feat(tasks): Refactor task workflow with payload/refresh separation

Major changes: - Split crawl into payload_fetch (API → disk) and product_refresh (disk → DB) - Add task chaining: store_discovery → product_discovery → payload_fetch → product_refresh - Add payload storage utilities for gzipped JSON on filesystem - Add /api/payloads endpoints for payload access and diffing - Add DB-driven TaskScheduler with schedule persistence - Track newDispensaryIds through discovery promotion for chaining - Add stealth improvements: HTTP fingerprinting, proxy rotation enhancements - Add Workers dashboard K8s scaling controls New files: - src/tasks/handlers/payload-fetch.ts - Fetches from API, saves to disk - src/services/task-scheduler.ts - DB-driven schedule management - src/utils/payload-storage.ts - Payload save/load utilities - src/routes/payloads.ts - Payload API endpoints - src/services/http-fingerprint.ts - Browser fingerprint generation - docs/TASK_WORKFLOW_2024-12-10.md - Complete workflow documentation Migrations: - 078: Proxy consecutive 403 tracking - 079: task_schedules table - 080: raw_crawl_payloads table - 081: payload column and last_fetch_at 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
2025-12-10 22:15:35 -07:00
parent 1fb0eb94c2
commit 4949b22457
33 changed files with 4064 additions and 737 deletions
--- a/backend/src/utils/payload-storage.ts
+++ b/backend/src/utils/payload-storage.ts
@@ -0,0 +1,406 @@
+/**
+ * Payload Storage Utility
+ *
+ * Per TASK_WORKFLOW_2024-12-10.md: Store raw GraphQL payloads for historical analysis.
+ *
+ * Design Pattern: Metadata/Payload Separation
+ * - Metadata in PostgreSQL (raw_crawl_payloads table): Small, indexed, queryable
+ * - Payload on filesystem: Gzipped JSON at storage_path
+ *
+ * Storage structure:
+ *   /storage/payloads/{year}/{month}/{day}/store_{dispensary_id}_{timestamp}.json.gz
+ *
+ * Benefits:
+ * - Compare any two crawls to see what changed
+ * - Replay/re-normalize historical data if logic changes
+ * - Debug issues by seeing exactly what the API returned
+ * - DB stays small, backups stay fast
+ * - ~90% compression (1.5MB -> 150KB per crawl)
+ */
+
+import * as fs from 'fs';
+import * as path from 'path';
+import * as zlib from 'zlib';
+import { promisify } from 'util';
+import { Pool } from 'pg';
+import * as crypto from 'crypto';
+
+const gzip = promisify(zlib.gzip);
+const gunzip = promisify(zlib.gunzip);
+
+// Base path for payload storage (matches image storage pattern)
+const PAYLOAD_BASE_PATH = process.env.PAYLOAD_STORAGE_PATH || './storage/payloads';
+
+/**
+ * Result from saving a payload
+ */
+export interface SavePayloadResult {
+  id: number;
+  storagePath: string;
+  sizeBytes: number;
+  sizeBytesRaw: number;
+  checksum: string;
+}
+
+/**
+ * Result from loading a payload
+ */
+export interface LoadPayloadResult {
+  payload: any;
+  metadata: {
+    id: number;
+    dispensaryId: number;
+    crawlRunId: number | null;
+    productCount: number;
+    fetchedAt: Date;
+    storagePath: string;
+  };
+}
+
+/**
+ * Generate storage path for a payload
+ *
+ * Format: /storage/payloads/{year}/{month}/{day}/store_{dispensary_id}_{timestamp}.json.gz
+ */
+function generateStoragePath(dispensaryId: number, timestamp: Date): string {
+  const year = timestamp.getFullYear();
+  const month = String(timestamp.getMonth() + 1).padStart(2, '0');
+  const day = String(timestamp.getDate()).padStart(2, '0');
+  const ts = timestamp.getTime();
+
+  return path.join(
+    PAYLOAD_BASE_PATH,
+    String(year),
+    month,
+    day,
+    `store_${dispensaryId}_${ts}.json.gz`
+  );
+}
+
+/**
+ * Ensure directory exists for a file path
+ */
+async function ensureDir(filePath: string): Promise<void> {
+  const dir = path.dirname(filePath);
+  await fs.promises.mkdir(dir, { recursive: true });
+}
+
+/**
+ * Calculate SHA256 checksum of data
+ */
+function calculateChecksum(data: Buffer): string {
+  return crypto.createHash('sha256').update(data).digest('hex');
+}
+
+/**
+ * Save a raw crawl payload to filesystem and record metadata in DB
+ *
+ * @param pool - Database connection pool
+ * @param dispensaryId - ID of the dispensary
+ * @param payload - Raw JSON payload from GraphQL
+ * @param crawlRunId - Optional crawl_run ID for linking
+ * @param productCount - Number of products in payload
+ * @returns SavePayloadResult with file info and DB record ID
+ */
+export async function saveRawPayload(
+  pool: Pool,
+  dispensaryId: number,
+  payload: any,
+  crawlRunId: number | null = null,
+  productCount: number = 0
+): Promise<SavePayloadResult> {
+  const timestamp = new Date();
+  const storagePath = generateStoragePath(dispensaryId, timestamp);
+
+  // Serialize and compress
+  const jsonStr = JSON.stringify(payload);
+  const rawSize = Buffer.byteLength(jsonStr, 'utf8');
+  const compressed = await gzip(Buffer.from(jsonStr, 'utf8'));
+  const compressedSize = compressed.length;
+  const checksum = calculateChecksum(compressed);
+
+  // Write to filesystem
+  await ensureDir(storagePath);
+  await fs.promises.writeFile(storagePath, compressed);
+
+  // Record metadata in DB
+  const result = await pool.query(`
+    INSERT INTO raw_crawl_payloads (
+      crawl_run_id,
+      dispensary_id,
+      storage_path,
+      product_count,
+      size_bytes,
+      size_bytes_raw,
+      fetched_at,
+      checksum_sha256
+    ) VALUES ($1, $2, $3, $4, $5, $6, $7, $8)
+    RETURNING id
+  `, [
+    crawlRunId,
+    dispensaryId,
+    storagePath,
+    productCount,
+    compressedSize,
+    rawSize,
+    timestamp,
+    checksum
+  ]);
+
+  console.log(`[PayloadStorage] Saved payload for store ${dispensaryId}: ${storagePath} (${(compressedSize / 1024).toFixed(1)}KB compressed, ${(rawSize / 1024).toFixed(1)}KB raw)`);
+
+  return {
+    id: result.rows[0].id,
+    storagePath,
+    sizeBytes: compressedSize,
+    sizeBytesRaw: rawSize,
+    checksum
+  };
+}
+
+/**
+ * Load a raw payload from filesystem by metadata ID
+ *
+ * @param pool - Database connection pool
+ * @param payloadId - ID from raw_crawl_payloads table
+ * @returns LoadPayloadResult with parsed payload and metadata
+ */
+export async function loadRawPayloadById(
+  pool: Pool,
+  payloadId: number
+): Promise<LoadPayloadResult | null> {
+  const result = await pool.query(`
+    SELECT id, dispensary_id, crawl_run_id, storage_path, product_count, fetched_at
+    FROM raw_crawl_payloads
+    WHERE id = $1
+  `, [payloadId]);
+
+  if (result.rows.length === 0) {
+    return null;
+  }
+
+  const row = result.rows[0];
+  const payload = await loadPayloadFromPath(row.storage_path);
+
+  return {
+    payload,
+    metadata: {
+      id: row.id,
+      dispensaryId: row.dispensary_id,
+      crawlRunId: row.crawl_run_id,
+      productCount: row.product_count,
+      fetchedAt: row.fetched_at,
+      storagePath: row.storage_path
+    }
+  };
+}
+
+/**
+ * Load a raw payload directly from filesystem path
+ *
+ * @param storagePath - Path to gzipped JSON file
+ * @returns Parsed JSON payload
+ */
+export async function loadPayloadFromPath(storagePath: string): Promise<any> {
+  const compressed = await fs.promises.readFile(storagePath);
+  const decompressed = await gunzip(compressed);
+  return JSON.parse(decompressed.toString('utf8'));
+}
+
+/**
+ * Get the latest payload for a dispensary
+ *
+ * @param pool - Database connection pool
+ * @param dispensaryId - ID of the dispensary
+ * @returns LoadPayloadResult or null if none exists
+ */
+export async function getLatestPayload(
+  pool: Pool,
+  dispensaryId: number
+): Promise<LoadPayloadResult | null> {
+  const result = await pool.query(`
+    SELECT id, dispensary_id, crawl_run_id, storage_path, product_count, fetched_at
+    FROM raw_crawl_payloads
+    WHERE dispensary_id = $1
+    ORDER BY fetched_at DESC
+    LIMIT 1
+  `, [dispensaryId]);
+
+  if (result.rows.length === 0) {
+    return null;
+  }
+
+  const row = result.rows[0];
+  const payload = await loadPayloadFromPath(row.storage_path);
+
+  return {
+    payload,
+    metadata: {
+      id: row.id,
+      dispensaryId: row.dispensary_id,
+      crawlRunId: row.crawl_run_id,
+      productCount: row.product_count,
+      fetchedAt: row.fetched_at,
+      storagePath: row.storage_path
+    }
+  };
+}
+
+/**
+ * Get two payloads for comparison (latest and previous, or by IDs)
+ *
+ * @param pool - Database connection pool
+ * @param dispensaryId - ID of the dispensary
+ * @param limit - Number of recent payloads to retrieve (default 2)
+ * @returns Array of LoadPayloadResult, most recent first
+ */
+export async function getRecentPayloads(
+  pool: Pool,
+  dispensaryId: number,
+  limit: number = 2
+): Promise<LoadPayloadResult[]> {
+  const result = await pool.query(`
+    SELECT id, dispensary_id, crawl_run_id, storage_path, product_count, fetched_at
+    FROM raw_crawl_payloads
+    WHERE dispensary_id = $1
+    ORDER BY fetched_at DESC
+    LIMIT $2
+  `, [dispensaryId, limit]);
+
+  const payloads: LoadPayloadResult[] = [];
+
+  for (const row of result.rows) {
+    const payload = await loadPayloadFromPath(row.storage_path);
+    payloads.push({
+      payload,
+      metadata: {
+        id: row.id,
+        dispensaryId: row.dispensary_id,
+        crawlRunId: row.crawl_run_id,
+        productCount: row.product_count,
+        fetchedAt: row.fetched_at,
+        storagePath: row.storage_path
+      }
+    });
+  }
+
+  return payloads;
+}
+
+/**
+ * List payload metadata without loading files (for browsing/pagination)
+ *
+ * @param pool - Database connection pool
+ * @param options - Query options
+ * @returns Array of metadata rows
+ */
+export async function listPayloadMetadata(
+  pool: Pool,
+  options: {
+    dispensaryId?: number;
+    startDate?: Date;
+    endDate?: Date;
+    limit?: number;
+    offset?: number;
+  } = {}
+): Promise<Array<{
+  id: number;
+  dispensaryId: number;
+  crawlRunId: number | null;
+  storagePath: string;
+  productCount: number;
+  sizeBytes: number;
+  sizeBytesRaw: number;
+  fetchedAt: Date;
+}>> {
+  const conditions: string[] = [];
+  const params: any[] = [];
+  let paramIndex = 1;
+
+  if (options.dispensaryId) {
+    conditions.push(`dispensary_id = $${paramIndex++}`);
+    params.push(options.dispensaryId);
+  }
+
+  if (options.startDate) {
+    conditions.push(`fetched_at >= $${paramIndex++}`);
+    params.push(options.startDate);
+  }
+
+  if (options.endDate) {
+    conditions.push(`fetched_at <= $${paramIndex++}`);
+    params.push(options.endDate);
+  }
+
+  const whereClause = conditions.length > 0 ? `WHERE ${conditions.join(' AND ')}` : '';
+  const limit = options.limit || 50;
+  const offset = options.offset || 0;
+
+  params.push(limit, offset);
+
+  const result = await pool.query(`
+    SELECT
+      id,
+      dispensary_id,
+      crawl_run_id,
+      storage_path,
+      product_count,
+      size_bytes,
+      size_bytes_raw,
+      fetched_at
+    FROM raw_crawl_payloads
+    ${whereClause}
+    ORDER BY fetched_at DESC
+    LIMIT $${paramIndex++} OFFSET $${paramIndex}
+  `, params);
+
+  return result.rows.map(row => ({
+    id: row.id,
+    dispensaryId: row.dispensary_id,
+    crawlRunId: row.crawl_run_id,
+    storagePath: row.storage_path,
+    productCount: row.product_count,
+    sizeBytes: row.size_bytes,
+    sizeBytesRaw: row.size_bytes_raw,
+    fetchedAt: row.fetched_at
+  }));
+}
+
+/**
+ * Delete old payloads (for retention policy)
+ *
+ * @param pool - Database connection pool
+ * @param olderThan - Delete payloads older than this date
+ * @returns Number of payloads deleted
+ */
+export async function deleteOldPayloads(
+  pool: Pool,
+  olderThan: Date
+): Promise<number> {
+  // Get paths first
+  const result = await pool.query(`
+    SELECT id, storage_path FROM raw_crawl_payloads
+    WHERE fetched_at < $1
+  `, [olderThan]);
+
+  // Delete files
+  for (const row of result.rows) {
+    try {
+      await fs.promises.unlink(row.storage_path);
+    } catch (err: any) {
+      if (err.code !== 'ENOENT') {
+        console.warn(`[PayloadStorage] Failed to delete ${row.storage_path}: ${err.message}`);
+      }
+    }
+  }
+
+  // Delete DB records
+  await pool.query(`
+    DELETE FROM raw_crawl_payloads
+    WHERE fetched_at < $1
+  `, [olderThan]);
+
+  console.log(`[PayloadStorage] Deleted ${result.rows.length} payloads older than ${olderThan.toISOString()}`);
+
+  return result.rows.length;
+}