def createNewAtomicPomset(self, name=None,
executableObject=None,
commandBuilderType=None,
executeEnvironmentType=None,
*args, **kwds):
newAtomicPomset = DefinitionModule.AtomicDefinition(*args, **kwds)
if name is None:
name = 'pomset %s' % uuid.uuid4().hex[:3]
newAtomicPomset.name(name)
newAtomicPomset.functionToExecute(
DefinitionModule.executeTaskInEnvironment)
newAtomicPomset.executable(executableObject)
# create the parameter orderings
parameterOrderings = DefinitionModule.createParameterOrderingTable()
newAtomicPomset.parameterOrderingTable(parameterOrderings)
if commandBuilderType is None:
commandBuilderType = 'shell process'
newAtomicPomset.commandBuilderType(commandBuilderType)
if executeEnvironmentType is None:
executeEnvironmentType = 'shell process'
newAtomicPomset.executeEnvironmentType(executeEnvironmentType)
newPomsetContext = ContextModule.wrapPomsetInContext(newAtomicPomset)
return newPomsetContext
def createHadoopWordcountDefinition():
parameterOrdering = DefinitionModule.createParameterOrderingTable()
row = parameterOrdering.addRow()
row.setColumn('source', 'input file')
row.setColumn('target', 'output file')
# TODO:
# need to be able to customize this for each host
executable = HadoopModule.JarExecutable()
executable.stageable(False)
executable.path([HadoopModule.getExecutablePath()])
executable.jarFile([getExamplesJar()])
executable.jarClass(['wordcount'])
definition = DefinitionModule.createShellProcessDefinition(
inputParameters = {
'input file':{
ParameterModule.PORT_ATTRIBUTE_COMMANDLINE:True,
ParameterModule.PORT_ATTRIBUTE_ISINPUTFILE:True,
},
'output file':{
ParameterModule.PORT_ATTRIBUTE_COMMANDLINE:True,
ParameterModule.PORT_ATTRIBUTE_ISSIDEEFFECT:True,
}
},
parameterOrderings = parameterOrdering,
executable = executable
)
return definition
def createWordCountDefinition(dir=None):
parameterOrdering = DefinitionModule.createParameterOrderingTable()
row = parameterOrdering.addRow()
row.setColumn('source', 'input file')
row.setColumn('target', 'output file')
if dir is None:
dir = os.getcwd().split(os.path.sep) + ['resources', 'testdata', 'TestExecute']
command = dir + ['wordcount.py']
executable = TaskCommandModule.Executable()
executable.stageable(True)
executable.path(command)
executable.staticArgs([])
definition = DefinitionModule.createShellProcessDefinition(
inputParameters = {
'input file':{
ParameterModule.PORT_ATTRIBUTE_COMMANDLINE:True,
ParameterModule.PORT_ATTRIBUTE_ISINPUTFILE:True,
},
'output file':{
ParameterModule.PORT_ATTRIBUTE_COMMANDLINE:True,
ParameterModule.PORT_ATTRIBUTE_ISSIDEEFFECT:True,
}
},
parameterOrderings = parameterOrdering,
executable = executable
)
definition.name('wordcount mapper')
definition.id(ID_WORDCOUNT)
definition.isLibraryDefinition(True)
return definition
def createHadoopStreamingDefinition():
parameterOrdering = DefinitionModule.createParameterOrderingTable()
row = parameterOrdering.addRow()
row.setColumn('source', 'input file')
row.setColumn('target', 'output file')
row = parameterOrdering.addRow()
row.setColumn('source', 'output file')
row.setColumn('target', 'mapper')
row = parameterOrdering.addRow()
row.setColumn('source', 'mapper')
row.setColumn('target', 'reducer')
# TODO:
# need to be able to customize this for each host
executable = HadoopModule.JarExecutable()
executable.stageable(False)
executable.path([HadoopModule.getExecutablePath()])
executable.jarFile([getStreamingJar()])
executable.jarClass([])
definition = DefinitionModule.createShellProcessDefinition(
inputParameters = {
'input file':{
ParameterModule.PORT_ATTRIBUTE_COMMANDLINE:True,
ParameterModule.PORT_ATTRIBUTE_ISINPUTFILE:True,
ParameterModule.PORT_ATTRIBUTE_COMMANDLINE_OPTIONS:{
ParameterModule.COMMANDLINE_PREFIX_FLAG:['-input']
},
},
'output file':{
ParameterModule.PORT_ATTRIBUTE_COMMANDLINE:True,
ParameterModule.PORT_ATTRIBUTE_ISSIDEEFFECT:True,
ParameterModule.PORT_ATTRIBUTE_COMMANDLINE_OPTIONS:{
ParameterModule.COMMANDLINE_PREFIX_FLAG:['-output']
},
},
'mapper':{
ParameterModule.PORT_ATTRIBUTE_COMMANDLINE:True,
ParameterModule.PORT_ATTRIBUTE_ISINPUTFILE:True,
ParameterModule.PORT_ATTRIBUTE_COMMANDLINE_OPTIONS:{
ParameterModule.COMMANDLINE_PREFIX_FLAG:['-mapper']
},
},
'reducer':{
ParameterModule.PORT_ATTRIBUTE_COMMANDLINE:True,
ParameterModule.PORT_ATTRIBUTE_ISINPUTFILE:True,
ParameterModule.PORT_ATTRIBUTE_COMMANDLINE_OPTIONS:{
ParameterModule.COMMANDLINE_PREFIX_FLAG:['-reducer']
},
},
},
parameterOrderings = parameterOrdering,
executable = executable
)
return definition
def createHadoopPipesDefinition():
parameterOrdering = DefinitionModule.createParameterOrderingTable()
row = parameterOrdering.addRow()
row.setColumn('source', 'input file')
row.setColumn('target', 'output file')
# TODO:
# need to be able to customize this for each host
command = ['pipesProgram']
executable = HadoopModule.PipesExecutable()
executable.stageable(False)
executable.path([HadoopModule.getExecutablePath()])
executable.pipesFile(command)
definition = DefinitionModule.createShellProcessDefinition(
inputParameters = {
'input file':{
ParameterModule.PORT_ATTRIBUTE_COMMANDLINE:True,
ParameterModule.PORT_ATTRIBUTE_ISINPUTFILE:True,
ParameterModule.PORT_ATTRIBUTE_COMMANDLINE_OPTIONS:{
ParameterModule.COMMANDLINE_PREFIX_FLAG:['-input']
},
},
'output file':{
ParameterModule.PORT_ATTRIBUTE_COMMANDLINE:True,
ParameterModule.PORT_ATTRIBUTE_ISSIDEEFFECT:True,
ParameterModule.PORT_ATTRIBUTE_COMMANDLINE_OPTIONS:{
ParameterModule.COMMANDLINE_PREFIX_FLAG:['-output']
},
},
},
parameterOrderings = parameterOrdering,
executable = executable
)
return definition
