Parsing @RoamResearch JSON exports enables us to do lot's of fun stuff. Starting with simple things, like detecting duplicates 
Of course we can export the text of a page, resolving both (()) block references and {{embed: ()}} blocks.
We can of course also get all the linked references.
Oh you want to see all the tags, and run some queries? :) More to come. github.com/houshuang/roam… @Conaw @wirtzdan
How about being able to insert auto-completed [[links]], but also ((blocks)) and {{embed: ((blocks))}} anywhere, in any app? Why not!
Today I decided to play around with @RoamResearch queries. I had a toy implementation before, but it wasn't actually parsing proper Roam queries, so that was first project. I'm sure there are better CS algo approaches, but fun to write from scratch.
Then for the implementation, actually easier. Including helper functions, just 100 lines. We get all the usual goodies, and, or, between, not etc.
In the last ten minutes, I decided to throw in a few more. Let's start with searching for instapaper blocks added today. There were not put on Daily Page, don't have any [[date links]]. But still works, because we're looking at the updated-at property.
Let me try to narrow down a bit. I only want blocks with highlights, or with bold text.
Or maybe I want to run a substring search as well. Let's see if Conor has ever talked about smoking. github.com/houshuang/roam…
Still pretty hacky, and I'm sure there are queries that would break it etc, but this both enables us to process queries on pages when processing Markdown exports etc, but also to experiment with new kinds of queries... (Or exporting properties to CSV files etc)...
