Hi, For the splitjson, it should be $ instead of $. But as suggested, best is to use record processors.
Pierre Le 25 août 2017 19:06, "Austin Duncan" <adun...@pyaanalytics.com> a écrit : > How would you use convert record? > > On Fri, Aug 25, 2017 at 2:02 PM, Austin Duncan <adun...@pyaanalytics.com> > wrote: > >> heres a sample flow. Sorry about it being vague i just dont want to risk >> it. >> >> On Fri, Aug 25, 2017 at 2:00 PM, Bryan Bende <bbe...@gmail.com> wrote: >> >>> Would it help to go directly from Avro to CSV using ConvertRecord? >>> >>> Rather then making two conversions and splitting. >>> >>> On Fri, Aug 25, 2017 at 1:53 PM, Austin Duncan <adun...@pyaanalytics.com> >>> wrote: >>> > It's hipaa data but I can probably do a work around. Give me a minute >>> > >>> > On Fri, Aug 25, 2017 at 1:51 PM, Joe Witt <joe.w...@gmail.com> wrote: >>> >> >>> >> Austin, >>> >> >>> >> Can you share the sample data and flow template by chance? >>> >> >>> >> Thanks >>> >> Joe >>> >> >>> >> On Fri, Aug 25, 2017 at 1:50 PM, Austin Duncan < >>> adun...@pyaanalytics.com> >>> >> wrote: >>> >> > Hey all, >>> >> > >>> >> > So I am converting an avro schema to json. Then I am trying to >>> split the >>> >> > json so that each of item can be its own row in a csv. I am unable >>> to >>> >> > split >>> >> > the json. i used a JSONPath Expression tester and it seems like the >>> >> > expression '$.' should work. But when I put it into the SplitJson >>> >> > processor >>> >> > i get the error: "'' is invalind because Failed to run validation >>> due to >>> >> > java.langStringIndexOutOfBondsException: String index out of >>> range:2." >>> >> > It >>> >> > seems like there should be a record "results" in the schema but it >>> isnt >>> >> > present in the json and it wont let me split on it. I have tried >>> using >>> >> > the >>> >> > UpdateRecord processor but couldn't figure out how to nest the >>> schema >>> >> > objects with in a new object called "results". Any ideas? >>> >> > >>> >> > Thanks, >>> >> > Austin >>> > >>> > >>> >> >> >
