
+                : [];
+        if (fromChunks.length === 0 && toChunks.length === 0)
+            return [];
+        const changes = [];
+        const fromMap = new Map(fromChunks.map(c => [c.title, c]));
+        const toMap = new Map(toChunks.map(c => [c.title, c]));
+        for (const [title, toC] of toMap.entries()) {
+            const fromC = fromMap.get(title);
+            if (!fromC) {
+                changes.push({
+                    title,
+                    changeType: 'added',
+                    summary: `New documentation section: "${title}"`,
+                });
+            }
+            else if (fromC.contentHash !== toC.contentHash) {
+                // Compare normalized text to ignore formatting-only whitespace
+                const normFrom = fromC.content.replace(/\s+/g, ' ').trim();
+                const normTo = toC.content.replace(/\s+/g, ' ').trim();
+                if (normFrom !== normTo) {
+                    changes.push({
+                        title,
+                        changeType: 'modified',
+                        summary: `Content updated for section "${title}"`,
+                    });
+                }
+            }
+        }
+        for (const title of fromMap.keys()) {
+            if (!toMap.has(title)) {
+                changes.push({
+                    title,
+                    changeType: 'removed',
+                    summary: `Documentation section removed: "${title}"`,
+                });
+            }
+        }
+        return changes;
+    }
+}

package/dist/packages/verification/src/extractor.d.ts

@@ -0,0 +1,29 @@
+import type { ExtractedApiCall } from '../../shared/src/index.ts';
+/**
+ * Normalizes an API path for comparison against schema paths.
+ * E.g. extracts pathname from full URLs ("https://api.stripe.com/v1/webhook_endpoints" -> "/v1/webhook_endpoints")
+ */
+export declare function normalizeApiPath(urlOrPath: string): string;
+/**
+ * Checks if an extracted path matches an OpenAPI schema path template.
+ * E.g. candidate "/v1/webhook_endpoints/we_123" matches schema "/v1/webhook_endpoints/{id}"
+ */
+export declare function matchEndpointPath(candidatePath: string, schemaPath: string): boolean;
+/**
+ * Deterministically extracts API calls, parameters, methods, and version syntax from code.
+ */
+export declare class CodeApiExtractor {
+    /**
+     * Extracts API calls from code.
+     */
+    extract(code: string, language?: string): ExtractedApiCall[];
+    private findLineNumber;
+    private extractNextJsParams;
+    private extractFetchCalls;
+    private extractAxiosCalls;
+    private extractPythonRequestsCalls;
+    private extractCurlCalls;
+    private extractSdkCalls;
+    private extractDynamicCalls;
+    private extractResponseFields;
+}