Refactor the Hypothesis API communication code, to remove boilerplate and make changes easier. This is helpful for me to implement additional methods for bi-directional annotations syncing in my fork.
With this, annotations are always iterated through search_after (not limit), which fixes the edge case of missing annotations in later syncs. I wrangled the Hypothesis API for another project before, and this seems to be the best way to work around the pagination issues.
Refactor the Hypothesis API communication code, to remove boilerplate and make changes easier. This is helpful for me to implement additional methods for bi-directional annotations syncing in my fork.
With this, annotations are always iterated through
search_after(notlimit), which fixes the edge case of missing annotations in later syncs. I wrangled the Hypothesis API for another project before, and this seems to be the best way to work around the pagination issues.