merge master

Author: witgo

.gitignore

@@ -7,7 +7,7 @@
 sbt/*.jar
 .settings
 .cache
-.mima-excludes
+.generated-mima-excludes
 /build/
 work/
 out/

.rat-excludes

@@ -3,6 +3,7 @@ target
 .project
 .classpath
 .mima-excludes
+.generated-mima-excludes
 .rat-excludes
 .*md
 derby.log

README.md

@@ -9,13 +9,14 @@ You can find the latest Spark documentation, including a programming
 guide, on the project webpage at <http://spark.apache.org/documentation.html>.
 This README file only contains basic setup instructions.
 
-
 ## Building Spark
 
 Spark is built on Scala 2.10. To build Spark and its example programs, run:
 
     ./sbt/sbt assembly
 
+(You do not need to do this if you downloaded a pre-built package.)
+
 ## Interactive Scala Shell
 
 The easiest way to start using Spark is through the Scala shell:
@@ -41,9 +42,9 @@ And run the following command, which should also return 1000:
 Spark also comes with several sample programs in the `examples` directory.
 To run one of them, use `./bin/run-example <class> [params]`. For example:
 
-    ./bin/run-example org.apache.spark.examples.SparkLR
+    ./bin/run-example SparkPi
 
-will run the Logistic Regression example locally.
+will run the Pi example locally.
 
 You can set the MASTER environment variable when running examples to submit
 examples to a cluster. This can be a mesos:// or spark:// URL, 

assembly/pom.xml

@@ -96,7 +96,7 @@
             <filter>
               <artifact>*:*</artifact>
               <excludes>
-                <exclude>org.datanucleus:*</exclude>
+                <exclude>org/datanucleus/**</exclude>
                 <exclude>META-INF/*.SF</exclude>
                 <exclude>META-INF/*.DSA</exclude>
                 <exclude>META-INF/*.RSA</exclude>

bagel/src/test/scala/org/apache/spark/bagel/BagelSuite.scala

@@ -38,8 +38,6 @@ class BagelSuite extends FunSuite with Assertions with BeforeAndAfter with Timeo
       sc.stop()
       sc = null
     }
-    // To avoid Akka rebinding to the same port, since it doesn't unbind immediately on shutdown
-    System.clearProperty("spark.driver.port")
   }
 
   test("halting by voting") {

bin/compute-classpath.cmd

@@ -20,6 +20,13 @@ rem
 rem This script computes Spark's classpath and prints it to stdout; it's used by both the "run"
 rem script and the ExecutorRunner in standalone cluster mode.
 
+rem If we're called from spark-class2.cmd, it already set enabledelayedexpansion and setting
+rem it here would stop us from affecting its copy of the CLASSPATH variable; otherwise we
+rem need to set it here because we use !datanucleus_jars! below.
+if "%DONT_PRINT_CLASSPATH%"=="1" goto skip_delayed_expansion
+setlocal enabledelayedexpansion
+:skip_delayed_expansion
+
 set SCALA_VERSION=2.10
 
 rem Figure out where the Spark framework is installed
@@ -31,7 +38,7 @@ if exist "%FWDIR%conf\spark-env.cmd" call "%FWDIR%conf\spark-env.cmd"
 rem Build up classpath
 set CLASSPATH=%FWDIR%conf
 if exist "%FWDIR%RELEASE" (
-  for %%d in ("%FWDIR%jars\spark-assembly*.jar") do (
+  for %%d in ("%FWDIR%lib\spark-assembly*.jar") do (
     set ASSEMBLY_JAR=%%d
   )
 ) else (
@@ -42,6 +49,21 @@ if exist "%FWDIR%RELEASE" (
 
 set CLASSPATH=%CLASSPATH%;%ASSEMBLY_JAR%
 
+rem When Hive support is needed, Datanucleus jars must be included on the classpath.
+rem Datanucleus jars do not work if only included in the uber jar as plugin.xml metadata is lost.
+rem Both sbt and maven will populate "lib_managed/jars/" with the datanucleus jars when Spark is
+rem built with Hive, so look for them there.
+if exist "%FWDIR%RELEASE" (
+  set datanucleus_dir=%FWDIR%lib
+) else (
+  set datanucleus_dir=%FWDIR%lib_managed\jars
+)
+set "datanucleus_jars="
+for %%d in ("%datanucleus_dir%\datanucleus-*.jar") do (
+  set datanucleus_jars=!datanucleus_jars!;%%d
+)
+set CLASSPATH=%CLASSPATH%;%datanucleus_jars%
+
 set SPARK_CLASSES=%FWDIR%core\target\scala-%SCALA_VERSION%\classes
 set SPARK_CLASSES=%SPARK_CLASSES%;%FWDIR%repl\target\scala-%SCALA_VERSION%\classes
 set SPARK_CLASSES=%SPARK_CLASSES%;%FWDIR%mllib\target\scala-%SCALA_VERSION%\classes

bin/pyspark

@@ -17,14 +17,20 @@
 # limitations under the License.
 #
 
-# Figure out where the Scala framework is installed
+# Figure out where Spark is installed
 FWDIR="$(cd `dirname $0`/..; pwd)"
 
 # Export this as SPARK_HOME
 export SPARK_HOME="$FWDIR"
 
 SCALA_VERSION=2.10
 
+if [[ "$@" = *--help ]] || [[ "$@" = *-h ]]; then
+  echo "Usage: ./bin/pyspark [options]"
+  $FWDIR/bin/spark-submit --help 2>&1 | grep -v Usage 1>&2
+  exit 0
+fi
+
 # Exit if the user hasn't compiled Spark
 if [ ! -f "$FWDIR/RELEASE" ]; then
   # Exit if the user hasn't compiled Spark
@@ -52,13 +58,34 @@ export PYTHONPATH=$SPARK_HOME/python/lib/py4j-0.8.1-src.zip:$PYTHONPATH
 export OLD_PYTHONSTARTUP=$PYTHONSTARTUP
 export PYTHONSTARTUP=$FWDIR/python/pyspark/shell.py
 
+# If IPython options are specified, assume user wants to run IPython
 if [ -n "$IPYTHON_OPTS" ]; then
   IPYTHON=1
 fi
 
-# Only use ipython if no command line arguments were provided [SPARK-1134]
-if [[ "$IPYTHON" = "1" && $# = 0 ]] ; then
-  exec ipython $IPYTHON_OPTS
+# Build up arguments list manually to preserve quotes and backslashes.
+# We export Spark submit arguments as an environment variable because shell.py must run as a
+# PYTHONSTARTUP script, which does not take in arguments. This is required for IPython notebooks.
+
+PYSPARK_SUBMIT_ARGS=""
+whitespace="[[:space:]]"
+for i in "$@"; do
+  if [[ $i =~ \" ]]; then i=$(echo $i | sed 's/\"/\\\"/g'); fi
+  if [[ $i =~ $whitespace ]]; then i=\"$i\"; fi
+  PYSPARK_SUBMIT_ARGS="$PYSPARK_SUBMIT_ARGS $i"
+done
+export PYSPARK_SUBMIT_ARGS
+
+# If a python file is provided, directly run spark-submit.
+if [[ "$1" =~ \.py$ ]]; then
+  echo -e "\nWARNING: Running python applications through ./bin/pyspark is deprecated as of Spark 1.0." 1>&2
+  echo -e "Use ./bin/spark-submit <python file>\n" 1>&2
+  exec $FWDIR/bin/spark-submit "$@"
 else
-  exec "$PYSPARK_PYTHON" "$@"
+  # Only use ipython if no command line arguments were provided [SPARK-1134]
+  if [[ "$IPYTHON" = "1" ]]; then
+    exec ipython $IPYTHON_OPTS
+  else
+    exec "$PYSPARK_PYTHON"
+  fi
 fi

bin/pyspark2.cmd

@@ -31,7 +31,7 @@ set FOUND_JAR=0
 for %%d in ("%FWDIR%assembly\target\scala-%SCALA_VERSION%\spark-assembly*hadoop*.jar") do (
   set FOUND_JAR=1
 )
-if "%FOUND_JAR%"=="0" (
+if [%FOUND_JAR%] == [0] (
   echo Failed to find Spark assembly JAR.
   echo You need to build Spark with sbt\sbt assembly before running this program.
   goto exit
@@ -42,15 +42,30 @@ rem Load environment variables from conf\spark-env.cmd, if it exists
 if exist "%FWDIR%conf\spark-env.cmd" call "%FWDIR%conf\spark-env.cmd"
 
 rem Figure out which Python to use.
-if "x%PYSPARK_PYTHON%"=="x" set PYSPARK_PYTHON=python
+if [%PYSPARK_PYTHON%] == [] set PYSPARK_PYTHON=python
 
 set PYTHONPATH=%FWDIR%python;%PYTHONPATH%
 set PYTHONPATH=%FWDIR%python\lib\py4j-0.8.1-src.zip;%PYTHONPATH%
 
 set OLD_PYTHONSTARTUP=%PYTHONSTARTUP%
 set PYTHONSTARTUP=%FWDIR%python\pyspark\shell.py
+set PYSPARK_SUBMIT_ARGS=%*
 
 echo Running %PYSPARK_PYTHON% with PYTHONPATH=%PYTHONPATH%
 
-"%PYSPARK_PYTHON%" %*
+rem Check whether the argument is a file
+for /f %%i in ('echo %1^| findstr /R "\.py"') do (
+  set PYTHON_FILE=%%i
+)
+
+if [%PYTHON_FILE%] == [] (
+  %PYSPARK_PYTHON%
+) else (
+  echo.
+  echo WARNING: Running python applications through ./bin/pyspark.cmd is deprecated as of Spark 1.0.
+  echo Use ./bin/spark-submit ^<python file^>
+  echo.
+  "%FWDIR%\bin\spark-submit.cmd" %PYSPARK_SUBMIT_ARGS%
+)
+
 :exit

bin/run-example

@@ -23,6 +23,16 @@ FWDIR="$(cd `dirname $0`/..; pwd)"
 export SPARK_HOME="$FWDIR"
 EXAMPLES_DIR="$FWDIR"/examples
 
+if [ -n "$1" ]; then
+  EXAMPLE_CLASS="$1"
+  shift
+else
+  echo "Usage: ./bin/run-example <example-class> [example-args]"
+  echo "  - set MASTER=XX to use a specific master"
+  echo "  - can use abbreviated example class name (e.g. SparkPi, mllib.LinearRegression)"
+  exit 1
+fi
+
 if [ -f "$FWDIR/RELEASE" ]; then
   export SPARK_EXAMPLES_JAR=`ls "$FWDIR"/lib/spark-examples-*hadoop*.jar`
 elif [ -e "$EXAMPLES_DIR"/target/scala-$SCALA_VERSION/spark-examples-*hadoop*.jar ]; then
@@ -37,23 +47,12 @@ fi
 
 EXAMPLE_MASTER=${MASTER:-"local[*]"}
 
-if [ -n "$1" ]; then
-  EXAMPLE_CLASS="$1"
-  shift
-else 
-  echo "usage: ./bin/run-example <example-class> [example-args]" 
-  echo "  - set MASTER=XX to use a specific master"
-  echo "  - can use abbreviated example class name (e.g. SparkPi, mllib.MovieLensALS)"
-  echo
-  exit -1
-fi
-
 if [[ ! $EXAMPLE_CLASS == org.apache.spark.examples* ]]; then
   EXAMPLE_CLASS="org.apache.spark.examples.$EXAMPLE_CLASS"
 fi
 
 ./bin/spark-submit \
   --master $EXAMPLE_MASTER \
   --class $EXAMPLE_CLASS \
-  $SPARK_EXAMPLES_JAR \
+  "$SPARK_EXAMPLES_JAR" \
   "$@"

bin/run-example2.cmd

@@ -30,32 +30,59 @@ if exist "%FWDIR%conf\spark-env.cmd" call "%FWDIR%conf\spark-env.cmd"
 
 rem Test that an argument was given
 if not "x%1"=="x" goto arg_given
-  echo Usage: run-example ^<example-class^> [^<args^>]
+  echo Usage: run-example ^<example-class^> [example-args]
+  echo   - set MASTER=XX to use a specific master
+  echo   - can use abbreviated example class name (e.g. SparkPi, mllib.LinearRegression)
   goto exit
 :arg_given
 
 set EXAMPLES_DIR=%FWDIR%examples
 
 rem Figure out the JAR file that our examples were packaged into.
 set SPARK_EXAMPLES_JAR=
-for %%d in ("%EXAMPLES_DIR%\target\scala-%SCALA_VERSION%\spark-examples*assembly*.jar") do (
-  set SPARK_EXAMPLES_JAR=%%d
+if exist "%FWDIR%RELEASE" (
+  for %%d in ("%FWDIR%lib\spark-examples*.jar") do (
+    set SPARK_EXAMPLES_JAR=%%d
+  )
+) else (
+  for %%d in ("%EXAMPLES_DIR%\target\scala-%SCALA_VERSION%\spark-examples*.jar") do (
+    set SPARK_EXAMPLES_JAR=%%d
+  )
 )
 if "x%SPARK_EXAMPLES_JAR%"=="x" (
   echo Failed to find Spark examples assembly JAR.
   echo You need to build Spark with sbt\sbt assembly before running this program.
   goto exit
 )
 
-rem Compute Spark classpath using external script
-set DONT_PRINT_CLASSPATH=1
-call "%FWDIR%bin\compute-classpath.cmd"
-set DONT_PRINT_CLASSPATH=0
-set CLASSPATH=%SPARK_EXAMPLES_JAR%;%CLASSPATH%
+rem Set master from MASTER environment variable if given
+if "x%MASTER%"=="x" (
+  set EXAMPLE_MASTER=local[*]
+) else (
+  set EXAMPLE_MASTER=%MASTER%
+)
+
+rem If the EXAMPLE_CLASS does not start with org.apache.spark.examples, add that
+set EXAMPLE_CLASS=%1
+set PREFIX=%EXAMPLE_CLASS:~0,25%
+if not %PREFIX%==org.apache.spark.examples (
+  set EXAMPLE_CLASS=org.apache.spark.examples.%EXAMPLE_CLASS%
+)
+
+rem Get the tail of the argument list, to skip the first one. This is surprisingly
+rem complicated on Windows.
+set "ARGS="
+:top
+shift
+if "%~1" neq "" (
+  set ARGS=%ARGS% "%~1"
+  goto :top
+)
+if defined ARGS set ARGS=%ARGS:~1%
 
-rem Figure out where java is.
-set RUNNER=java
-if not "x%JAVA_HOME%"=="x" set RUNNER=%JAVA_HOME%\bin\java
+call "%FWDIR%bin\spark-submit.cmd" ^
+  --master %EXAMPLE_MASTER% ^
+  --class %EXAMPLE_CLASS% ^
+  "%SPARK_EXAMPLES_JAR%" %ARGS%
 
-"%RUNNER%" -cp "%CLASSPATH%" %JAVA_OPTS% %*
 :exit