What is an elegant way to parse for a specific part of a JSON string using a different key:value pair? (Scala)

Question

I need to find an Id based on it's associated name. My program makes REST calls to an API. The API returns results in JSON format. The Name is unique, so I would like to use it to get it's Id value. Note the ... can contain anything and does include some {Such}Id keys. The Id's can be nested in any number of {...{...{...}...}...} The Id's are always immediately before the name.

Note: ... is code that for privacy reasons cannot be shown. The code itself(before exclusion of private data) is the result of a REST call as returned by Advanced Rest Client and verified at http://jsonlint.com/ to be valid JSON.

The code is returned as such:

{
  Id: "d5a94d1a-afb7-4e1d-ae0d-a22e01393666"
  ProjectId: "ed61c45a-f208-4115-8584-a21a00c51ac0"
  Name: "Automated Runs"
  OrderNumber: 0
  Expands: [3]
    0:  "Children"
    1:  "Parent"
    2:  "Project"
    ...
    scripts: [4]
    0:  {
      Id: "0b70a55c-5e68-4b27-bfcf-a22f00c5dc48"
      Name: "3816"
      PackageId: "d5a94d1a-afb7-4e1d-ae0d-a22e01393666"
      ProjectId: "ed61c45a-f208-4115-8584-a21a00c51ac0"
      ...
      Expands: [6]
      0:  "Assignments"
      1:  "Attachments"
      2:  "FieldControls"
      3:  "FieldValues"
      4:  "Package"
      5:  "Steps"
      ...
    1:  {
      Id: "14e5c663-0d5a-46bb-ac48-a22f00c15998"
      Name: "3814"
      PackageId: "d5a94d1a-afb7-4e1d-ae0d-a22e01393666"
      ProjectId: "ed61c45a-f208-4115-8584-a21a00c51ac0"
      ...
      Expands: [6]
      0:  "Assignments"
      1:  "Attachments"
      2:  "FieldControls"
      3:  "FieldValues"
      4:  "Package"
      5:  "Steps"
      ...
    2:  {
      Id: "00d52fcd-b611-4f69-aeb6-a22f00c263a9"
      Name: "3815"
      ProjectId: "ed61c45a-f208-4115-8584-a21a00c51ac0"
      ...
      Expands: [6]
      0:  "Assignments"
      1:  "Attachments"
      2:  "FieldControls"
      3:  "FieldValues"
      4:  "Package"
      5:  "Steps"
      ...

    3:  {
      Id: "4d3a6132-8497-4b6b-a064-a22f00c669ff"
      Name: "3817"
      ...
      Expands: [6]
      0:  "Assignments"
      1:  "Attachments"
      2:  "FieldControls"
      3:  "FieldValues"
      4:  "Package"
      5:  "Steps"
      ...
}

Things I have tried include regex (I'm new to it and having some troubles), and simple string splitting. While I have the string splitting working, it is semi hard coded.

What i would like is something like:

def getID(myJSON:String, myName:String){
  val pattern = "\"Id\": \"*\",\r\n\"Name\":\"" + myName + "\",\""
  get the id (*) from result using pattern
}

Or Even better convert it to be generic.

def getID(myJSON:String, myValue:String, searchKey:String, findKey:String){
  val pattern = { ... findKey: *...} in the inner most  { ... searchKey: * ...} scope
  get the id (*) from result using the pattern in the found {...searchKey...} scope
}

Either would be great and very much appreciated. My current code looks like:

result.split("Id\": \"")(3).split("\"")(0)

It might be pretty, but it has lots of room for mishaps. An Id might be created by a user which sets the count to be incorrect, etc...

Thank you, Erick Stone

Answer 1

How about using the for comprehension in json4s

scala> :paste
// Entering paste mode (ctrl-D to finish)

  import org.json4s._
  import org.json4s.native.JsonMethods._

  val json = """
      {
        "a": {
          "Id": "1",
          "Name": "Name1",
          "b": {
            "Id": "2",
            "Name": "Name2",
          }
        }
      }
    """


  def getId(json: String, name: String) = {
    val res = for {
      JObject(child) <- parse(json)
      JField("Name", JString(n)) <- child
      JField("Id", JString(id)) <- child
      if n == name
    } yield id

    res.headOption
  }

// Exiting paste mode, now interpreting.

scala> getId(json, "Name1")
res4: Option[String] = Some(1)

scala> getId(json, "Name2")
res5: Option[String] = Some(2)