Merge pull request sbt#520 from shuttie/fix-dedup-heap-usage

do not buffer entry contents in memory while doing MergeStrategy.deduplicate

Author: eed3si9n

src/main/scala/sbtassembly/Assembly.scala

@@ -699,21 +699,22 @@ object Assembly {
 
   private[sbtassembly] def sha1 = MessageDigest.getInstance("SHA-1")
 
-  private[sbtassembly] def sha1Content(b: Array[Byte]): String =
-    byteArrayInputStreamResource(b) { in =>
+  private[sbtassembly] def sha1Content(stream: InputStream): String =
+    {
       val messageDigest = sha1
       val buffer = new Array[Byte](8192)
 
       @tailrec
       def read(): Unit = {
-        val byteCount = in.read(buffer)
+        val byteCount = stream.read(buffer)
         if (byteCount >= 0) {
           messageDigest.update(buffer, 0, byteCount)
           read()
         }
       }
 
       read()
+      stream.close()
       bytesToString(messageDigest.digest())
     }
 

src/main/scala/sbtassembly/MergeStrategy.scala

@@ -122,10 +122,11 @@ object MergeStrategy {
    * Verifies if all the conflicts have the same content, otherwise error out
    */
   val deduplicate: MergeStrategy = MergeStrategy("Deduplicate") { conflicts =>
-    val conflictContents = conflicts.map(_.stream()).map(Streamable.bytes(_))
-    val fingerprints = Set() ++ conflictContents.map(sbtassembly.Assembly.sha1Content)
+    val fingerprints = conflicts.map(dep =>
+      sbtassembly.Assembly.sha1Content(dep.stream())
+    ).toSet
     if (fingerprints.size == 1)
-      Right(Vector(JarEntry(conflicts.head.target, () => new ByteArrayInputStream(conflictContents.head))))
+      Right(Vector(JarEntry(conflicts.head.target, conflicts.head.stream)))
     else
       Left(
         s"Deduplicate found different file contents in the following:$newLineIndented${conflicts.mkString(newLineIndented)}"