Steampipe.io lets you run sql against a variety of services, including hypothes.is. This strikes me as a nice way to perhaps develop a workflow from hypothesis to obsidian (there are any number of ways one could do this).

These are my notes for getting it up and running.
At the terminal, with brew installed (running on mac)

brew tap turbot/tap
brew install steampipe
steampipe -v
steampipe plugin install hypothesis

Then add my developer key for hypothesis to the config file for steampipe, which lives at

~/.steampipe/config/hypothesis.spc

There is a dashboard for hypothesis at https://github.com/turbot/steampipe-samples/tree/main/all/hypothesis . Copy those files to text editor, save with .sp extension. Then at the terminal, fire ’em up with

steampipe dashboard`

Queries can be run at the query prompt, via

steampipe query

. Or you can save your queries as a file, and pipe the results to output. So, I make an empty file and put this query into it:

select
uri,
tags,
exact,
text
from
hypothesis_search
where
query = 'tag=media&tag=review';

I save it as `query1`, no file extension. Then, at the terminal prompt,

$ steampipe query query1 > output.csv

The resulting output file actually uses the `|` character as a separator, so once I remember to specify that when opening in Excel, a lovely file of annotations. I have to run now, but I can see opening this file in Obsidian.md and then refactoring it so that I end up with one note per annotation.

references
https://hub.steampipe.io/plugins/turbot/hypothesis/tables/hypothesis_search
https://hub.steampipe.io/plugins/turbot/hypothesis

 

 

…god I hate the gutenberg editor here on wordpress.com. It sucks…