Crawlr-Browser / src /queue.js
senku21230's picture
Upload 11 files
ce9e053 verified
Raw
History Blame Contribute Delete
2 kB
// src/queue.js β€” Rate limiting, retry, per-domain throttle
const pRetry = require("p-retry");
// ── Per-domain rate limiter ───────────────────────────────────────────────
const domainLastRequest = new Map();
const DOMAIN_DELAY_MS = 1000; // একই domain-এ ΰ§§ সেকেন্঑ gap
async function domainThrottle(url) {
try {
const domain = new URL(url).hostname;
const last = domainLastRequest.get(domain) || 0;
const wait = DOMAIN_DELAY_MS - (Date.now() - last);
if (wait > 0) await new Promise((r) => setTimeout(r, wait));
domainLastRequest.set(domain, Date.now());
} catch (_) {}
}
// ── Retry wrapper ─────────────────────────────────────────────────────────
async function withRetry(fn, options = {}) {
const { retries = 3, minTimeout = 1000, onRetry } = options;
return pRetry(fn, {
retries,
minTimeout,
maxTimeout: 10000,
randomize: true,
onFailedAttempt: (err) => {
console.warn(`⚠️ Retry ${err.attemptNumber}/${retries + 1}: ${err.message}`);
if (onRetry) onRetry(err);
},
});
}
// ── Webhook sender ────────────────────────────────────────────────────────
async function sendWebhook(url, data) {
if (!url) return;
const fetch = require("node-fetch");
try {
await fetch(url, {
method: "POST",
headers: { "Content-Type": "application/json" },
body: JSON.stringify({ ...data, timestamp: new Date().toISOString() }),
timeout: 10000,
});
console.log(`πŸ“‘ Webhook sent β†’ ${url}`);
} catch (err) {
console.warn(`⚠️ Webhook failed: ${err.message}`);
}
}
module.exports = { domainThrottle, withRetry, sendWebhook };