Thanks, that's what I figured. I probably can't use InferAvroSchema because this data is pretty tricky in terms of them suddenly deciding to change the types.
On Thu, Aug 17, 2017 at 6:18 PM, Matt Burgess <[email protected]> wrote: > Mike, > > String is usually the safe bet, InferAvroSchema and ExecuteSQL default to > String if they can't figure out what type to use. > > Regards, > Matt > > > > On Aug 17, 2017, at 6:00 PM, Mike Thomsen <[email protected]> > wrote: > > > > Is it safe to choose "string" as a default type with Avro? I'm trudging > through some really dirty data right now and that seems to behave fine when > I I do something like this: > > > > Flowfile content: > > > > { > > "x": 1 > > } > > > > Avro field definition: > > > > { "name": "x", "type": ["null", "string"] } > > > > Where sometimes X is an int and others it's a string (don't ask me why, > I didn't set up ETL on this data). > > > > I did a simple flow that did just this and it **seemed** to work great, > but I thought I'd get a second opinion before spending my client's time > building something big around it. > > > > Thanks, > > > > Mike >
