Preserving Jules Jordan base photos in deep scrape.

src/scrapers/julesjordan.js

@@ -51,28 +51,16 @@ function scrapeAll(scenes, site, entryIdFromTitle) {
 				const prefixedSrc = qu.prefixUrl(src, site.url);
 
 				if (src) {
-					return [
-						{
-							src: prefixedSrc.replace(/.jpg$/, '-full.jpg'),
-							referer: site.url,
-							verifyType: 'image', // sometimes returns 200 OK with text/html instead of 403
-						},
-						{
-							src: prefixedSrc.replace(/-1x.jpg$/, '-4x.jpg'),
-							referer: site.url,
-							verifyType: 'image',
-						},
-						{
-							src: prefixedSrc.replace(/-1x.jpg$/, '-2x.jpg'),
-							referer: site.url,
-							verifyType: 'image',
-						},
-						{
-							src: prefixedSrc,
-							referer: site.url,
-							verifyType: 'image',
-						},
-					];
+					return Array.from(new Set([
+						prefixedSrc.replace(/.jpg$/, '-full.jpg'),
+						prefixedSrc.replace(/-1x.jpg$/, '-4x.jpg'),
+						prefixedSrc.replace(/-1x.jpg$/, '-2x.jpg'),
+						prefixedSrc,
+					])).map((source) => ({
+						src: source,
+						referer: site.url,
+						verifyType: 'image',
+					}));
 				}
 
 				return null;
@@ -209,7 +197,21 @@ async function scrapeScene({ html, query }, context) {
 	if (argv.jjFullPhotos) {
 		release.photos = getPhotos(query, release, context);
 	} else {
-		release.photos = query.imgs('#images img');
+		// base release photos are usually better, but deep photos have additional thumbs
+		// the filenames are not chronological, so sorting after appending only worsens the mix
+		release.photos = [
+			...context.baseRelease?.photos?.map((sources) => sources.at(-1).src) || [],
+			...query.imgs('#images img'),
+		].map((source) => Array.from(new Set([
+			source.replace(/.jpg$/, '-full.jpg'),
+			source.replace(/-1x.jpg$/, '-4x.jpg'),
+			source.replace(/-1x.jpg$/, '-2x.jpg'),
+			source,
+		])).map((fallbackSource) => ({
+			src: fallbackSource,
+			referer: context.entity.url,
+			verifyType: 'image',
+		})));
 	}
 
 	if (query.exists('.update_dvds a')) {