Hi - is there a way to create the SQL for { it[name] = "foo" } and { it[director] = "bar") }, to then later combine them and call Movies.update() { it[name] = "foo") ; it[director] = "bar" }

It's unclear to me how the default connection works. The docs say the latest connection is the default. Does this truly mean that the latest call of

Database.connect

is the default db? This is inconvenient when working with Spring Boot because bean creation order is undefined.

@tapac thanks for the postgres index fix! Am I looking for a 0.26.2?

Hello, everyone! Long time no see. Sorry for "silence mode" while I was on vacation with almost no internet. But I'm back and plan to check all new issues and PR in a few days. There are a few bug fixes pending in a master branch, so let's start with a bug-fix release 0.26.2. Please, check release note to see the full list of changes.

Hi, I want to create a custom count, normally when I want to count a column I just

columnX.count()

in the slice part of the query. I want to do something like this

count(*) filter ( where columnX in ('a', 'b') ) as custom_count

is there an easy way to achieve this? Thanks 🙂

@tapac welcome back! Hopefully an easy question as you dive back into code - does the new composite column support nullability?

Hi, Thank you for creating Exposed, I'm having a great time learning the library 🙂 I was just wondering if it would be a good idea to change the

adjustSlice

function from

fun adjustSlice(body: ColumnSet.() -> FieldSet): Query = apply { set = set.source.body() }

to something like

fun adjustSlice(body: ColumnSet.(FieldSet) -> FieldSet): Query = apply { set = set.source.body(set) }

That way you can get a hold of fields that are currently a part of the fieldset for the query and do things like adding additional fields (or even removing fields). To me this was useful when I wanted to add a single extra field to the slice, depending on some condition. Currently I've created this extension function that I'm using, and it works quite well:

fun Query.addToSlice(vararg columns: Expression<*>): Query {
    return adjustSlice {
        slice(this@addToSlice.set.fields + columns)
    }
}

The JDBC layer is parsing the value to

BigDecimal

correctly but it appears that Exposed is rounding the precision to that of the original column.

Hi! I have a

UUIDTable

`Intervention`with a child:

val foreignEnterprise = reference("foreign_enterprise", ForeignEnterprises.id)

. I want to get the foreign enteprise for an intervention. I have the

InterventionUuid

. How can I do ? In SQL, I do: SELECT foreign_enterprise.* from foreign_enterprise, intervention WHERE intervention.foreign_enteprise_id = foreign_enterprise_id AND intervention.id ='`INPUT_UUID`'.

Hey guys! I'm currently working with Exposed and its a breeze! However, I stumbled upon a thing that I don't quite understand yet. I have the following table:

object Sensors : IntIdTable() {
    val roomId = reference("id", Rooms.id).nullable()
    val name = text("name")
}

And now I just want to select all rows and map them to my domain Object:

class Sensor(
        val id: Int?,
        val roomId: Int?,
        val name: String
)

Sensors.selectAll().map {
      Sensor(it[id], it[roomId], it[name])
}

However, if I do it as shown, I get an alert saying that

it[id]

is of type string (while I expect an int). Ist htere a better way? Why is it like that, even though my table definition is different? Thanks in advance!

Hi @tapac. When do you think to update exposed to kotlin 1.4? Thx

Exposed doesn't work with NoSQL DBs, correct?

the question is, how to create it?

How do I simply have a collection of strings in an object? (Using DAO) This basic use case doesn't seem to be catered for in the DAO guide. I can see how to do it with a full 'Entity' for each string, each with its ID and reference etc, but this is unnecessarily heavy - just a table with 'parentId, string' should be all that's necessary to implement the one-to-many relationship. I can't believe Exposed has missed this out...

Hi all I have defined a table with custom primary key:

object Genders : Table() {
    val value: Column<String> = varchar("value", 1)
    val description: Column<String> = varchar("description", 20)
    override val primaryKey = PrimaryKey(value)
}

As next, I would like to create an Entity class for it and I have tried:

class Gender(v: EntityID<String>) : Entity<String>(v) {
    companion object : EntityClass<String, Gender>(???)

    var value by Genders.value
    var description by Genders.description
}

How to create an entity without

transaction

and later do the transaction.

Hi, I've been working on a project using exposed for a bit and was wondering if there is any convenience functionality for audit fields like created & updated timestamps or soft delete logic with a deleted column? If so, is it documented anywhere? If not, are there any plans/issues for adding it, (but just super low priority)? Didn't want to reinvent the wheel here if I didn't have to.

What is the correct way to create a NULL valued expression? Let's say I want a branch of a case/when/else to result in NULL.

hi there. guys if we assume that we have some kind of sealed hierarchy of classes (with common fields in base classes, and some specific in the, subclasses)

sealed class A {
    abstract val a: Int

    data class B(val b: String, override val a: Int) : A()
    data class C(val c: Boolean, override val a: Int) : A()
}

persist those B,C classes to single table, with nullable b and c fields - make it any sense from exposed point of view? I’ve tried to google some samples for this but without success. thats why interesting to ask about it( maybe this idea some example of sick imagination of mine 😃)

Hello, is it possible to wrap row into entity, that was left joined and can be null?

Hi! Is it possible to get transaction context inside other thread? I have code like this:

transaction {
            FirstObject.selectAll()
            supplyAsync {
                runBlocking {
                    SecondObject.selectAll()
                }
            }
        }

I need to do both calls inside one transaction. When calling

SecondObject.selectAll()

, I got exception

No transaction in context

. https://github.com/JetBrains/Exposed/wiki/Transactions#working-with-coroutines -> This should probably help me, but I did not fully understand how to use it.

How do I make a column refer to multiple rows in a different table?

Hi! I am manually wrapping outer join result in DAO objects. How can I check that a column is null?

Is there a way to wrap a constant as a query? i.e.

SELECT 1

INSERT IGNORE...

throws an exception when using H2. I'm up-to-date with exposed and h2 drivers... Am I missing anything else?

Is it possible for me to somehow make two Tables with the same structure without copying the code?

Good Morning

Good night! It may be a getting started coroutine question, but since its subject is strictly related to exposed, maybe someone can help me with an expecifc use case insight. So if I do it like this:

fun addA() = transaction{A insert statement}
fun addB() = trabsaction{B insert statement}

suspend fun applyAandB() = coroutineScope {
    launch { addA() }
    launch { addB() }
}

would them execute asyncronously?

Speak up, guys! I'm putting together a clause where exposed this way

Employee.select { Employee.id = id and (Employee.name = name)}

is that correct?