' + esc(key) + ' (' + items.length + ')
';
+ ch += '';
+ items.slice(0, 6).forEach(function(t) {
var priceStr = '';
if (t.latest_price) {
var _pAmt = parseFloat(t.latest_price);
var _pCur = (t.latest_currency || 'USD').toUpperCase();
var _pUSD = toUSD(_pAmt, _pCur);
- var _pEUR = toEUR(_pAmt, _pCur);
- priceStr = ' — ' + (_pUSD !== null ? fmtUSD(_pUSD) : _pCur + ' ' + _pAmt.toFixed(2))
- + (_pEUR !== null ? ' / ' + fmtEUR(_pEUR) : '');
+ priceStr = _pUSD !== null ? fmtUSD(_pUSD) : _pCur + ' ' + _pAmt.toFixed(2);
}
- var verBadge = (t.price_verified === true)
- ? '✓ Verified' : '';
- var fullyBadge = (t.fully_verified === true)
- ? '★ 100%' : '';
- var foUrl = t.product_page_url
- ? 'Shop ↗' : '';
- ch += ' string | null> = {
- cisco: buildCiscoUrl,
- arista: buildAristaUrl,
- juniper: buildJuniperUrl,
- "nvidia-networking": buildNvidiaUrl,
- edgecore: buildEdgecoreUrl,
- celestica: buildCelesticaUrl,
- asterfusion: buildAsterfusionUrl,
- dell: buildDellUrl,
- "hpe-aruba": buildHpeArubaUrl,
- huawei: buildHuaweiUrl,
- nokia: buildNobelUrl,
- extreme: buildExtremeUrl,
- mikrotik: buildMikroTikUrl,
- ubiquiti: buildUbiquitiUrl,
- "fs-com": buildFsComUrl,
- supermicro: buildSupermicroUrl,
- wistron: (_m) => null, // no public product pages
+ cisco: buildCiscoUrl,
+ arista: buildAristaUrl,
+ juniper: buildJuniperUrl,
+ "nvidia-networking": buildNvidiaUrl,
+ edgecore: buildEdgecoreUrl,
+ celestica: buildCelesticaUrl,
+ asterfusion: buildAsterfusionUrl,
+ dell: buildDellUrl,
+ "hpe-aruba": buildHpeArubaUrl,
+ huawei: buildHuaweiUrl,
+ nokia: buildNobelUrl,
+ extreme: buildExtremeUrl,
+ mikrotik: buildMikroTikUrl,
+ ubiquiti: buildUbiquitiUrl,
+ "fs-com": buildFsComUrl,
+ supermicro: buildSupermicroUrl,
+ "alcatel-lucent": buildAlcatelLucentUrl,
+ "ale": buildAlcatelLucentUrl,
+ wistron: (_m) => null, // no public product pages
};
+// ── Generic marketing image detector ────────────────────────────────────────
+// Rejects URLs that are clearly stock photos, homepages, lifestyle shots or
+// any other non-product image. Patterns found from real-world scrapes.
+
+const GENERIC_IMAGE_PATTERNS: RegExp[] = [
+ // ── Logo / brand marks (never product photos) ────────────────────────────
+ /[-/_]logo[-_.]|\/logos?\//i,
+ /cisco[-_]?logo/i,
+ /juniper[-_]networks[-_]logo/i,
+ /arista[-_]?logo/i,
+ /brand[-_]?logo/i,
+ /company[-_]?logo/i,
+ // SVG logos often have these in path
+ /\/svg\//i,
+ /\.svg(\?|$)/i,
+ // ── Alcatel-Lucent Enterprise generic hero images ────────────────────────
+ /naas-homepag/i,
+ /al-enterprise.*\/images\/naas/i,
+ // ── Generic OG / social sharing defaults ─────────────────────────────────
+ /og[-_]default/i,
+ /default[-_](?:og|social|share|image)/i,
+ /site[-_](?:default|image|og)/i,
+ /social[-_](?:default|share)/i,
+ /twitter[-_]default/i,
+ /default[-_]thumbnail/i,
+ // ── Homepage / banner / lifestyle ────────────────────────────────────────
+ /\/homepage\//i,
+ /hero[-_](?:banner|bg|background|image)/i,
+ /banner[-_](?:bg|background)/i,
+ /lifestyle/i,
+ /stock[-_]?photo/i,
+ /people[-_](?:at|in|with)/i,
+ // ── Placeholder / fallback ────────────────────────────────────────────────
+ /placeholder/i,
+ /no[-_]?image/i,
+ /image[-_]?not[-_]?found/i,
+ /\/fallback[/-]/i,
+ /missing[-_]image/i,
+ // ── Generic about/press/brand pages ──────────────────────────────────────
+ /\/press[-_]kit/i,
+ /\/media[-_]kit/i,
+];
+
+function isGenericImage(url: string): boolean {
+ return GENERIC_IMAGE_PATTERNS.some((re) => re.test(url));
+}
+
// ── og:image extractor ──────────────────────────────────────────────────────
function extractOgImage(html: string, baseUrl: string): string | null {
+ const resolve = (url: string): string | null => {
+ if (!url) return null;
+ let abs = url;
+ if (url.startsWith("/")) {
+ try { abs = new URL(url, baseUrl).toString(); } catch { return null; }
+ }
+ if (!abs.startsWith("http")) return null;
+ if (isGenericImage(abs)) return null; // ← reject logos/marketing images
+ return abs;
+ };
+
// Primary: og:image
const ogM = html.match(/ with product hint
- const imgM = html.match(/]+src="([^"]+(?:product|hero|switch|router)[^"]*\.(?:jpg|jpeg|png|webp))"/i);
+ // Fallback: large product image in ![]()
with product keyword in path
+ const imgM = html.match(/]+src="([^"]+(?:product|switch|router|hardware)[^"]*\.(?:jpg|jpeg|png|webp))"/i);
if (imgM?.[1]) {
- try {
- const abs = new URL(imgM[1], baseUrl).toString();
- if (abs.startsWith("http")) return abs;
- } catch { /* ignore */ }
+ const resolved = resolve(imgM[1]);
+ if (resolved) return resolved;
}
return null;
diff --git a/sql/043-clear-generic-switch-images.sql b/sql/043-clear-generic-switch-images.sql
new file mode 100644
index 0000000..1969ffd
--- /dev/null
+++ b/sql/043-clear-generic-switch-images.sql
@@ -0,0 +1,29 @@
+-- Migration 043 — Clear generic / logo images from switches
+-- Removes image_url entries that are clearly logos or marketing photos,
+-- not actual product hardware images. The scraper will re-fetch with the
+-- improved isGenericImage() filter on the next daily run.
+
+UPDATE switches
+SET image_url = NULL, assets_scraped_at = NULL
+WHERE image_url IS NOT NULL
+ AND (
+ -- Generic logos (Cisco, vendor brand marks)
+ image_url ILIKE '%logo%'
+ OR image_url ILIKE '%.svg'
+ -- Alcatel-Lucent Enterprise lifestyle / naas hero images
+ OR image_url ILIKE '%naas-homepag%'
+ OR image_url ILIKE '%al-enterprise%naas%'
+ -- Generic OG defaults
+ OR image_url ILIKE '%og-default%'
+ OR image_url ILIKE '%default-social%'
+ OR image_url ILIKE '%og_default%'
+ -- Placeholder/fallback images
+ OR image_url ILIKE '%placeholder%'
+ OR image_url ILIKE '%no-image%'
+ OR image_url ILIKE '%noimage%'
+ -- Generic homepage images
+ OR image_url ILIKE '%/homepage/%'
+ -- Lifestyle / stock photos (common CDN path patterns)
+ OR image_url ILIKE '%lifestyle%'
+ OR image_url ILIKE '%stock-photo%'
+ );
'
- + '' + esc(t.part_number || t.standard_name || t.slug) + ''
- + '' + esc(t.reach_label || '') + priceStr + ''
- + fullyBadge + verBadge + foUrl
+ ch += '
';
});
-
- // Form-factor matches (spec_match) — collapsed summary
- if (foSpecMatch.length > 0) {
- ch += ''
+ + '' + esc(t.part_number || t.standard_name) + ''
+ + '' + esc(t.vendor_name || '') + ''
+ + (priceStr ? '' + priceStr + '' : '')
+ '
';
});
- if (items.length > 8) ch += '+' + (items.length - 8) + ' more Flexoptix options
';
+ if (items.length > 6) ch += '+' + (items.length - 6) + ' more
';
ch += ''
- + '
';
- }
}
- // ── ALL COMPATIBLE (other vendors) ────────────────────────────────────
- if (otherList.length > 0) {
- var verifiedOthers = otherList.filter(function(t) { return t.verification_method === 'vendor_matrix'; });
- var specOthers = otherList.filter(function(t) { return t.verification_method !== 'vendor_matrix'; });
-
- ch += '+ ' + foSpecMatch.length + ' more by form factor match
' - + ''
- + foSpecMatch.slice(0, 20).map(function(t) {
- return ''
- + esc(t.standard_name || t.part_number) + '';
- }).join('')
- + (foSpecMatch.length > 20 ? '+' + (foSpecMatch.length - 20) + ' more' : '')
- + '
Competitor Transceivers ' + otherList.length + ''
- + (verifiedOthers.length > 0 ? '(' + verifiedOthers.length + ' vendor-tested)' : '') + '
';
-
- // Show vendor-tested ones with price info
- if (verifiedOthers.length > 0) {
- var groups = {};
- verifiedOthers.forEach(function(t) {
- var key = (t.form_factor || '?') + ' ' + (t.speed || '?');
- if (!groups[key]) groups[key] = [];
- groups[key].push(t);
- });
- Object.keys(groups).sort().forEach(function(key) {
- var items = groups[key];
- ch += '' + esc(key) + ' (' + items.length + ')
';
- ch += '';
- items.slice(0, 6).forEach(function(t) {
- var priceStr = '';
- if (t.latest_price) {
- var _pAmt = parseFloat(t.latest_price);
- var _pCur = (t.latest_currency || 'USD').toUpperCase();
- var _pUSD = toUSD(_pAmt, _pCur);
- priceStr = _pUSD !== null ? fmtUSD(_pUSD) : _pCur + ' ' + _pAmt.toFixed(2);
- }
- ch += '
';
- });
- }
-
- // Show spec-match ones as compact chips
- if (specOthers.length > 0) {
- ch += ''
- + '' + esc(t.part_number || t.standard_name) + ''
- + '' + esc(t.vendor_name || '') + ''
- + (priceStr ? '' + priceStr + '' : '')
- + '
';
- });
- if (items.length > 6) ch += '+' + (items.length - 6) + ' more
';
- ch += '';
- ch += '
';
- }
+ // Spec-match as compact chips
+ if (specOthers.length > 0) {
+ ch += 'Form factor compatible
';
- ch += '';
- specOthers.slice(0, 20).forEach(function(t) {
- var fullyBadge = (t.fully_verified === true) ? '★ ' : '';
- ch += ''
- + fullyBadge + esc(t.standard_name || t.slug || t.part_number) + '';
- });
- if (specOthers.length > 20) ch += '+' + (specOthers.length - 20) + ' more';
- ch += '
';
+ ch += '
';
}
el('panel-content').insertAdjacentHTML('beforeend', ch);
diff --git a/packages/scraper/src/scrapers/switch-image-fetcher.ts b/packages/scraper/src/scrapers/switch-image-fetcher.ts
index ccd5603..4f96917 100644
--- a/packages/scraper/src/scrapers/switch-image-fetcher.ts
+++ b/packages/scraper/src/scrapers/switch-image-fetcher.ts
@@ -40,7 +40,6 @@ function buildCiscoUrl(model: string): string | null {
const slug = m.replace("N9K-C", "").toLowerCase().replace(/[^a-z0-9]/g, "-");
return `https://www.cisco.com/c/en/us/products/switches/nexus-${slug}-switch/index.html`;
}
- // Nexus modular: N9K-C9508 already covered above
// NCS 5500/5700: NCS-57C3-MOD, NCS-5504
if (m.startsWith("NCS-")) {
const num = m.replace("NCS-", "").toLowerCase().replace(/[^a-z0-9]/g, "-");
@@ -51,9 +50,20 @@ function buildCiscoUrl(model: string): string | null {
const slug = m.toLowerCase().replace(/[^a-z0-9]/g, "-");
return `https://www.cisco.com/c/en/us/products/switches/catalyst-${slug}/index.html`;
}
+ // Cisco 8000 SP series: 8101-32FH, 8202-32FH, 8608
+ if (/^8[0-9]{3}/.test(m)) {
+ const slug = m.toLowerCase().replace(/[^a-z0-9]/g, "-");
+ return `https://www.cisco.com/c/en/us/products/routers/8000-series-routers/${slug}/index.html`;
+ }
return null;
}
+function buildAlcatelLucentUrl(model: string): string | null {
+ // OmniSwitch 6900-X72, OmniSwitch 9900-C32D
+ const slug = model.toLowerCase().replace(/\s+/g, "-").replace(/[^a-z0-9-]/g, "");
+ return `https://www.al-enterprise.com/en/products/switches/${slug}`;
+}
+
function buildAristaUrl(model: string): string | null {
// 7060X6-64PE → https://www.arista.com/en/products/7060x6-series/7060cx6-64pe
// 7050CX3-32S → https://www.arista.com/en/products/7050x3-series/7050cx3-32s
@@ -147,52 +157,111 @@ function buildAsterfusionUrl(model: string): string | null {
// ── URL dispatcher by vendor slug ───────────────────────────────────────────
const URL_BUILDERS: RecordForm factor compatible
';
+ ch += '';
+ specOthers.slice(0, 20).forEach(function(t) {
+ var fullyBadge = (t.fully_verified === true) ? '★ ' : '';
+ ch += ''
+ + fullyBadge + esc(t.standard_name || t.slug || t.part_number) + '';
+ });
+ if (specOthers.length > 20) ch += '+' + (specOthers.length - 20) + ' more';
+ ch += '