MoultingYAML is a Scala wrapper for SnakeYAML based on spray-json.
Its basic idea is to provide a simple immutable model of the YAML language, built on top of SnakeYAML models, as well as a type-class based serialization and deserialization of custom objects.
MoultingYAML's latest release is 0.4.0
and is built against Scala 2.12.0, 2.11.8 and 2.10.6.
To use it in an existing SBT project, add the following dependency to your
build.sbt
:
libraryDependencies += "net.jcazevedo" %% "moultingyaml" % "0.4.0"
In order to use MoultingYAML, bring all relevant elements into scope with:
import net.jcazevedo.moultingyaml._
import net.jcazevedo.moultingyaml.DefaultYamlProtocol._ // if you don't supply your own protocol
You can then parse a YAML string into its Abstract Syntax Tree (AST) representation with:
val source = """- Mark McGwire
|- Sammy Sosa
|- Ken Griffey""".stripMargin
val yamlAst = source.parseYaml
It is also possible to print a YAML AST back to a String using the prettyPrint
method:
val yaml = yamlAst.prettyPrint
If more fine-grained control over the printed yaml is needed, it is possible to
use the configurable print
method. For example, to enclose everything in
double quotes:
val yaml = yamlAst.print(scalarStyle = DoubleQuoted)
YAML provide different scalar styles to choose from, controlled by the argument
scalarStyle
of the print
method. The possible values for scalarStyle
are
Plain
, SingleQuoted
, DoubleQuoted
, Literal
and Folded
. Refer to the
YAML specification for details on
each representation.
In addition, YAML also has flow styles, in order to be able to use explicit
indicators instead of indentation to denote scope. The flow style is controlled
by the flowStyle
argument of the print
method. The possible values for
flowStyle
are Flow
, Block
and Auto
. Block
style uses indentation,
whereas Flow
style relies on explicit indicators to denote scope. The Auto
flow style attempts to combine both the Block
and Flow
style within the same
document.
Scala objects can be converted to a YAML AST using the pimped toYaml
method:
val yamlAst = List(1, 2, 3).toYaml
Convert a YAML AST to a Scala object with the convertTo
method:
val myList = yamlAst.convertTo[List[Int]]
In order to support calling the toYaml
and convertTo
methods for an object of
type T
, you need to have implicit values in scope that provide YamlFormat[T]
instances for T
and all types used by T
(directly or indirectly). You
normally do that through a YamlProtocol.
YamlProtocols follow the same design as spray-json's
JsonProtocols, which in turn are based on SJSON's. It's a type-class
based approach that connects an existing type T
with the logic of how to
(de)serialize its instances to and from YAML.
A YamlProtocol is a bunch of implicit values of type YamlFormat[T]
, where
each YamlFormat[T]
contains the logic of how to convert instances of T
to
and from YAML.
MoultingYAML comes with a DefaultYamlProtocol
, which already covers all of
Scala's value types as well as the most important reference and collection
types. The following are types already taken care of by the
DefaultYamlProtocol
:
Byte
,Short
,Int
,Long
,Float
,Double
,Char
,Unit
,Boolean
String
,Symbol
BigInt
,BigDecimal
Option
,Either
,Tuple1
-Tuple7
List
,Array
immutable.{Map, Iterable, Seq, IndexedSeq, LinearSeq, Set, Vector}
collection.{Iterable, Seq, IndexedSeq, LinearSeq, Set}
When you want to convert types not covered by the DefaultYamlProtocol
, you
need to provide a YamlFormat[T]
for your custom types.
If your custom type T
is a case class then augmenting the
DefaultYamlProtocol
with a YamlFormat[T]
can be done using the yamlFormatX
helpers, where X
stands for the number of fields in the case class:
case class Color(name: String, red: Int, green: Int, blue: Int)
object MyYamlProtocol extends DefaultYamlProtocol {
implicit val colorFormat = yamlFormat4(Color)
}
import MyYamlProtocol._
import net.jcazevedo.moultingyaml._
val yaml = Color("CadetBlue", 95, 158, 160).toYaml
val color = yaml.convertTo[Color]
Example combining custom YamlFormats:
case class Color(name: String, red: Int, green: Int, blue: Int)
case class Palette(name: String, colors: Option[List[Color]] = None)
object PaletteYamlProtocol extends DefaultYamlProtocol {
implicit val colorFormat = yamlFormat4(Color)
implicit val paletteFormat = yamlFormat2(Palette)
}
import PaletteYamlProtocol._
import net.jcazevedo.moultingyaml._
val yaml = """name: My Palette
|colors:
|- name: color 1
| red: 1
| green: 1
| blue: 1
|- name: color 2
| red: 2
| green: 2
| blue: 2
|""".stripMargin.parseYaml
val palette = yaml.convertTo[Palette]
If you explicitly declare the companion object for your case class the notation
above will stop working. You'll have to explicitly refer to the companion
object's apply
method to fix this:
case class Color(name: String, red: Int, green: Int, blue: Int)
object Color
object MyYamlProtocol extends DefaultYamlProtocol {
implicit val colorFormat = yamlFormat4(Color.apply)
}
If your case class has a type parameter the yamlFormat
methods can also help
you. However, there is a little more boilerplate required as you need to add
context bounds for all type parameters and explicitly refer to the case classes
apply method as in this example:
case class NamedList[A](name: String, items: List[A])
object MyYamlProtocol extends DefaultYamlProtocol {
implicit def namedListFormat[A: YamlFormat] = yamlFormat2(NamedList.apply[A])
}
As in spray-json, the NullOptions
trait supplies an alternative
rendering mode for optional case class members. Normally optional members that
are undefined (None
) are not rendered at all. By mixing in this trait into
your custom YamlProtocol you can enforce the rendering of undefined members as
null
. (Note that this only affects YAML writing, MoultingYAML will always read
missing optional members as well as null
optional members as None
)
To provide (de)serialization logic for types that aren't case classes, one has
to define the write
and read
methods of YamlFormat
. Here is one example:
class Color(val name: String, val red: Int, val green: Int, val blue: Int)
object MyYamlProtocol extends DefaultYamlProtocol {
implicit object ColorYamlFormat extends YamlFormat[Color] {
def write(c: Color) =
YamlArray(
YamlString(c.name),
YamlNumber(c.red),
YamlNumber(c.green),
YamlNumber(c.blue))
def read(value: YamlValue) = value match {
case YamlArray(
Vector(
YamlString(name),
YamlNumber(red: Int),
YamlNumber(green: Int),
YamlNumber(blue: Int))) =>
new Color(name, red, green, blue)
case _ => deserializationError("Color expected")
}
}
}
import MyYamlProtocol._
val yaml = new Color("CadetBlue", 95, 158, 160).toYaml
val color = yaml.convertTo[Color]
This serializes Color
instances as a YAML array. Another way would be to
serialize Color
s as YAML mappings, which are called YAML objects in
MoultingYAML:
object MyYamlProtocol extends DefaultYamlProtocol {
implicit object ColorYamlFormat extends YamlFormat[Color] {
def write(c: Color) = YamlObject(
YamlString("name") -> YamlString(c.name),
YamlString("red") -> YamlNumber(c.red),
YamlString("green") -> YamlNumber(c.green),
YamlString("blue") -> YamlNumber(c.blue)
)
def read(value: YamlValue) = {
value.asYamlObject.getFields(
YamlString("name"),
YamlString("red"),
YamlString("green"),
YamlString("blue")) match {
case Seq(
YamlString(name),
YamlNumber(red: Int),
YamlNumber(green: Int),
YamlNumber(blue: Int)) =>
new Color(name, red, green, blue)
case _ => deserializationError("Color expected")
}
}
}
}
Most of MoultingYAML's type-class (de)serialization code was inspired by spray-json, by Mathias Doenitz. spray-json was, in turn, inspired by the SJSON library by Debasish Ghosh. Both deserve credits here.
MoultingYAML is licensed under the MIT
license. See LICENSE.md
for details.
Feedback and contributions to the project are very welcome. Use GitHub's issue tracker to report any issues you might have when using the project. Submit code contributions via GitHub's pull requests.