R2dbcConverter should provide read functionality that accepts alias prefixes for columns

[pmaedel]

There is an example project available here

I used the r2dbc mapping functionality to map joined entities via alias prefixes:

    override suspend fun findByIdJoined(id: Long): Fruit? {

        return databaseClient.execute("select f.*, t.id as t_id, t.kind as t_kind from fruit f left join tree t on t.id = f.tree_id where f.id = :id")
            .bind("id", id)
            .map { row, rowMetaData ->
                val fruitEntity = converter.read(Fruit::class.java, row, rowMetaData)
                val joinedTreeEntity = converter.read<Tree>(row, "t_")
                fruitEntity.copy(joinedTree = joinedTreeEntity)
            }.awaitFirstOrNull()
    }

    private inline fun <reified T> R2dbcConverter.read(row: Row, aliasPrefix: String): T = read(T::class.java,
        object : Row {
            override fun <T : Any?> get(index: Int, type: Class<T>) = row[index, type]
            override fun <T : Any?> get(name: String, type: Class<T>) = row["$aliasPrefix$name", type]
        }
    )

I can easily map the root entity (tree) via read(type, row, rowMetadata). Mapping the aliased joined entity (fruit) however seems only to be possible by hacking a customized read functionality, which would better be provided via R2dbcConverter. I have seen that functionality for reading values with column prefixes is available internally via MappingR2dbcConverter#readFrom but could not find a way to make use of it for my use case.

[mp911de]

We already use a prefix internally and with #288, we plan to extend on reading objects with a prefix applied so it makes sense to surface the functionality a bit more. We should extend EntityRowMapper and R2dbcConverter to accept both column prefixes.