.iterate does this que up results of a query in memory?

[LRagji]

The reason i choose for .iterate() over .all() is simply cause i could stream result set from disk to lets say http response without caching the response in the server memory and bloating it..

Some of non scientific tests that i have done are pointing that .iterate() method is caching some results into memory if not all, question is can we have an option to not cache any results in process memory and stream the results from the disks?, another connected question is what happens to read concurrency when such a iterate method is active, ideally it should go down unless i am missing something ..

A new bee to sqlite so please answer accordingly..

[JoshuaWise]

You could benefit from reading this: #406

[LRagji]

This is perfectly my problem thanks for pointing to the thread....

But I still don't understand how much exact memory is required by iterate? It should be irrespective of results ie row count is what I feel can you shed some light on this? Relative or BigO notation should be fine..

My question comes from a project where I am trying to open 10 different sqlite database on worker threads and doing a k-way merge between result set, and I just see my memory multiply everytime maybe it's the order by clause, but even after removing it I need to estimate what would one iterate cost in terms of memory..

Maybe like all requested the api documentation should be updated atleast with the thread link...