Pig Python Stream Example

register 'hdfs:///user/jniemie8/pigpythonudf.py' using org.apache.pig.scripting.streaming.python.PythonScriptEngine as pythonudfs;
rawdata = LOAD 'pigpythonudfexample/*' using TextLoader();
sortedrecords = FOREACH rawdata GENERATE FLATTEN(pythonudfs.reorderColumns($0));
store sortedrecords into 'pigpythonudfexample/';


/*##datafile## <- remove this line
1, 7, 5, 4, 3
10, 2, 5, 4, 7
1, 9, 4, 4, 3
*/

# pigpythonudf.py
from pig_util import outputSchema
@outputSchema("t1:tuple(col0:int, cal1:int, col2:int, col3:int, col4:int)")
def reorderColumns(line):
                columns=sorted([int(x.strip()) for x in line.split(',')])
                return tuple(columns)