Our goal is to make Drill as easy as possible to write Scan operators for (in fact, any operator should be easy to implement). In fact, one of our other goals to allow people to utilize their existing Hive SerDes with a slight performance penalty. We're in the process now of working on our first full scanning implementation. Timothy Chen is leading that work. Once we get the first one done, we'll work through building a set of documentation and steps for Scan implementation.
On Wed, Jun 5, 2013 at 12:07 AM, Ted Dunning <[email protected]> wrote: > Btw... in the Drill talk that Michael and I gave at Buzzwords, one of the > most popular requests was documentation for how to write scan operators. > > > > > On Tue, Jun 4, 2013 at 8:19 PM, Avinash Dongre <[email protected]>wrote: > >> Hi Drillers, >> Is there any document on how to write SCAN Operators. >> I would like to see possibility of writing SCAN Operator for our in -house >> database. >> >> >> Thanks >> Avinash >>
