Scanning a page

last updated

When a page holds many images and no supported extractor - a forum thread, a blog post, an image dump - scan it and pick what to keep.

Open the popup and click scan page. The chooser opens docked beside the page (Chrome’s side panel, Firefox’s sidebar), so it stays open while you scroll the page to cross-check.

Some pages cannot be read at all - browser-internal pages, the add-on store, the built-in PDF viewer. The chooser says “this page cannot be scanned” for those, which is a different answer from “no images found” on a page it read and came up empty on.

Almost everything a scan finds is a direct file URL, so scanned images arrive in monbooru untagged (see tagged or untagged).

What the chooser lists#

Two filters run before anything is shown:

  • Images smaller than the min image size setting (64 pixels by default) are dropped, which removes icons, buttons and tracking pixels. Set it to 0 in settings to see everything.
  • File types monbooru cannot store are left out: the chooser lists jpg, png, gif and webp images and mp4 and webm video, nothing else. A URL with no file extension is still listed, since most CDN images carry none and only monloader can tell.

A scan lists at most the configured scan cap (100 by default, up to 1000). When a page holds more, the header reads “scan: 100 of 350 images” so a truncated scan is never mistaken for the whole page.

If some picks fail to queue, the message under the send button says why and the selection is kept, so you can send again once the cause is fixed - re-sending the ones that made it is harmless, they come back as duplicates.

Picking a resolution#

When the page offers an image in several sizes, the card shows one token per resolution (“1500w”, “800w”, “src”). The largest is selected by default. A token reading “src” is one the page gave no width for; pick it and it relabels itself with the real width once that version loads.

Previews#

Thumbnails load straight from wherever the images live, by your browser, with three guards: no referrer is sent, they load lazily as you scroll, and the grid asks for the smallest version each image offers - while the card still sends the big version you picked.

If you would rather the chooser contact no outside host at all, turn off the previews checkbox in settings. Cards then show the file name instead of a thumbnail, and picking works the same.

A blank tile marked “preview blocked” means the host refused to serve the image outside its own site (hotlink protection). You can select and send a blocked card anyway; monloader may or may not be able to fetch it, and the queue outcome will tell you.