Step towards replacing bzip2/xz compression with Java implementation

com.oracle.truffle.r.runtime/src/com/oracle/truffle/r/runtime/RCompression.java

@@ -22,19 +22,22 @@
  */
 package com.oracle.truffle.r.runtime;
 
+import java.io.ByteArrayInputStream;
 import java.io.FileInputStream;
 import java.io.IOException;
 import java.io.InputStream;
 import java.io.OutputStream;
 import java.lang.ProcessBuilder.Redirect;
+import java.util.Arrays;
 import java.util.zip.GZIPInputStream;
 
 import com.oracle.truffle.r.runtime.conn.GZIPConnections.GZIPRConnection;
 import com.oracle.truffle.r.runtime.ffi.RFFIFactory;
 
+import org.tukaani.xz.LZMA2InputStream;
+
 /**
- * Abstracts the implementation of the various forms of compression used in R. Since the C API for
- * LZMA is very complex (as used by GnuR), we use an 'xz' subprocess to do the work.
+ * Abstracts the implementation of the various forms of compression used in R.
  */
 public class RCompression {
     public enum Type {
@@ -98,7 +101,8 @@ public class RCompression {
             case BZIP2:
                 throw RInternalError.unimplemented("BZIP2 compression");
             case LZMA:
-                return lzmaUncompress(udata, cdata);
+                byte[] udataCopy = Arrays.copyOf(udata, udata.length);
+                return lzmaUncompressInternal(udataCopy, cdata);
             default:
                 assert false;
                 return false;
@@ -160,6 +164,20 @@ public class RCompression {
 
     }
 
+    private static boolean lzmaUncompressInternal(byte[] udata, byte[] data) {
+        int dictSize = udata.length < LZMA2InputStream.DICT_SIZE_MIN ? LZMA2InputStream.DICT_SIZE_MIN : udata.length;
+        try (LZMA2InputStream lzmaStream = new LZMA2InputStream(new ByteArrayInputStream(data), dictSize)) {
+            int totalRead = 0;
+            int n;
+            while ((n = lzmaStream.read(udata, totalRead, udata.length - totalRead)) > 0) {
+                totalRead += n;
+            }
+            return totalRead == udata.length;
+        } catch (IOException ex) {
+            return false;
+        }
+    }
+
     private static boolean lzmaUncompress(byte[] udata, byte[] data) {
         int rc;
         ProcessBuilder pb = new ProcessBuilder("xz", "--decompress", "--format=raw", "--lzma2", "--stdout");

mx.fastr/suite.py

@@ -73,6 +73,16 @@ suite = {
         "resource" : "true"
     },
 
+    "APACHE-COMMONS-COMPRESS" : {
+        "urls" : ["http://central.maven.org/maven2/org/apache/commons/commons-compress/1.12/commons-compress-1.12.jar"],
+        "sha1" : "84caa68576e345eb5e7ae61a0e5a9229eb100d7b"
+    },
+
+    "TUKAANI_XZ" : {
+        "urls" : ["http://central.maven.org/maven2/org/tukaani/xz/1.5/xz-1.5.jar"],
+        "sha1" : "9c64274b7dbb65288237216e3fae7877fd3f2bee",
+    },
+
     "ANTLR-3.5" : {
       "path" : "libdownloads/antlr-runtime-3.5.jar",
       "urls" : ["http://central.maven.org/maven2/org/antlr/antlr-runtime/3.5/antlr-runtime-3.5.jar"],
@@ -315,6 +325,8 @@ suite = {
       "dependencies" : [
         "truffle:TRUFFLE_API",
         "truffle:TRUFFLE_DEBUG",
+        "APACHE-COMMONS-COMPRESS",
+        "TUKAANI_XZ",
       ],
       "checkstyle" : "com.oracle.truffle.r.runtime",
       "javaCompliance" : "1.8",