Normalize YouTube Shorts links to watch URLs

media-scrape/src/extract.ts

@@ -36,6 +36,15 @@ export interface ExtractedMedia {
     title: string | null;
 }
 
+/**
+ * Shorts links break the embed player: it needs `watch?v=`. Rewrite them for storage so the
+ * viewer, "View full", and refreshes all use the canonical form. Other URLs pass through.
+ */
+export function normalizeYouTubeUrl(url: string): string {
+    const match = /^https?:\/\/(?:www\.|m\.)?youtube\.com\/shorts\/([A-Za-z0-9_-]{6,})/.exec(url);
+    return match ? `https://www.youtube.com/watch?v=${match[1]}` : url;
+}
+
 function extension(filename: string | null | undefined): string {
     const match = /\.([a-z0-9]+)$/i.exec(filename ?? "");
     return match ? match[1].toLowerCase() : "";
@@ -75,9 +84,11 @@ function hostname(url: string | null): string | null {
 
 function extractEmbed(embed: CoreEmbed): ExtractedMedia | null {
     const provider = embed.provider ?? hostname(embed.url);
+    const embedUrl = embed.url ? normalizeYouTubeUrl(embed.url) : null;
+    const videoUrl = embed.videoUrl ? normalizeYouTubeUrl(embed.videoUrl) : null;
     const common = {
         source: "embed" as const,
-        playUrl: embed.videoUrl ?? null,
+        playUrl: videoUrl,
         provider,
         downloaded: false,
         size: null,
@@ -91,7 +102,7 @@ function extractEmbed(embed: CoreEmbed): ExtractedMedia | null {
         return {
             ...common,
             kind: "video",
-            url: embed.url ?? embed.videoUrl,
+            url: embedUrl ?? videoUrl ?? "",
             contentType: null,
             previewUrl: embed.thumbnailUrl ?? embed.imageUrl
         };
@@ -102,7 +113,7 @@ function extractEmbed(embed: CoreEmbed): ExtractedMedia | null {
         return {
             ...common,
             kind: "gif",
-            url: embed.url ?? embed.imageUrl ?? "",
+            url: embedUrl ?? embed.imageUrl ?? "",
             contentType: null,
             previewUrl: embed.thumbnailUrl ?? embed.imageUrl
         };
@@ -112,7 +123,7 @@ function extractEmbed(embed: CoreEmbed): ExtractedMedia | null {
         return {
             ...common,
             kind: "image",
-            url: embed.url ?? embed.imageUrl,
+            url: embedUrl ?? embed.imageUrl,
             contentType: null,
             previewUrl: embed.thumbnailUrl ?? embed.imageUrl
         };

media-scrape/src/store.ts

@@ -7,7 +7,7 @@ import { mkdirSync, readdirSync, readFileSync, statSync, writeFileSync } from "n
 import { join } from "node:path";
 import { DatabaseSync } from "node:sqlite";
 import { fileURLToPath } from "node:url";
-import { extractMedia } from "./extract.js";
+import { extractMedia, normalizeYouTubeUrl } from "./extract.js";
 import { aspectCondition, resolutionCondition } from "./filter-presets.js";
 import { emptyMediaCounts, type CoreMessage, type MediaKind, type ScrapeMeta } from "./types.js";
 
@@ -281,6 +281,86 @@ export class ScrapeStore {
             this.db.exec("ALTER TABLE media ADD COLUMN title_attempts INTEGER NOT NULL DEFAULT 0");
             this.db.exec("ALTER TABLE media ADD COLUMN title_error TEXT");
         }
+
+        this.migrateYouTubeShorts();
+    }
+
+    /**
+     * One-time rewrite of youtube.com/shorts links to watch URLs. The embed player needs `v=`,
+     * so old rows with a shorts URL would never play. `PRAGMA user_version` makes it idempotent
+     * per scrape database.
+     */
+    private migrateYouTubeShorts(): void {
+        const version = this.db.prepare("PRAGMA user_version").get() as { user_version?: number } | undefined;
+        if ((version?.user_version ?? 0) >= 1) return;
+
+        const rows = this.db
+            .prepare(
+                "SELECT * FROM media WHERE url LIKE '%youtube.com/shorts/%' OR play_url LIKE '%youtube.com/shorts/%'"
+            )
+            .all() as unknown as MediaRow[];
+
+        if (rows.length) {
+            const update = this.db.prepare("UPDATE media SET url = ?, play_url = ? WHERE id = ?");
+            const findTarget = this.db.prepare("SELECT id FROM media WHERE url = ? AND id != ?");
+            const merge = this.db.prepare(
+                `UPDATE media SET
+                    preview_hash = COALESCE(preview_hash, ?),
+                    preview_bytes = COALESCE(preview_bytes, ?),
+                    preview_ext = COALESCE(preview_ext, ?),
+                    file_path = COALESCE(file_path, ?),
+                    file_hash = COALESCE(file_hash, ?),
+                    file_bytes = COALESCE(file_bytes, ?),
+                    downloaded = MAX(downloaded, ?),
+                    title = COALESCE(title, ?),
+                    title_key = COALESCE(title_key, ?),
+                    play_url = COALESCE(play_url, ?)
+                 WHERE id = ?`
+            );
+            const relink = this.db.prepare(
+                "UPDATE OR IGNORE media_sources SET media_id = ? WHERE media_id = ?"
+            );
+            const dropLinks = this.db.prepare("DELETE FROM media_sources WHERE media_id = ?");
+            const drop = this.db.prepare("DELETE FROM media WHERE id = ?");
+
+            this.db.exec("BEGIN");
+            try {
+                for (const row of rows) {
+                    const url = normalizeYouTubeUrl(row.url);
+                    const playUrl = row.play_url ? normalizeYouTubeUrl(row.play_url) : null;
+                    const target = findTarget.get(url, row.id) as { id: number } | undefined;
+
+                    // The same video was already indexed under its watch URL: fold this row in.
+                    if (!target) {
+                        update.run(url, playUrl, row.id);
+                        continue;
+                    }
+
+                    merge.run(
+                        row.preview_hash,
+                        row.preview_bytes,
+                        row.preview_ext,
+                        row.file_path,
+                        row.file_hash,
+                        row.file_bytes,
+                        row.downloaded,
+                        row.title,
+                        row.title_key,
+                        playUrl,
+                        target.id
+                    );
+                    relink.run(target.id, row.id);
+                    dropLinks.run(row.id);
+                    drop.run(row.id);
+                }
+                this.db.exec("COMMIT");
+            } catch (error) {
+                this.db.exec("ROLLBACK");
+                throw error;
+            }
+        }
+
+        this.db.exec("PRAGMA user_version = 1");
     }
 
     close(): void {