VEYNDO
Handbook
frontend/src/shared/media/README.mdView source ↗

English source · 1 min

Shared media captions

  • MediaCaptionEditor.tsx edits the selected content language and reviews a bounded AI proposal. Image, language, content and request changes discard stale responses. The native product, page or app editor owns saving, revision and publication.
  • caption-i18n.ts contains EN/DE/FR/ES interface copy; enabled content locales come from ContentLanguage.
  • captions.css fits existing media inspectors without a second workspace or AI conversation owner.

Captions are visible text, separate from accessibility alt text. merchant.media.caption uses central inference, current assistant permission and enabled shop locales. Context-only proposals are explicitly labelled. No generation call saves or publishes content.

product-media.ts owns the shared typed image/model distinction and scoped model URLs. ModelViewer.tsx loads the pinned viewer on demand, with localized controls and error recovery. model-decoder-disabled.ts excludes an unused WebAssembly decoder; native admission requires uncompressed embedded GLB. model-i18n.ts owns interface and accessibility strings.