28683 Sample paper from JSON data

[carolinepaulic]

Description

This covers the bulk of https://github.com/allenai/scholar/issues/28339

Changes in this PR:

• Created new data dir where our sample data lives
  • Added a JSON file with a shortened version of an annotated paper (just 3 citations)
  • Added a JSON file with citation paper data for one of the paper IDs from the sample paper
• Reader updates:
  • On page load, it kicks off async call to grab sample paper JSON data
  • Once paper data is retrieved, it "loads" annotations by grabbing the raw entities from our data and massaging it into a nicer format. We map each page number to its annotations so that the Reader can just loop through each page and easily get the appropriate annotations from the map. Currently citations are the only entity supported.
• CitationPopover updates:
  • On citation click, it kicks off an async call to grab sample citation JSON data
  • Displays 'Loading...' while we wait for data to return
  • Populates popover once we have data
  • Truncated paper abstracts to 300 chars because Ant popover placement gets messed up when the popover contents are too long
• CitationsDemo updates:
  • Removed hardcoded sample data
  • Generates CitationPopovers based on new paperAnnotated prop
  • Generates CitationPopovers based on new annotations prop
• Updated HighlightOverlayDemo, TextHighlightDemo, and ScrollToDemo since we're using a different sample paper now
• Added new BoundingBox function boundingBoxRawToScaled to transform raw bounding box coordinates (expressed as a percentage of page size) into coordinates that make sense for our current page size
• Added new BoundingBox function scaleRawBoundingBoxWithContext- this is used to transform raw bounding box coordinates (expressed as a percentage of page size) into coordinates that make sense for our current page size, scaling, and rotation
• Data model updates:
  • Updated types related to entities, citations, and papers. "Raw" types represent data that is coming from our data source and non-raw types are formatted for the UI.
  • Turned BoundingBox page attribute from optional to required
  • Upgraded a lot of types up to classes because we need the constructors

Testing Plan

• Manual testing
  • New PDF loads
  • Existing functionality works as expected (zoom, rotate, outline, highlight overlay demo, text highlight demo, scroll to demo)
  • Citation popovers lazy load data, display 'Loading...' while waiting and update automatically when data comes in
  • Added console logs in various rendering/data retrieval functions to ensure that components are not rendering or grabbing new data unnecessarily
• Ran all unit tests

[carolinepaulic]

@smitar I updated the Paper/Annotation data types and simplified the Reader effects that set the paper and annotation data based on your suggestions.

@mjlangan I made BoundingBox.page required and fixed the key issue.