Rene Fichtmueller ba24d33858 fix(scrapers): replace bot User-Agents with Chrome UA + disable dead domain
- 16 commercial scrapers: replace TIP-Bot/1.0 with Chrome/120 UA
  (GBICS confirmed returning 0 bytes for bot UA, Chrome UA returns 200KB)
- gbics.ts: fix User-Agent (was returning empty HTML, now returns products)
- optictransceiver.ts: disable — domain repurposed as plant shop (2026-04-06)
  Alocasia Regal Shield is not a transceiver.
2026-04-06 02:17:50 +02:00

232 lines
7.9 KiB
TypeScript

/**
* 10Gtek.com Scraper — Chinese OEM Transceiver Vendor
*
* 10gtek.com is a direct competitor to FS.com at lower price points.
* Uses plain fetch (server-rendered HTML).
* Rate limited: 1 req/2sec.
*
* Categories: SFP, SFP+, SFP28, QSFP+, QSFP28, QSFP-DD, OSFP
*/
import { pool, findOrCreateScrapedTransceiver, ensureVendor, upsertPriceObservation } from "../utils/db";
import { contentHash, parsePrice } from "../utils/hash";
const BASE = "https://www.10gtek.com";
const HEADERS = {
"User-Agent": "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/120.0.0.0 Safari/537.36",
Accept: "text/html,application/xhtml+xml",
};
const CATEGORIES = [
{ path: "/sfp", formFactor: "SFP", speed: "1G", speedGbps: 1 },
{ path: "/10g-sfp+", formFactor: "SFP+", speed: "10G", speedGbps: 10 },
{ path: "/sfp28", formFactor: "SFP28", speed: "25G", speedGbps: 25 },
{ path: "/qsfp", formFactor: "QSFP+", speed: "40G", speedGbps: 40 },
{ path: "/qsfp28", formFactor: "QSFP28", speed: "100G", speedGbps: 100 },
{ path: "/qsfpdd", formFactor: "QSFP-DD", speed: "400G", speedGbps: 400 },
{ path: "/xfp", formFactor: "XFP", speed: "10G", speedGbps: 10 },
];
interface Product {
partNumber: string;
name: string;
url: string;
price?: number;
currency?: string;
formFactor: string;
speed: string;
speedGbps: number;
reachLabel?: string;
reachMeters?: number;
fiberType?: string;
}
function sleep(ms: number): Promise<void> {
return new Promise((resolve) => setTimeout(resolve, ms));
}
function detectReach(text: string): { label: string; meters: number } | undefined {
const patterns: [RegExp, string, number][] = [
[/\b80\s*km\b/i, "80km", 80000],
[/\b40\s*km\b/i, "40km", 40000],
[/\b20\s*km\b/i, "20km", 20000],
[/\b10\s*km\b/i, "10km", 10000],
[/\b2\s*km\b/i, "2km", 2000],
[/\b500\s*m\b/i, "500m", 500],
[/\b300\s*m\b/i, "300m", 300],
[/\b100\s*m\b/i, "100m", 100],
[/\bLR4\b/, "10km", 10000],
[/\bLR\b/, "10km", 10000],
[/\bER\b/, "40km", 40000],
[/\bZR\b/, "80km", 80000],
[/\bSR4?\b/, "100m", 100],
[/\bDR4?\b/, "500m", 500],
[/\bFR4?\b/, "2km", 2000],
];
for (const [regex, label, meters] of patterns) {
if (regex.test(text)) return { label, meters };
}
return undefined;
}
function detectFiber(text: string): string {
if (/single.?mode|smf|[^a-z]lx[^a-z]|[^a-z]lr[^a-z]|[^a-z]er[^a-z]|[^a-z]zr[^a-z]|bidi|cwdm|dwdm/i.test(text)) return "SMF";
if (/multi.?mode|mmf|[^a-z]sx[^a-z]|[^a-z]sr[^a-z]/i.test(text)) return "MMF";
return "";
}
/** Strip HTML tags and decode common entities */
function stripHtml(s: string): string {
return s.replace(/<[^>]+>/g, "").replace(/&amp;/g, "&").replace(/&lt;/g, "<")
.replace(/&gt;/g, ">").replace(/&nbsp;/g, " ").replace(/&deg;/g, "°")
.replace(/&#\d+;/g, "").trim();
}
function parseDistance(text: string): { label: string; meters: number } | undefined {
const km = text.match(/(\d+)\s*km/i);
if (km) return { label: `${km[1]}km`, meters: parseInt(km[1]) * 1000 };
const m = text.match(/(\d+)\s*m\b/i);
if (m) return { label: `${m[1]}m`, meters: parseInt(m[1]) };
return undefined;
}
function parseProductList(html: string, cat: typeof CATEGORIES[number]): Product[] {
const products: Product[] = [];
// 10Gtek uses HTML tables with columns:
// Part No. | Spec | Data Rate | Wavelength | Fiber Type | Distance | Optical Comp. | Tx Power | E.R | Rx Sens. | Temp.
// Extract all <tr> rows and parse cells
const rowRegex = /<tr[^>]*>([\s\S]*?)<\/tr>/gi;
let rowMatch;
while ((rowMatch = rowRegex.exec(html)) !== null) {
const rowHtml = rowMatch[1];
// Extract all <td> cell contents
const cellRegex = /<td[^>]*>([\s\S]*?)<\/td>/gi;
const cells: string[] = [];
let cellMatch;
while ((cellMatch = cellRegex.exec(rowHtml)) !== null) {
cells.push(stripHtml(cellMatch[1]));
}
// Need at least 6 columns, first cell must look like a part number (starts with A or contains letters+digits)
if (cells.length < 6) continue;
const partNumber = cells[0];
if (!partNumber || partNumber.length < 3) continue;
// Skip header rows
if (/^Part\s*No/i.test(partNumber) || /^Spec/i.test(partNumber)) continue;
// Part numbers typically start with A (ASF, AXS, AXQ, AQS, etc.) or contain alphanumeric
if (!/^[A-Z][A-Z0-9]/i.test(partNumber)) continue;
const spec = cells[1] || "";
const dataRate = cells[2] || "";
const wavelength = cells.length >= 4 ? cells[3] : "";
const fiberType = cells.length >= 5 ? cells[4] : "";
const distance = cells.length >= 6 ? cells[5] : "";
const txPower = cells.length >= 8 ? cells[7] : "";
// Build descriptive name
const name = `${partNumber} ${spec} ${dataRate}`.trim();
const reach = parseDistance(distance) || detectReach(spec + " " + distance);
// Determine fiber type from table cell or spec
let fiber = "";
if (/SMF|single/i.test(fiberType)) fiber = "SMF";
else if (/MMF|multi/i.test(fiberType)) fiber = "MMF";
else if (/CAT|RJ|copper/i.test(fiberType)) fiber = "Copper";
else fiber = detectFiber(spec);
// Extract wavelength
const wl = wavelength.replace(/[^0-9]/g, "");
products.push({
partNumber,
name,
url: `${BASE}${cat.path}#${partNumber}`,
formFactor: cat.formFactor,
speed: cat.speed,
speedGbps: cat.speedGbps,
reachLabel: reach?.label,
reachMeters: reach?.meters,
fiberType: fiber,
});
}
// Dedupe by part number
const seen = new Set<string>();
return products.filter((p) => {
if (seen.has(p.partNumber)) return false;
seen.add(p.partNumber);
return true;
});
}
async function fetchPage(url: string): Promise<string> {
const resp = await fetch(url, { headers: HEADERS, signal: AbortSignal.timeout(30000) });
if (!resp.ok) throw new Error(`HTTP ${resp.status} for ${url}`);
return resp.text();
}
export async function scrape10Gtek(): Promise<void> {
console.log("=== 10Gtek Scraper Starting ===\n");
const vendorId = await ensureVendor("10Gtek", "compatible", "https://www.10gtek.com", "https://www.10gtek.com");
let totalProducts = 0;
let priceUpdates = 0;
for (const cat of CATEGORIES) {
console.log(`\n--- ${cat.formFactor} (${cat.speed}) ---`);
try {
const html = await fetchPage(BASE + cat.path);
const catProducts = parseProductList(html, cat);
console.log(` Found ${catProducts.length} products`);
for (const product of catProducts) {
try {
const txId = await findOrCreateScrapedTransceiver({
partNumber: product.partNumber,
vendorId,
formFactor: product.formFactor,
speedGbps: product.speedGbps,
speed: product.speed,
reachMeters: product.reachMeters,
reachLabel: product.reachLabel,
fiberType: product.fiberType,
category: "DataCenter",
});
if (product.price && product.price > 0) {
const hash = contentHash(JSON.stringify({ price: product.price, part: product.partNumber }));
const updated = await upsertPriceObservation({
transceiverId: txId,
sourceVendorId: vendorId,
price: product.price,
currency: product.currency || "USD",
stockLevel: "in_stock",
url: product.url,
contentHash: hash,
});
if (updated) priceUpdates++;
}
totalProducts++;
} catch (err) {
console.warn(` Error: ${(err as Error).message.slice(0, 80)}`);
}
}
} catch (err) {
console.error(` Category failed: ${(err as Error).message}`);
}
await sleep(2000);
}
console.log(`\n=== 10Gtek Complete: ${totalProducts} products, ${priceUpdates} prices ===`);
}
if (require.main === module) {
scrape10Gtek()
.then(() => pool.end())
.catch((err) => { console.error("Fatal:", err); pool.end(); process.exit(1); });
}