MongoDB Scala Driver - Rendering BSON Documents

2019-05-11 16:59发布

问题:

We currently have a Type-Safe query language at work, this custom DSL allows us to easily write database queries that are interpreted and converted into Mongo Queries.

We recently swapped over from Casbah to the new Mongo Scala Driver and rewrote our interpreter. I am however having some issues when dealing with optional values.

This is an example query:

dao.headOption(Order.id === orderId.some)

The type stored on the Order Object is an Option so we lift the provided id into an option as well. However whenever I try to render out the generated query for debugging as well as for test cases in the following manner:

import org.mongodb.scala.bson.{BsonDocument, codecs}

query.toBsonDocument(BsonDocument.getClass, codecs.DEFAULT_CODEC_REGISTRY)

The following exception ends up being thrown:

Can't find a codec for class scala.Some.
org.bson.codecs.configuration.CodecConfigurationException: Can't find a codec for class scala.Some.

I am not sure how to rectify this, or if I need to implement my own codec for Options, and if I do I have no idea how I would do that.

Any help would be gladly appreciated. Thank you in advance

Update

I have seen that I can try and implement the Codec Interface as shown here:

http://mongodb.github.io/mongo-java-driver/3.0/bson/codecs/

Would I need to implement it for each possible sub-type of Option?

Example Option[Int], Option[UUID], Option[List[String]] etc.

回答1:

you could use something like this to fix the issue

class SomeCodec extends Codec[Some[_]] {
  override def encode(writer: BsonWriter, value: Some[_], encoderContext: EncoderContext): Unit = value match {
    case Some(v: String) ⇒ writer.writeString(v)
    case Some(v: Int) ⇒ writer.writeInt32(v)
    case Some(v: Long) ⇒ writer.writeInt64(v)
    case Some(v: Boolean) ⇒ writer.writeBoolean(v)
  }

  override def getEncoderClass: Class[Some[_]] = classOf[Some[_]]

  override def decode(reader: BsonReader, decoderContext: DecoderContext): Some[_] = {
    reader.getCurrentBsonType match {
      case BsonType.BOOLEAN ⇒ Some(reader.readBoolean())
      case BsonType.STRING ⇒ Some(reader.readString())
      case BsonType.INT64 ⇒ Some(reader.readInt64())
      case BsonType.INT32 ⇒ Some(reader.readInt32())
    }
  }
}