Deduplicate media across scrapes and polish the viewer controls

docs/media-scraper.md

@@ -111,6 +111,8 @@ Downloads run after a finished walk when the scrape used `mode: "download"`:
   `<destination>/<scrape-id>/thumbnails/` while the local copy stays for the Media tab.
 - Downloaded videos get a poster frame (`ffmpeg -ss 0.5`, 480px WebP) cached in `thumbs/`, because
   Discord serves no thumbnails for uploaded video files.
+- The ffprobe pass fills missing dimensions and `duration_secs` after a download. Attachment
+  durations come from Discord; the 0 that Discord sends for non-audio files is treated as unknown.
 - Progress counters: `downloaded`, `downloadErrors`.
 
 Previews run after every walk:

docs/media.md

@@ -67,6 +67,10 @@ media-scrape/db/
 - The database is an index. It holds the scrape state, the messages seen, the media items, and the
   link between a media item and its source messages. Source channel, author, and timestamp are
   indexed so filters can be added later without a re-scrape.
+- Scrapes stay self-contained, but the API and UI **deduplicate by URL across scrapes**: one card
+  per URL, the newest scrape wins; stats count unique URLs too. Re-scraping a channel does not
+  duplicate anything in the grid or the counters. (Files are still stored per scrape, so the
+  on-disk size doubles with overlapping scrapes; a shared content store is a possible later step.)
 - Storage uses `node:sqlite` (experimental in Node 22, prints a warning). If its API changes,
   `better-sqlite3` is the fallback; it needs a native build, so that switch is an owner install.
 - The file name in `media/` is the SHA-256 of the bytes, never the original name. The original
@@ -147,9 +151,13 @@ that deletion is irreversible. The delete button is disabled for a 5 second coun
   tiles carry a translucent play icon top-right. Discord serves no thumbnails for uploaded videos,
   so a video without a cached preview renders a first-frame `<video>` in the grid, and downloaded
   videos get an ffmpeg poster cached in `thumbs/`. Clicking a tile opens a viewer: wide media
-  shows the metadata banner below, vertical media shows it to the right; the banner has the full
-  link (with an external-site warning) and **View full**, and the play action embeds a YouTube,
-  Vimeo, Twitch, or Streamable player when possible. The status row reads
+  shows the metadata banner below, vertical media shows it to the right, and videos carry a large
+  play icon top-right. The play action embeds a YouTube, Vimeo, Twitch, or Streamable player, or
+  plays a direct mp4/webm/m4v; when neither is possible it opens the external-site warning, since
+  the page is the only way to watch. **View full** opens the embedded media (player, local file,
+  or cached preview) in a new tab; the full link is a separate button that opens the original
+  page with the warning. Length comes from Discord's attachment duration and, for downloads, an
+  ffprobe probe. The status row reads
   `(Images/Videos) | Downloaded A/B | Cached C/B (excl Downl) | Loaded L/B`, where B is the total
   for the active filters. A scrape strip shows live progress with Stop.
   **New scrape…** (rightmost) opens a popup with Server and Group chat trees (servers expand to