N4v41

i did some tests and a query like this will work:

to do taks

SELECT * FROM blocks WHERE hpath LiKE '/daily note/2023/sep%'  AND markdown LIKE '%[ ]%' AND subtype = 't' AND type = 'i';

done tasks

SELECT * FROM blocks WHERE hpath LiKE '/daily note/2023/sep%'  AND markdown LIKE '%[X]%' AND subtype = 't' AND type = 'i';

what is exactly you doubt ? a full tutorial would require a r2 subscription wich i don't have.

in simple steps:

1. go to the cloudflare dashboard
2. create a bucket
3. create an token to authenticate on the s3 api: https://developers.cloudflare.com/r2/api/s3/tokens/
4. go to siyuan configure the s3 sync with the information gathered on previous steps
5. after configuring the conection information go to cloud sync dir and add a directory for the sincronization

in general i think that under normal use "not adding/modifying more than 20k assets/documents per day" the billing probably would stay on the free tier.
for comparison i have a minio instance, my bucket has 615MB with a total count of 23,305 objects,
on R2 this would stay on the free tier
if the chages are much frequent i would recommend increase the sync time from the default 30 second to avoid surpassing the 1mm write operations on the free tier.