I checked my environnement. I do not see anything wrong. I am on Mac OS X, I installed flink 1.3.2 with Sierra. I start the Scala shell with script.
The integer flow is generated by a simulator. I can send the python code if it helps. Thanks a lot for your help Philippe > Le 10 sept. 2017 à 18:31, Ted Yu <[email protected]> a écrit : > > What if you add the following at the beginning: > > import > org.apache.flink.table.plan.nodes.datastream.DataStreamGroupWindowAggregate._ > > On Sun, Sep 10, 2017 at 9:18 AM, philippe <[email protected]> wrote: > >> Thanks. >> >> Sorry, that’s a copy/paste artefact. Here is an exact copy: >> >> // Entering paste mode (ctrl-D to finish) >> >> case class MonEntier(classe: Int, valeur: Int) >> val stream = senv.socketTextStream("localhost", 9000, '\n') >> >> val w = stream.map ( { x => Tuple1(x.toInt) } ) >> .map( {y => MonEntier(y._1 % 3, y._1) } ) >> .windowAll(TumblingProcessingTimeWindows. >> of(Time.seconds(5))) >> .fold("") { (acc, v) => acc + " - " + v.valeur } >> .print() >> >> senv.execute("Ma gestion de fenêtres ") >> >> >> // Exiting paste mode, now interpreting. >> >> <console>:68: error: not found: value TumblingProcessingTimeWindows >> .windowAll(TumblingProcessingTimeWindows. >> of(Time.seconds(5))) >> >> >> >> >>> Le 10 sept. 2017 à 18:14, Ted Yu <[email protected]> a écrit : >>> >>> Is there extra '=' before the windowAll() call ? >>> >>> On Sun, Sep 10, 2017 at 1:01 AM, philippe <[email protected]> >> wrote: >>> >>>> >>>> Hi all, >>>> >>>> I am trying to run a simple example in the Scala shell: >>>> >>>> >>>> case class MonEntier(classe: Int, valeur: Int) >>>> val stream =3D senv.socketTextStream("localhost", 9000, '\n') >>>> val w =3D stream.map ( { x =3D> Tuple1(x.toInt) } ) >>>> .map( {y =3D> MonEntier(y._1 % 3, y._1) } ) >>>> = >>>> .windowAll(TumblingProcessingTimeWindows.of(Time.seconds(5))) >>>> .print() >>>> >>>> >>>> The socket at 9000 sends a stream of integers >>>> >>>> I get a compilation error: >>>> >>>> <console>:68: error: not found: value TumblingProcessingTimeWindows >>>> .windowAll(new = >>>> TumblingProcessingTimeWindows.of(Time.seconds(5))) >>>> >>>> >>>> The snippet directly comes from the doc. Help welcome ! >>>> >>>> Philippe >>>> >> >>
