brightdata-sdk-patterns

Bright Data SDK Patterns

Overview

Production-ready patterns for Bright Data proxy integrations. Since Bright Data uses HTTP proxy protocols (not a dedicated SDK), these patterns wrap proxy configuration, retry logic, session management, and response parsing into reusable modules.

Prerequisites

• Completed brightdata-install-auth setup
• Familiarity with async/await and HTTP proxy protocols
• axios or node-fetch installed

Instructions

Step 1: Proxy Client Singleton

// src/brightdata/client.ts
import axios, { AxiosInstance, AxiosRequestConfig } from 'axios';
import https from 'https';
import 'dotenv/config';

let instance: AxiosInstance | null = null;

export function getBrightDataClient(options?: {
  country?: string;
  session?: string;
  zone?: string;
}): AxiosInstance {
  const { BRIGHTDATA_CUSTOMER_ID, BRIGHTDATA_ZONE, BRIGHTDATA_ZONE_PASSWORD } = process.env;
  const zone = options?.zone || BRIGHTDATA_ZONE!;

  let username = `brd-customer-${BRIGHTDATA_CUSTOMER_ID}-zone-${zone}`;
  if (options?.country) username += `-country-${options.country}`;
  if (options?.session) username += `-session-${options.session}`;

  if (!instance || options) {
    instance = axios.create({
      proxy: {
        host: 'brd.superproxy.io',
        port: 33335,
        auth: { username, password: BRIGHTDATA_ZONE_PASSWORD! },
      },
      httpsAgent: new https.Agent({ rejectUnauthorized: false }),
      timeout: 60000,
      headers: { 'User-Agent': 'Mozilla/5.0 (compatible; Scraper/1.0)' },
    });
  }
  return instance;
}

Step 2: Retry Wrapper with Proxy Error Handling

// src/brightdata/retry.ts
export async function scrapeWithRetry<T>(
  url: string,
  parser: (html: string) => T,
  config = { maxRetries: 3, baseDelayMs: 2000, maxDelayMs: 30000 }
): Promise<T> {
  const client = getBrightDataClient();

  for (let attempt = 0; attempt <= config.maxRetries; attempt++) {
    try {
      const response = await client.get(url);
      return parser(response.data);
    } catch (error: any) {
      const status = error.response?.status;
      // Bright Data proxy errors that warrant retry
      const retryable = [502, 503, 407, 429].includes(status) || error.code === 'ETIMEDOUT';

      if (attempt === config.maxRetries || !retryable) throw error;

      const delay = Math.min(
        config.baseDelayMs * Math.pow(2, attempt) + Math.random() * 1000,
        config.maxDelayMs
      );
      console.log(`Attempt ${attempt + 1} failed (${status || error.code}), retrying in ${delay.toFixed(0)}ms`);
      await new Promise(r => setTimeout(r, delay));
    }
  }
  throw new Error('Unreachable');
}

Step 3: Session Management for Sticky IPs

// src/brightdata/sessions.ts — maintain same IP across requests
import { v4 as uuidv4 } from 'uuid';

export class StickySession {
  private sessionId: string;
  private client: AxiosInstance;

  constructor(country?: string) {
    this.sessionId = uuidv4();
    this.client = getBrightDataClient({
      country,
      session: this.sessionId, // Same session = same exit IP
    });
  }

  async get(url: string) {
    return this.client.get(url);
  }

  // Create new session (rotates IP)
  rotate(): void {
    this.sessionId = uuidv4();
    this.client = getBrightDataClient({ session: this.sessionId });
  }
}

// Usage: login flow that needs consistent IP
const session = new StickySession('us');
await session.get('https://example.com/login'); // IP: 1.2.3.4
await session.get('https://example.com/dashboard'); // IP: 1.2.3.4 (same)
session.rotate();
await session.get('https://example.com/other'); // IP: 5.6.7.8 (new)

Step 4: HTML Response Parser with Cheerio

// src/brightdata/parser.ts
import * as cheerio from 'cheerio';

export function parseProductPage(html: string) {
  const $ = cheerio.load(html);
  return {
    title: $('h1').first().text().trim(),
    price: $('[data-price], .price').first().text().trim(),
    description: $('meta[name="description"]').attr('content') || '',
    images: $('img[src]').map((_, el) => $(el).attr('src')).get().slice(0, 10),
    inStock: !$('.out-of-stock').length,
  };
}

export function parseSearchResults(html: string) {
  const $ = cheerio.load(html);
  return $('div.g, [data-result]').map((i, el) => ({
    rank: i + 1,
    title: $(el).find('h3').text().trim(),
    link: $(el).find('a').attr('href') || '',
    snippet: $(el).find('.VwiC3b, .st').text().trim(),
  })).get();
}

Step 5: Python Context Manager

# brightdata/client.py
import os, requests
from contextlib import contextmanager
from dotenv import load_dotenv

load_dotenv()

@contextmanager
def brightdata_session(country=None, city=None):
    """Context manager for Bright Data proxy sessions."""
    cid = os.environ['BRIGHTDATA_CUSTOMER_ID']
    zone = os.environ['BRIGHTDATA_ZONE']
    pwd = os.environ['BRIGHTDATA_ZONE_PASSWORD']

    username = f'brd-customer-{cid}-zone-{zone}'
    if country: username += f'-country-{country}'
    if city: username += f'-city-{city}'

    proxy_url = f'http://{username}:{pwd}@brd.superproxy.io:33335'
    session = requests.Session()
    session.proxies = {'http': proxy_url, 'https': proxy_url}
    session.verify = './brd-ca.crt'
    session.headers['User-Agent'] = 'Mozilla/5.0'

    try:
        yield session
    finally:
        session.close()

# Usage
with brightdata_session(country='us') as s:
    resp = s.get('https://example.com')
    print(resp.status_code)

Output

• Reusable proxy client singleton with geo-targeting
• Retry wrapper handling 502, 503, 407, 429, and timeouts
• Sticky session management for multi-step scraping flows
• HTML parsing utilities with cheerio
• Python context manager pattern

Error Handling

Pattern	Use Case	Benefit
Singleton client	All proxy requests	Consistent config, connection reuse
Retry wrapper	Transient proxy errors	Auto-recovery from 502/503
Sticky sessions	Login flows, pagination	Same IP across requests
Response cache	Development	Avoids burning proxy credits

Resources

• Bright Data Proxy Docs
• Session Management
• Cheerio Documentation

Next Steps

Apply patterns in brightdata-core-workflow-a for real-world browser scraping.