Hi bryan,
Thanks for your suggestion.
i have followed your steps and am have an one doubt regarding your schema
creation for "Sex" column.
i haven't using schema in created flow.
You have said that flow have a field called "Sex" in the schema being used
by the record writer.i'm FreeFormSetWriter only not "CSVRecordWriter"
In that CSVReader i have configured *"Schema Access Strategy"-->"Use String
fields from Header".*
FreeFormSetWriter,i have specified the ${ID},${Name},${Sex} only.
*i wants you to know when i have to create schema for "Sex".*
Please stop me if am understands anything wrong.
*Can you please guide me to create schema for my requirement?*
Thanks,
prabhu
On Fri, Jul 7, 2017 at 5:05 PM, Bryan Bende <[email protected]> wrote:
> Hi Prabhu,
>
> The SimpleCsvFileLookupService is meant to look up one value and add
> it back to the records.
>
> So lets say you want to lookup the gender and add it to the original
> records...
>
> You would configure SimpleCsvFileLookupService with the following:
>
> - Lookup Key Column = ID
> - Lookup Key Value = Sex
>
> When the service starts it will then make a map of ID to Sex so you would
> have:
>
> 2201 -> Male
> 3300 -> Female
>
> Now in LookupRecord you would add a user-defined property of "key" =
> "ID" since ID is the column from the incoming records that would to
> use as the key into the above map.
>
> Then "Result Record Path" should be the field in the records where you
> want the result of the look up go to, so you would want something like
> "/Sex".
>
> You'll also need to have a field called "Sex" in the schema being used
> by the record writer. You could make one schema that has a nullable
> Sex field and have the CsvReader and FreeFormTextWriter both reference
> that schema, or you could let the CsvReader infer the schema from the
> fields (it won't have a sex field) and then use a different schema for
> the writer.
>
> -Bryan
>
>
> On Fri, Jul 7, 2017 at 6:28 AM, prabhu Mahendran
> <[email protected]> wrote:
> > I tried to join two csv file based on id with respect to the below
> > reference.
> >
> >
> > How to join two CSVs with Apache Nifi
> >
> >
> > i'm using NiFi-1.3.0
> >
> >
> > Now i have two csv files.
> >
> >
> > 1.custom.csv
> >
> >
> > No,Name,ID,Age
> >
> > 1,Hik,2201,33
> >
> > 2,Kio,3300,22
> >
> >
> > 2.gender.csv
> >
> >
> > ID,Name,Sex
> >
> > 2201,Hik,Male
> >
> > 3300,Kio,Female
> >
> > I try to combine those tables with "ID" like following endresult.
> >
> >
> > No,Name,Sex,ID,Age
> >
> > 1,Hik,Male,2201,33
> >
> > 2,Kio,Female,3300,22
> >
> > I have using following processor structure.
> >
> >
> > GetFile-SplitText-ExtractText-LookUpRecord-PutFile
> >
> > In that lookup record i have configured
> >
> >
> > RecordReader = "CSVReader"
> >
> > RecordWriter="FreeFormTextRecordSetWriter"
> >
> > LookUpService="SimpleCSVFileLookUpService"
> >
> >
> > ResultRecordPath-->/Key
> >
> > key-->/ID
> >
> > In that LookUpService i have given path of the "gender.csv" and setted
> > LookUpKeyColumn and LookUpValueColumn to be "ID".
> >
> >
> > In that FreeFormTextRecordSetWriter i have given text
> > value"${No},${Name},${ID},${Age},${Sex}"
> >
> >
> > It yields below result only.
> >
> >
> > No,Name,Sex,ID,Age,
> >
> > 1,Hik,Male,2201,33,
> >
> > 2,Kio,Female,3300,22,
> >
> > It doesn't have "sex" column.
> >
> >
> > I think i could not configured correctly.
> >
> >
> > i don't know how to use ResultRecordPath & one dynamic attribute(Key)
> > specification in LookUpRecord?
> >
> >
> > Can anyone guide me to solve my issue?
>
