issue_comments

This didn't quite work: the column type is incorrect, so the foreign key relationship isn't sticking:

https://github-to-sqlite.dogsheep.net/github/repos?organization=53015001

[organization] TEXT REFERENCES [users]([id]) - should be INTEGER.

The problem is that if the first repo inserted has no organization it's set to null, which sqlite-utils derives as a TEXT column.

One solution would be to create the column explicitly with a type, but this could get messy.

I think I want a new sqlite-utils feature for this instead.

Reminder: the current list of arguments for those methods is: https://github.com/simonw/sqlite-utils/blob/13528faa817d79bc3900d3af7473300686b145d7/sqlite_utils/db.py#L951-L967

The syntax for explicitly creating a table looks like this: python db["cats"].create({ "id": int, "name": str, "weight": float, }, pk="id") So using a dictionary of column name -> Python type here would be good.

I think it becomes a new optional argument to .insert_all() and .insert() and .upsert_all() and .upsert(). Question is, what should that argument be called?

Options:

• columns
• types
• type_overrides
• extra_columns

I like the first two best. I'm leaning towards columns - it would serve two purposes: you can use it to ensure extra columns are created, and you can use it to specify the type on extra OR deteced columns.

Maybe something like this:

python db["authors"].insert_all( [{"id": 1, "name": "Sally", "score": "2"}], pk="id", not_null={"name", "score"}, defaults={"score": 1}, columns={"score": int, "weight": float} )

I'm going to make organization another foreign key to the users table just in case it IS possible (maybe with GitHub Enterprise or similar?)

Is it possible for a repo to have an owner that differs from its organization?

Turns out the main problem was #26 - now fixed.

Aha! it was missing from the fetch_repo() function.

I'm already doing this here: https://github.com/dogsheep/github-to-sqlite/blob/c4aaa50e167cfa9021c7c94260bc3e89e10947bf/github_to_sqlite/utils.py#L246-L250

Documentation here: https://sqlite-utils.readthedocs.io/en/stable/python-api.html#listing-rows

Released in 2.6

I think a neat way to do this would be with an optional argument for .rows_where():

python rows = db["table"].rows_where("age > 10", order_by="age desc")

If you want everything you can use this:

python rows = db["table"].rows_where(order_by="age desc")

It's a tiny bit weird calling .rows_where() without a where clause, but I think it makes sense here - especially since .rows is a property that can't take any arguments - though under the hood it actually does this:

https://github.com/simonw/sqlite-utils/blob/ad6ac19470a67867b96cb4c086450b8e4e46bf02/sqlite_utils/db.py#L436-L443