feat: EXPOSED-258 Enhance upsert to allow exclusion of columns set on conflict

[bog-walk]

Currently, if the onUpdate parameter of upsert() is specified, this argument will be used in the update clause of the upsert statement. If the argument is left null, a duplicate of the column-value assignments from the insert body block will be used instead.
So if a user wants to only update a subset of columns, or not update columns with default values for examples, the only option is to provide an argument to onUpdate that duplicates the insert block:

object TestTable : Table("tester") {
    val id = integer("id").uniqueIndex()
    val created = datetime("created").defaultExpression(CurrentDateTime)
    val updated = datetime("updated").defaultExpression(CurrentDateTime)
    val name = varchar("name", 32)
    val age = integer("age")
}

// for example, column "created" should never be updated
val updateValues = listOf(
    TestTable.updated to CurrentDateTime,
    TestTable.name to stringParam("B"),
    TestTable.age to intParam(100)
)
TestTable.upsert(onUpdate = updateValues) {
    it[id] = 1
    it[name] = "B"
    it[age] = 100
}

This can become verbose, so an additional parameter, onUpdateExclude, has been introduced, which duplicates all insert column-value pairs except those for the specified columns:

// the above example is reduced to this
TestTable.upsert(onUpdateExclude = listOf(TestTable.created)) {
    it[id] = 1
    it[name] = "B"
    it[age] = 100
}

This parameter could also be used in the reverse case when the number of columns to update is small but the user wants to avoid duplicating the insert values:

// for example, only column "updated" should be updated
// could provide a set of all other columns or
TestTable.upsert(onUpdateExclude = TestTable.columns - TestTable.updated) {
    it[id] = 1
    it[name] = "B"
    it[age] = 100
}

Next steps:

• Update documentation
• Update Wiki once released