transceiver-db/packages/scraper/src/scrapers/champion-one.ts

/**
 * Champion ONE Scraper — US-based compatible transceiver vendor
 *
 * championone.com — Server-rendered HTML, no JS required.
 * Large catalog: SFP, SFP+, SFP28, QSFP+, QSFP28, QSFP-DD, OSFP, XFP, X2, GBIC
 *
 * Rate limited: 1 req/2sec.
 */
import { pool, findOrCreateScrapedTransceiver, ensureVendor, upsertPriceObservation } from "../utils/db";
import { contentHash } from "../utils/hash";

const BASE = "https://www.championone.com";
const HEADERS = {
  "User-Agent": "Mozilla/5.0 (compatible; TIP-Bot/1.0; research)",
  Accept: "text/html,application/xhtml+xml",
};

const CATEGORIES = [
  { path: "/sfp-transceivers", formFactor: "SFP", speed: "1G", speedGbps: 1 },
  { path: "/sfp-plus-transceivers", formFactor: "SFP+", speed: "10G", speedGbps: 10 },
  { path: "/25g-sfp28-transceivers", formFactor: "SFP28", speed: "25G", speedGbps: 25 },
  { path: "/qsfp-plus-transceivers", formFactor: "QSFP+", speed: "40G", speedGbps: 40 },
  { path: "/qsfp28-transceivers", formFactor: "QSFP28", speed: "100G", speedGbps: 100 },
  { path: "/qsfp-dd-transceivers", formFactor: "QSFP-DD", speed: "400G", speedGbps: 400 },
  { path: "/xfp-transceivers", 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;
  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][] = [
    [/\b160\s*km\b/i, "160km", 160000],
    [/\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],
    [/\b300\s*m\b/i, "300m", 300],
    [/\b100\s*m\b/i, "100m", 100],
    [/\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],
    [/\bCWDM4\b/i, "2km", 2000],
    [/\bPSM4\b/i, "500m", 500],
  ];
  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|rj45|base-t/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[] = [];

  // Champion ONE uses standard ecommerce HTML with product cards
  const productRegex = /href="(\/[^"]*?(?:transceiver|sfp|qsfp|osfp|xfp|optic)[^"]*)"[^>]*>([^<]{5,})<\/a>/gi;
  let match;
  while ((match = productRegex.exec(html)) !== null) {
    const url = match[1];
    const name = match[2].trim();
    if (name.length < 8 || name.length > 200) continue;

    const context = html.slice(Math.max(0, match.index - 200), match.index + 500);
    const priceMatch = context.match(/\$\s*([\d,]+\.?\d{0,2})/) || context.match(/USD\s*([\d,]+\.?\d{0,2})/i);
    const price = priceMatch ? parseFloat(priceMatch[1].replace(",", "")) : undefined;

    const partNum = name.replace(/\s+/g, "-").replace(/[^a-zA-Z0-9\-]/g, "").slice(0, 80);
    const reach = detectReach(name);

    products.push({
      partNumber: partNum, name,
      url: url.startsWith("http") ? url : BASE + url,
      price: price && price > 0 && price < 50000 ? price : undefined,
      currency: price ? "USD" : undefined,
      formFactor: cat.formFactor, speed: cat.speed, speedGbps: cat.speedGbps,
      reachLabel: reach?.label, reachMeters: reach?.meters,
      fiberType: detectFiber(name), wavelength: detectWavelength(name),
    });
  }

  // Pattern 2: Generic product card pattern
  const cardRegex = /class="[^"]*product[^"]*"[\s\S]*?href="([^"]+)"[^>]*>[\s\S]*?(?:name|title)[^>]*>([^<]+)/gi;
  while ((match = cardRegex.exec(html)) !== null) {
    const url = match[1];
    const name = match[2].trim();
    if (products.find((p) => p.url === (url.startsWith("http") ? url : BASE + url))) continue;
    if (name.length < 8) continue;

    const context = html.slice(match.index, match.index + 500);
    const priceMatch = context.match(/\$\s*([\d,]+\.?\d{0,2})/);
    const price = priceMatch ? parseFloat(priceMatch[1].replace(",", "")) : undefined;
    const reach = detectReach(name);

    products.push({
      partNumber: name.replace(/\s+/g, "-").replace(/[^a-zA-Z0-9\-]/g, "").slice(0, 80),
      name,
      url: url.startsWith("http") ? url : BASE + url,
      price: price && price > 0 && price < 50000 ? price : undefined,
      currency: price ? "USD" : undefined,
      formFactor: cat.formFactor, speed: cat.speed, speedGbps: cat.speedGbps,
      reachLabel: reach?.label, reachMeters: reach?.meters,
      fiberType: detectFiber(name), wavelength: detectWavelength(name),
    });
  }

  const seen = new Set<string>();
  return products.filter((p) => {
    if (seen.has(p.url)) return false;
    seen.add(p.url);
    return true;
  });
}

function getMaxPage(html: string): number {
  const pageMatches = html.match(/[?&]page=(\d+)/g) || html.match(/\/page\/(\d+)/g);
  if (!pageMatches) return 1;
  let max = 1;
  for (const m of pageMatches) {
    const n = parseInt(m.replace(/[^0-9]/g, ""));
    if (n > max) max = n;
  }
  return Math.min(max, 30);
}

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 scrapeChampionOne(): Promise<void> {
  console.log("=== Champion ONE Scraper Starting ===\n");

  const vendorId = await ensureVendor("Champion ONE", "compatible", "https://www.championone.com", "https://www.championone.com");

  let totalProducts = 0;
  let priceUpdates = 0;

  for (const cat of CATEGORIES) {
    console.log(`\n--- ${cat.formFactor} (${cat.speed}) ---`);

    try {
      const firstPage = await fetchPage(BASE + cat.path);
      const maxPage = getMaxPage(firstPage);
      console.log(`  Pages: ${maxPage}`);

      let catProducts: Product[] = parseProductList(firstPage, cat);

      for (let page = 2; page <= maxPage; page++) {
        await sleep(2000);
        try {
          const html = await fetchPage(`${BASE}${cat.path}?page=${page}`);
          catProducts.push(...parseProductList(html, cat));
        } catch (err) {
          console.warn(`  Page ${page} failed: ${(err as Error).message}`);
        }
      }

      const seen = new Set<string>();
      catProducts = catProducts.filter((p) => {
        if (seen.has(p.url)) return false;
        seen.add(p.url);
        return true;
      });

      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,
            wavelengths: product.wavelength, 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=== Champion ONE Complete: ${totalProducts} products, ${priceUpdates} prices ===`);
}

if (require.main === module) {
  scrapeChampionOne()
    .then(() => pool.end())
    .catch((err) => { console.error("Fatal:", err); pool.end(); process.exit(1); });
}