Scala hex literal for bytes

问题

Hex literal containing A-F digit are converting to int by default. When I am trying to declear an Int with 0x it is creating correctly.

val a: Int = 0x34
val b: Int = 0xFF

But when I am trying to declear a Byte with 0x second line is not compiling

val a: Byte = 0x34
val b: Byte = 0xFF // compilation error

I have found a workaround that is

val a: Byte = 0x34
val b: Byte = 0xFF.toByte

But is there any decent way to declear a Byte from its hex literal?

For example I am trying to declear a Byte array in a Test method in this way

anObject.someMethod(1, 1.1f, 0xAB, "1") shouldBe
  Array[Byte](0xAF, 0xAF, 0xAF, 0xAF, 0xAF, 0xAF, 0xAF, 0xAF, 0xAF, 0xAF, 0xAF, 0xAF, 0xAF, 0xAF, 0xAF, 0xAF, 0xAF, 0xAF, 0xAF, 0xAF, 0xAF, 0xAF, 0xAF, 0xAF)
anObject.someMethod(2, 2.2f, 0xCD, "2") shouldBe
  Array[Byte](0xBE, 0xBE, 0xBE, 0xBE, 0xBE, 0xBE, 0xBE, 0xBE, 0xBE, 0xBE, 0xBE, 0xBE, 0xBE, 0xBE, 0xBE, 0xBE, 0xBE, 0xBE, 0xBE, 0xBE, 0xBE, 0xBE, 0xBE, 0xBE)
anObject.someMethod(3, 3.2f, 0xEF, "3") shouldBe
  Array[Byte](0xCD, 0xCD, 0xCD, 0xCD, 0xCD, 0xCD, 0xCD, 0xCD, 0xCD, 0xCD, 0xCD, 0xCD, 0xCD, 0xCD, 0xCD, 0xCD, 0xCD, 0xCD, 0xCD, 0xCD, 0xCD, 0xCD, 0xCD, 0xCD)

But not in this way

anObject.someMethod(1, 1.1f, 0xAB.toByte, "1") shouldBe
  Array[Byte](0xAF.toByte, 0xAF.toByte, 0xAF.toByte, 0xAF.toByte, 0xAF.toByte, 0xAF.toByte, 0xAF.toByte, 0xAF.toByte, 0xAF.toByte, 0xAF.toByte, 0xAF.toByte, 0xAF.toByte, 0xAF.toByte, 0xAF.toByte, 0xAF.toByte, 0xAF.toByte, 0xAF.toByte, 0xAF.toByte, 0xAF.toByte, 0xAF.toByte, 0xAF.toByte, 0xAF.toByte, 0xAF.toByte, 0xAF.toByte)
anObject.someMethod(2, 2.2f, 0xCD.toByte, "2") shouldBe
  Array[Byte](0xBE.toByte, 0xBE.toByte, 0xBE.toByte, 0xBE.toByte, 0xBE.toByte, 0xBE.toByte, 0xBE.toByte, 0xBE.toByte, 0xBE.toByte, 0xBE.toByte, 0xBE.toByte, 0xBE.toByte, 0xBE.toByte, 0xBE.toByte, 0xBE.toByte, 0xBE.toByte, 0xBE.toByte, 0xBE.toByte, 0xBE.toByte, 0xBE.toByte, 0xBE.toByte, 0xBE.toByte, 0xBE.toByte, 0xBE.toByte)
anObject.someMethod(3, 3.2f, 0xEF.toByte, "3") shouldBe
  Array[Byte](0xCD.toByte, 0xCD.toByte, 0xCD.toByte, 0xCD.toByte, 0xCD.toByte, 0xCD.toByte, 0xCD.toByte, 0xCD.toByte, 0xCD.toByte, 0xCD.toByte, 0xCD.toByte, 0xCD.toByte, 0xCD.toByte, 0xCD.toByte, 0xCD.toByte, 0xCD.toByte, 0xCD.toByte, 0xCD.toByte, 0xCD.toByte, 0xCD.toByte, 0xCD.toByte, 0xCD.toByte, 0xCD.toByte, 0xCD.toByte)

Tested in scala 2.12.4

回答1:

You can do this with implicit conversions.

Before:

def f(b: Byte) = println(s"Byte = $b")
f(0x34)
f(0xFF) // compilation error

After:

implicit def int2Byte(i: Int) = i.toByte

def f(b: Byte) = println(s"Byte = $b")
f(0x34)
f(0xFF)

Output:

Byte = 52
Byte = -1

回答2:

Recall that in Scala we can easily define new ways to interpret arbitrary String literals by adding methods to a special class StringContext using the "pimp-my-library"-pattern. For example, we can add the method b for creating single bytes to StringContext so that we can write down bytes as follows:

val myByte = b"5C"

Here is how it can be implemented:

implicit class SingleByteContext(private val sc: StringContext) {
   def b(args: Any*): Byte = {
     val parts = sc.parts.toList
     assert(
       parts.size == 1 && args.size == 0, 
       "Expected a string literal with exactly one part"
     )
     Integer.parseInt(parts(0), 16).toByte
   }
}

In order to use this new syntax, we have to import the above object into implicit scope. Then we can do this:

/* Examples */ {

  def printByte(b: Byte) = println("%02X".format(b))

  printByte(b"01")
  printByte(b"7F")
  printByte(b"FF")
  printByte(b"80")
}

This will print:

01
7F
FF
80

You can obviously tweak the implementation (e.g. you can rename "b" to "hex" or "x" or "Ox" or something like this).

---

Note that this technique can be easily extended to deal with entire byte arrays, as described in this answer to a similar question. This would allow you to write down byte arrays without repeating the annoying 0x-prefix, e.g.:

val myBytes = hexBytes"AB,CD,5E,7F,5A,8C,80,BC"

来源：https://stackoverflow.com/questions/48926339/scala-hex-literal-for-bytes