need consolidate result , consolidate result need primary key. Cannot assume partial model has the pk.

You are right. That's the reason why into_model is private, I reckon.

@billy1624 please review as well

need consolidate result , consolidate result need primary key. Cannot assume partial model has the pk.

You are right. That's the reason why into_model is private, I reckon.

I have an a bit complex idea.

If a model have 2 primary key

struct Model{
    #[sea_orm(primary_key)]
    id1: i32,
    #[sea_orm(primary_key)]
    id2: i32,
    ...
}

then, might the derive macro can generate the following PartialModel

#[derive(DerivePartialModel,  FromQueryResult)]
#[sea_orm(entity = "Entity")]
struct PrimaryPartial{
    id1:i32,
    id2:i32
}

finally, I can use a PartialModel like following, the pk will be the model's primary key partial model, and the inner will be the partial model user want. We can using pk to consolidate result

#[derive(FromQueryResult, DerivePartialModel)]
struct WithPk<E: EntityTrait, P: PartialModelTrait> {
    #[sea_orm(flatten)]
    pk: <<E as EntityTrait>::PrimaryKey as PrimaryKeyTrait>::PartialModel,
    #[sea_orm(flatten)]
    inner: P,
}

However, those code cannot using in current sea-orm, because neither FromQueryResult nor DerivePartialModel not support flatten and DerivePartialModel not support any generic argument.

DerivePartialModel now support nest. It look like that

    #[derive(FromQueryResult, DerivePartialModel)]
    #[sea_orm(entity = "Entity")]
    struct TimeRecord {
        _access: DateTimeLocal,
        _modify: DateTimeLocal,
        _delete: Option<DateTimeLocal>,
    }

    #[derive(FromQueryResult, DerivePartialModel)]
    #[sea_orm(entity = "Entity")]
    struct Bitmap {
        _bitmap1: u64,
        _bitmap2: u64,
        _bitmap3: u64,
        _bitmap4: u64,
    }
    #[derive(DerivePartialModel)]

    struct NestPartialModel {
        #[sea_orm(flatten)]
        _record: TimeRecord,
        #[sea_orm(flatten)]
        _bitmap: Bitmap,
    }
    // manual implement
    impl FromQueryResult for NestPartialModel {
        fn from_query_result(
            res: &sea_orm::QueryResult,
            pre: &str,
        ) -> Result<Self, sea_orm::DbErr> {
            Ok(Self {
                _record: TimeRecord::from_query_result(res, pre)?,
                _bitmap: Bitmap::from_query_result(res, pre)?,
            })
        }
    }

It generate SQL like following(MySQL)

SELECT 
`foo_table`.`access`, `foo_table`.`modify`, `foo_table`.`delete`, 
`foo_table`.`bitmap1`, `foo_table`.`bitmap2`, `foo_table`.`bitmap3`, `foo_table`.`bitmap4` 
FROM `foo_table`

I think it is a good idea to support nest and generic argument for DerivePartialModel .

Just being curious, why do you have underscore in front of all attributes?

Just being curious, why do you have underscore in front of all attributes?

The code snippet is in derive_test, without it will result warning.

Yeah flatten makes sense. However, out of the scope of this PR, this is useful only if FromQueryResult supports flatten as well. So I would say let's merge everything up to 75fbe44, and add flatten in another PR. We will then tackle both FromQueryResult and DerivePartialModel there.

i.e. decoupling them would make them easier to review individually

Yeah flatten makes sense. However, out of the scope of this PR, this is useful only if FromQueryResult supports flatten as well. So I would say let's merge everything up to 75fbe44, and add flatten in another PR. We will then tackle both FromQueryResult and DerivePartialModel there.

i.e. decoupling them would make them easier to review individually

Aright, the flatten part and generic support part has been removed. I will open another PR to add flatten support on DerivePartialModel and FromQueryResult

🎉 Released In 0.12.1 🎉

Thank you everyone for the contribution!
This feature is now available in the latest release. Now is a good time to upgrade!
Your participation is what makes us unique; your adoption is what drives us forward.
You can support SeaQL 🌊 by starring our repos, sharing our libraries and becoming a sponsor ⭐.