Hello Varadha, I am actually working on the writer interface for Drill right now. Unfortunately we currently only have reading implementations for JSON and Parquet and neither of those storage engines can export data.
I am working on designing a unified reader/writer interface that will enable us to add more formats faster, while simultaneously adding read and write support with each implementation. I will be posting my patches to the list for review soon, I invite you to take a look if you have an interest in Drill from a development perspective. Regards, Jason Altekruse On Sat, Oct 19, 2013 at 1:06 PM, Varadharajan M <[email protected]>wrote: > Hi All, > > I've a logical plan, thats intended to read from a json file and write the > output to another file. I was under the impression that the write operation > could be done with the help of 'store' operation. Here is my logical plan: > > { > "head" : { > "type" : "APACHE_DRILL_LOGICAL", > "version" : "1", > "generator" : { > "type" : "optiq", > "info" : "na" > } > }, > > "storage" : { > "jsonl" : { > "type" : "json", > "dfsName" : "file:///" > }, > "filel" : { > "type" : "fs", > "root" : "file:///" > } > }, > > "query" : [ { > "op" : "scan", > "memo" : "initial_scan", > "ref" : "_MAP", > "storageengine" : "jsonl", > "selection" : [ { > "path" : "/tmp/students.json" > } ], > "@id" : 1 > }, > > { > "op" : "store", > "input" : 1, > "memo" : "output sink", > "storageengine": "filel", > "@id" : 2, > target: { > file: "file:///tmp/do.json" > } > } ] > } > > > Now, if i try to run this logical plan with the submit_plan shell script, i > get the output to the console rather than to the specified file (That is, > there is no data written to the file). Am i missing anything here? > > - Varadha >
