Feature to save annotations in JSON format following the Web Annotation Data Model

[jaime-rivas]

Hi everyone,

@marco-c @calixteman @Snuffleupagus it's absolutely fantastic that you are adding editing capabilities to pdf.js to bring full annotation support, big thank you!

Giving this current effort, I wanted to suggest a related capability that might be very useful and hopefully achievable at the same time. In addition to supporting saving annotations in the native PDF annotation format inside the PDF, it would be great to have the option to save annotations outside the PDF in JSON format following the Web Annotation Data Model.

This will allow annotation libraries like Apache Annotator to externally load those annotations, enable collaborative annotations scenarios, etc. using the W3C’s standard Web Annotation data model. Related issue here: Fragment selector for PDFs + PDF.js integration

Thank you again!

[Snuffleupagus]

[...] that you are adding editing capabilities to pdf.js to bring full annotation support

I don't believe it's going to be "full" editing support, given the sheer number of Annotation-types that the PDF specification support, but most likely rather a few commonly used types.

[...] it would be great to have the option to save annotations outside the PDF in JSON format following the Web Annotation Data Model.

It sounds like that could very well require a fair amount of code, for a use-case that's potentially not entirely relevant for the Firefox built-in PDF Viewer. (Without having really looked at the linked specification...)

However, note that you're already able to access modified Annotation-data through the PDF.js API (but please keep in mind that editing is still unfinished/unreleased); refer to:

• The AnnotationStorage-implementation in https://github.com/mozilla/pdf.js/blob/master/src/display/annotation_storage.js
• Accessing the AnnotationStorage-instance, for the current PDF document, in

  pdf.js/src/display/api.js

  Lines 724 to 729 in b9f5a70

  	/**
  	* @type {AnnotationStorage} Storage for annotation data in forms.
  	*/
  	get annotationStorage() {
  	return this._transport.annotationStorage;
  	}

Hence it'd likely be possible to implement the JSON-exporting you mention outside of the PDF.js library as well, since as outlined above it's possible to access the "raw" data for modified Annotations.

[jaime-rivas]

Thank you @Snuffleupagus!

We will work on an external library to access AnnotationStorage and exporting the annotations. We will keep this issue updated when we make progress.

[Snuffleupagus]

Let's close this as WONTFIX for now, for the reasons outlined in #15055 (comment).