Skip to content

Init thumbnail rework. - #145

Open
markpbaggett wants to merge 7 commits into
mainfrom
issue-141
Open

Init thumbnail rework.#145
markpbaggett wants to merge 7 commits into
mainfrom
issue-141

Conversation

@markpbaggett

@markpbaggett markpbaggett commented Sep 18, 2025

Copy link
Copy Markdown
Collaborator

What Does This Do?

Skips Cantaloupe for audio / video thumbnails and fixes Issue-141.
Manifest generation was slow for items with many audio/video files (e.g. taboca_201002_03, 251 canvases) because every thumbnail (including AV thumbnails) were routed through Cantaloupe's IIIF image service via create_thumbnail_from_iiif. For AV items this is unnecessary round-trip overhead per file and a needless dependency on the image server for content that isn't an image derivative.

This change runs things through addThumbnails based on mediatype:

  • movies, audio, etree: build the thumbnail directly from a static archive.org/download/... URL instead of calling Cantaloupe. Movies get no explicit width/height; audio/etree get width/height from file metadata if present, otherwise a 192×108 default.
  • Everything else (images, texts, etc.): unchanged. It still goes through Cantaloupe, but now falls back to the same static-URL approach

@markpbaggett
markpbaggett marked this pull request as ready for review September 18, 2025 20:39
@glenrobson

Copy link
Copy Markdown
Collaborator

Audio we think we can hard code dimensions

Video we are thinking of removing height and width and testing with viewers as this is a 'should' in the spec.

@glenrobson

Copy link
Copy Markdown
Collaborator

Discussing this in the 5th march call and wondered if the video thumbnail would be the same size as the video. We must have the height and width for the video as its required for the canvas.

Audio should be static sizes.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

Look at audio + video thumbnails to see if we can make the generation of manifests quicker.

2 participants