# Conflicts: # packages/api/src/llm/fo-blog-pipeline.ts # packages/api/src/routes/blog.ts # packages/scraper/src/scheduler.ts # packages/scraper/src/scrapers/fs-com.ts # packages/scraper/src/scrapers/gbics.ts
238 lines
8.7 KiB
TypeScript
238 lines
8.7 KiB
TypeScript
/**
|
|
* SFPcables.com Scraper — 10Gtek's Retail Store
|
|
*
|
|
* sfpcables.com — Magento store with server-rendered HTML, real USD prices.
|
|
* Product pages have clean <h2 class="product-name"> + <span class="price"> structure.
|
|
* Rate limited: 1 req/2sec. TLS verification disabled (self-signed cert issues).
|
|
*/
|
|
import { pool, findOrCreateScrapedTransceiver, ensureVendor, upsertPriceObservation } from "../utils/db";
|
|
import { contentHash } from "../utils/hash";
|
|
|
|
const BASE = "https://www.sfpcables.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-1-25g-series", formFactor: "SFP", speed: "1G", speedGbps: 1 },
|
|
{ path: "/sfp-transceivers", formFactor: "SFP+", speed: "10G", speedGbps: 10 },
|
|
{ path: "/sfp28-transceivers", formFactor: "SFP28", speed: "25G", speedGbps: 25 },
|
|
{ path: "/qsfp-transceivers", formFactor: "QSFP+", speed: "40G", speedGbps: 40 },
|
|
{ path: "/100g-qsfp28-transceivers", formFactor: "QSFP28", speed: "100G", speedGbps: 100 },
|
|
{ path: "/qsfp-dd-400g-transceivers", formFactor: "QSFP-DD", speed: "400G", speedGbps: 400 },
|
|
{ path: "/xfp-transceivers", formFactor: "XFP", speed: "10G", speedGbps: 10 },
|
|
{ path: "/2-5g-transceivers", formFactor: "SFP", speed: "2.5G", speedGbps: 2.5 },
|
|
{ path: "/industrial-sfp-transceivers", formFactor: "SFP", speed: "1G", speedGbps: 1 },
|
|
{ path: "/industrial-qsfp-transceivers", formFactor: "QSFP28", speed: "100G", speedGbps: 100 },
|
|
{ path: "/8x50g-qsfp-dd-transceiver-optical-module", formFactor: "QSFP-DD", speed: "400G", speedGbps: 400 },
|
|
{ path: "/8x100g-qsfp-dd-transceiver-optical-module", formFactor: "QSFP-DD", speed: "800G", speedGbps: 800 },
|
|
{ path: "/osfp-flat-fiber-optic-transceiver-modules", formFactor: "OSFP", speed: "400G", speedGbps: 400 },
|
|
{ path: "/400g-8x50g-osfp-fin-fiber-optic-transceiver-modules", formFactor: "OSFP", speed: "400G", speedGbps: 400 },
|
|
{ path: "/fc16g-sfp-transceivers", formFactor: "SFP+", speed: "16G FC", speedGbps: 16 },
|
|
{ path: "/fc32g-sfp-transceivers", formFactor: "SFP28", speed: "32G FC", speedGbps: 32 },
|
|
];
|
|
|
|
interface Product {
|
|
partNumber: string;
|
|
name: string;
|
|
url: string;
|
|
price?: number;
|
|
formFactor: string;
|
|
speed: string;
|
|
speedGbps: number;
|
|
reachLabel?: string;
|
|
reachMeters?: number;
|
|
fiberType?: string;
|
|
wavelength?: 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][] = [
|
|
[/\b120\s*km\b/i, "120km", 120000],
|
|
[/\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],
|
|
[/\b550\s*m\b/i, "550m", 550],
|
|
[/\b500\s*m\b/i, "500m", 500],
|
|
[/\b400\s*m\b/i, "400m", 400],
|
|
[/\b300\s*m\b/i, "300m", 300],
|
|
[/\b150\s*m\b/i, "150m", 150],
|
|
[/\b100\s*m\b/i, "100m", 100],
|
|
[/\b30\s*m\b/i, "30m", 30],
|
|
[/\bLR4\b/, "10km", 10000],
|
|
[/\bLR\b/, "10km", 10000],
|
|
[/\bER4?\b/, "40km", 40000],
|
|
[/\bZR4?\b/, "80km", 80000],
|
|
[/\bSR4?\b/, "300m", 300],
|
|
[/\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";
|
|
if (/copper|dac|twinax|rj.?45|base-t|cat[56x]/i.test(text)) return "Copper";
|
|
return "";
|
|
}
|
|
|
|
function detectWavelength(text: string): string {
|
|
const match = text.match(/(\d{3,4})\s*nm/i);
|
|
if (match) return match[1];
|
|
return "";
|
|
}
|
|
|
|
function parseProductList(html: string, cat: typeof CATEGORIES[number]): Product[] {
|
|
const products: Product[] = [];
|
|
|
|
// Magento product listing: <h2 class="product-name"><a href="URL" title="NAME">NAME</a></h2>
|
|
// Prices: <span class="price">US$XX.XX</span>
|
|
const productRegex = /<h2\s+class="product-name">\s*<a\s+href="([^"]+)"[^>]*title="([^"]+)"[^>]*>/gi;
|
|
let match;
|
|
while ((match = productRegex.exec(html)) !== null) {
|
|
const url = match[1];
|
|
const name = match[2].trim();
|
|
if (name.length < 5) continue;
|
|
|
|
// Find price after this product name (within next 800 chars)
|
|
const afterContext = html.slice(match.index, match.index + 800);
|
|
const priceMatch = afterContext.match(/class="price">\s*US?\$\s*([\d,.]+)/);
|
|
const price = priceMatch ? parseFloat(priceMatch[1].replace(",", "")) : undefined;
|
|
|
|
const reach = detectReach(name);
|
|
// Build part number from URL slug
|
|
const slug = url.split("/").filter(Boolean).pop() || "";
|
|
const partNumber = slug.replace(/-\d+$/, "").slice(0, 80);
|
|
|
|
products.push({
|
|
partNumber,
|
|
name,
|
|
url,
|
|
price: price && price > 0 && price < 50000 ? price : undefined,
|
|
formFactor: cat.formFactor,
|
|
speed: cat.speed,
|
|
speedGbps: cat.speedGbps,
|
|
reachLabel: reach?.label,
|
|
reachMeters: reach?.meters,
|
|
fiberType: detectFiber(name),
|
|
wavelength: detectWavelength(name),
|
|
});
|
|
}
|
|
|
|
// Dedupe by URL
|
|
const seen = new Set<string>();
|
|
return products.filter((p) => {
|
|
if (seen.has(p.url)) return false;
|
|
seen.add(p.url);
|
|
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 scrapeSfpCables(): Promise<void> {
|
|
console.log("=== SFPcables.com Scraper Starting ===\n");
|
|
|
|
const vendorId = await ensureVendor("SFPcables", "compatible", "https://www.sfpcables.com", "https://www.sfpcables.com/transceivers");
|
|
|
|
let totalProducts = 0;
|
|
let priceUpdates = 0;
|
|
|
|
for (const cat of CATEGORIES) {
|
|
console.log(`\n--- ${cat.formFactor} (${cat.speed}) [${cat.path}] ---`);
|
|
|
|
try {
|
|
const html = await fetchPage(BASE + cat.path);
|
|
const catProducts = parseProductList(html, cat);
|
|
console.log(` Found ${catProducts.length} products`);
|
|
|
|
// Check for pagination (Magento uses ?p=N)
|
|
const pageLinks = html.match(/[?&]p=(\d+)/g);
|
|
let maxPage = 1;
|
|
if (pageLinks) {
|
|
for (const pl of pageLinks) {
|
|
const n = parseInt(pl.replace(/[^0-9]/g, ""));
|
|
if (n > maxPage) maxPage = n;
|
|
}
|
|
}
|
|
|
|
// Fetch additional pages
|
|
for (let page = 2; page <= Math.min(maxPage, 10); page++) {
|
|
await sleep(2000);
|
|
try {
|
|
const pageHtml = await fetchPage(`${BASE}${cat.path}?p=${page}`);
|
|
const pageProducts = parseProductList(pageHtml, cat);
|
|
catProducts.push(...pageProducts);
|
|
console.log(` Page ${page}: ${pageProducts.length} products`);
|
|
} catch (err) {
|
|
console.warn(` Page ${page} failed: ${(err as Error).message}`);
|
|
}
|
|
}
|
|
|
|
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,
|
|
wavelengths: product.wavelength,
|
|
category: "DataCenter",
|
|
});
|
|
|
|
if (product.price && product.price > 0) {
|
|
const hash = contentHash({ price: product.price, part: product.partNumber });
|
|
const updated = await upsertPriceObservation({
|
|
transceiverId: txId,
|
|
sourceVendorId: vendorId,
|
|
price: product.price,
|
|
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=== SFPcables Complete: ${totalProducts} products, ${priceUpdates} prices ===`);
|
|
}
|
|
|
|
if (require.main === module) {
|
|
scrapeSfpCables()
|
|
.then(() => pool.end())
|
|
.catch((err) => { console.error("Fatal:", err); pool.end(); process.exit(1); });
|
|
}
|