jsmap - simple json transformations Subscribe Pub Share

I just touched on a simple utility I had created, to play with json documents (and similar tree structures) via Map and List, see jsmap - simplified json.

This utility can also do very simple structure transformations jt like:

      import razie.js._

      jt(root) {
        // prune empty lists
        case (_, "children", x: List[_]) if (x.isEmpty) => ("" -> x)
      }
 

It removes the properties called "children" if they are an empty list, anywhere they appear (path is _) and it also stops the original transformation from recursing into that node.

      import razie.js._

      jt(tojmap) {
        case ("/", "xurl", u) => ("data" -> Map(
          "$dim" -> "10",
          "url" -> u))
        case (path, "links", l: List[_]) => ("adjacencies" -> jt(l, path) {
          case (_, "to", t) => ("nodeTo", t)
          case (_, "type", t) => ("data" -> Map("weight" -> "1"))
        })
      }

This will do several transformations, as you see some are path dependent (case ("/"... matches the root node only). Note that if you want to recurse in the original tree, you do it yourself by invoking the jt again.

If there was a property "type" of a list of objects "links", it will be transformed into an object "data" with weight etc.

You can see this very simple transformation can be tremendously flexible. I found it easy to use...

So for instance, to change the name of the "depth" property of every third level object to "goodDepth" it would be something like:

      import razie.js._

      tojsons(jt(parse(jsonString)) {
        case (path, "depth", value) if path.split("/").size == 3 => ("goodDepth" -> "good")
        }
      })

Why use other libraries that encode transformations into JSON or XML or any other language, when you have the full power of scala and instant testing, debugging, content assist etc?

---

See more in: Cool Scala Subscribe You need to log in to post a comment!

---