Gzip Scala

Gzip.scala

import java.io.{ByteArrayOutputStream, ByteArrayInputStream}
import java.util.zip.{GZIPOutputStream, GZIPInputStream}

import scala.util.Try

object Gzip {

  def compress(input: Array[Byte]): Array[Byte] = {
    val bos = new ByteArrayOutputStream(input.length)
    val gzip = new GZIPOutputStream(bos)
    gzip.write(input)
    gzip.close()
    val compressed = bos.toByteArray
    bos.close()
    compressed
  }

  def decompress(compressed: Array[Byte]): Option[String] =
    Try {
      val inputStream = new GZIPInputStream(new ByteArrayInputStream(compressed))
      scala.io.Source.fromInputStream(inputStream).mkString
    }.toOption
}

GzipSpec.scala

class GzipSpec extends WordSpecLike with Matchers {

  "The GZIP object" should {

    "decompress a compressed string" in {
      val input = Gzip.compress("Hello World".getBytes("UTF-8"))
      Gzip.decompress(input) shouldBe Some("Hello World")
    }
  }
}

Should you not do Base64Encoding here ?

@sbhola only if you need it in an ASCII string for sending via a transport like SMTP or embedding within some other file format. The raw binary data will be smaller than the Base64 representation.

Is there any dependency we need include to run this code? I am getting following error message"Cannot resolve symbol wordspeclike"?

Thanks @owainlewis!
This is very nice and useful.
Just wondering - why does decompress return a String and not an Array[Byte]?
I think having Array[Byte] will make this useful for a variety of binary and non binary items.

I have same question as JThakrar.

Here's my version of decompress.

  def decompress(compressed: Array[Byte]): Array[Byte] = {
    val gzipInputStream = new GZIPInputStream(new ByteArrayInputStream(compressed))
    val output = new ArrayBuffer[Byte]()
    var totalByteCount = 0
    val bytes = new Array[Byte](1024)
    while (gzipInputStream.available() == 1) {
      val byteCount = gzipInputStream.read(bytes)
      if (byteCount > 0) {
        output ++= bytes.take(byteCount)
        totalByteCount += byteCount
      }
    }
    output.take(totalByteCount).toArray
  }

similar to @JThakrar's...probably simpler but requires Apache Commons IO

def decompress(compressed: Array[Byte]): Option[Array[Byte]] =
    Try {
      val inputStream = new GZIPInputStream(new ByteArrayInputStream(compressed))
      org.apache.commons.io.IOUtils.toByteArray(inputStream)
    }.toOption

Thanks. Extra example of sequential generating of gziped CSV files with os-lib:

Add the following to build.sbt

libraryDependencies += "com.lihaoyi" %% "os-lib" % "0.8.0"

then

package example

import scala.util.{Try, Success, Failure, Random}

import java.io.{ByteArrayOutputStream, ByteArrayInputStream}
import java.util.zip.{GZIPOutputStream, GZIPInputStream}

object Main extends  App {

  // create header for a csv file
  val header = List("column1","column2", "column3")
  val header_compressed = Gzip.compress(
        (header.mkString(",")+"\n").getBytes("UTF-8"))
  os.write.over(os.Path("/tmp/file.gz"), header_compressed)  

  (0 to 10).foreach { idx =>

    val rand = Random

    val lines_to_write = (0 to 100).map { _ =>
      (0 to 2).map(_ => rand.nextInt(100).toString).mkString(",")     
    }

    val new_lines = Gzip.compress(
      (lines_to_write.mkString("\n")+"\n").getBytes("UTF-8"))

    os.write.append(os.Path("/tmp/file.gz"), new_lines) 
  }
}

  
object Gzip {

  def compress(input: Array[Byte]): Array[Byte] = {
    val bos = new ByteArrayOutputStream(input.length)
    val gzip = new GZIPOutputStream(bos)
    gzip.write(input)
    gzip.close()
    val compressed = bos.toByteArray
    bos.close()
    compressed
  }

  def decompress(compressed: Array[Byte]): Option[String] =
    Try {
      val inputStream = new GZIPInputStream(new ByteArrayInputStream(compressed))
      scala.io.Source.fromInputStream(inputStream).mkString
    }.toOption
}