A script to dump bytecode in plaintext

README.md

This is a simple script to help me see the bytecode when plaing with Scala 3.

All it does is parallelises the calls to javap command that should available if you have JDK installed.

Installation

Just Ammonite - https://ammonite.io/

It's excellent.

usage

For example, running it in watch mode, pointing it at the target folder with classes:

amm -w dump-bytecode.sc --classesFolder $(pwd)/scala3/target/scala-3.0.0-M3/classes --bytecodeFolder $(pwd)/bytecode-scala

WARNING: the script will delete all *.bytecode files (and only them) from the target folder

dump-bytecode.sc

import $ivy.`org.scala-lang.modules::scala-parallel-collections:1.0.0`

import scala.collection.parallel.CollectionConverters._

import ammonite.ops._
import os.ProcessOutput

def process(classFile: os.Path, bytecodeFile: os.Path): Unit = {
  println(s"Processing ${classFile}")
  val args =
    Seq(
      "javap",
      "-v",
      "-c",
      classFile.toNIO.toAbsolutePath().toString()
    )

  os.proc(args)
    .call(
      stdout = ProcessOutput.makePathRedirect(bytecodeFile),
      stderr = os.Inherit
    )
}

@main
def main(classesFolder: os.Path, bytecodeFolder: os.Path) = {

  interp.watch(classesFolder)

  os.makeDir.all(bytecodeFolder)
  
  os.walk(bytecodeFolder).filter(_.ext == "bytecode").foreach { p =>
   os.remove(p)
  }

  val allClassfiles = os.walk(classesFolder).filter(_.ext == "class")


  allClassfiles.zip(allClassfiles.map(_.baseName + ".bytecode")).par.foreach {
    case (absPath, filename) =>
      val target = bytecodeFolder / filename

      process(absPath, target)
  }
}