feat: support Document type on update/remove etc

[Cologler]

Currently:

table = ...
items = table.all()
... # work with items, than:
for item in items:
    table.update(item, doc_ids=[item.doc_id])

Expected:

table = ...
items = table.all()
... # work with items, than:
for item in items:
    table.update(item)

[msiemens]

That's an interesting idea! When implementing the update method I had SQL's UPDATE statement in mind where you either pass a query (WHERE) to update selected rows or skip the WHERE argument to update all rows. Although this is a nice idea I'm not sure whether it deviates too much from general expectations.

One could use a replace method instead to replace one or more documents but this might bloat the API too much…

[Cologler]

SQLite has similar update stmt insert or replace values (...) to update single row without where query repr.

[montarion]

I too would like this

[stale]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Feel free to reopen this if needed. Thank you for your contributions ❤️

[benetherington]

I don't think a separate doc_ids argument is too much of a stretch here, but I do really like how clean that example code is! @Cologler, would you be willing to use upsert(item) instead? I just put in PR #395 which allows this, and it would mimic the insert or replace statement you mentioned.

If upsert won't do, I'd be happy to work on a replace method, Marcus. It could just be an alias for update(document, doc_ids=[document.doc_id]). I don't think it's a good idea to roll this into update() itself as the conditional logic would really dive into the weeds.

[msiemens]

If upsert won't do, I'd be happy to work on a replace method, Marcus.

I think upsert is the best way to support this feature as it matches the desired semantics most closely 🙂

[Cologler]

@benetherington @msiemens

update equals update or failed, upsert equals update or insert, which is different.

In most cases, I insert a dict into tinydb, then load as Document from tinydb, do some changes, finally update back to the database. If I use SQLite, I always use UPDATE instead of UPSERT to write back, this is strange that I should use upsert to update it on tinydb.

The doc_id is generated by the tinydb in most cases, and I never manually create a Document instance, they usually created by the tinydb when I query it.

[Cologler]

This issue is about what we can do after we load an entity from a tinydb, and update and remove are two basic operations of CRUD.