I ended up getting a little creative by chaining a few Puts with multiple RouteOnAttributes. Worked quite well so far.
On Thu, Dec 20, 2018 at 1:59 PM Mark Payne <[email protected]> wrote: > Mike, > > I do think that RouteOnAttribute is the correct play here, given the > details at hand. > Other options, should that be too cumbersome to configure may be > UpdateAttribute using the > Advanced configuration, which works sort of like a mini rules engine. Or, > to be honest, > this sounds like it could be a great use of ExecuteScript. > > If you want to create a new Processor, then by all means do so, but I > think we'd need a better > description of the use case for it and description of where the other > approaches fall short and how this new > Processor would fill in the gap in order for it to be warranted as a > standard processor. > > Thanks > -Mark > > > On Dec 20, 2018, at 12:19 PM, Mike Thomsen <[email protected]> > wrote: > > > > I might still hack out a prototype and send it as a PR because it would > be > > like a really beefed up version of RouteOnAttribute that is built around > > policy enforcement with attributes. > > > > On Thu, Dec 20, 2018 at 12:18 PM Mike Thomsen <[email protected]> > > wrote: > > > >> This is probably enough: > >> > https://nifi.apache.org/docs/nifi-docs/html/expression-language-guide.html#allattributes > >> > >> I missed that because I was thinking "how do I do a bunch of 'and must > >> exist' statements." > >> > >> Thanks, > >> > >> Mike > >> > >> On Thu, Dec 20, 2018 at 12:01 PM Mark Payne <[email protected]> > wrote: > >> > >>> Hey Mike, > >>> > >>> I would recommend RouteOnAttribute for that. > >>> > >>> Thanks > >>> -Mark > >>> > >>> Sent from my iPhone > >>> > >>>> On Dec 20, 2018, at 11:45 AM, Mike Thomsen <[email protected]> > >>> wrote: > >>>> > >>>> Do we have any processors that would be particularly good at checking > a > >>>> flowfile for the existence of certain mandatory attributes and then > >>> routing > >>>> flowfiles missing them to a failure relationship? > >>>> > >>>> I have a processor sketched out in my head for this already, but don't > >>> want > >>>> to reinvent the wheel. > >>>> > >>>> Thanks, > >>>> > >>>> Mike > >>> > >> > >
