From 59c2827bcd682a6c78714438ef8d7aaca2ccf13b Mon Sep 17 00:00:00 2001 From: Ellen Spertus Date: Fri, 1 Jun 2018 14:54:17 -0700 Subject: Factored common code out of HBaseRowCountJob and its test into a new companion object. --- scalding/src/main/scala/sandcrawler/HBaseRowCountJob.scala | 14 +++++++++----- 1 file changed, 9 insertions(+), 5 deletions(-) (limited to 'scalding/src/main') diff --git a/scalding/src/main/scala/sandcrawler/HBaseRowCountJob.scala b/scalding/src/main/scala/sandcrawler/HBaseRowCountJob.scala index d47fe60..79ebbb1 100644 --- a/scalding/src/main/scala/sandcrawler/HBaseRowCountJob.scala +++ b/scalding/src/main/scala/sandcrawler/HBaseRowCountJob.scala @@ -16,7 +16,15 @@ class HBaseRowCountJob(args: Args) extends JobBase(args) with HBasePipeConversio val output = args("output") - val hbs = new HBaseSource( + HBaseRowCountJob.getHBaseSource + .read + .debug + .groupAll { _.size('count) } + .write(Tsv(output)) +} + +object HBaseRowCountJob { + def getHBaseSource = new HBaseSource( //"table_name", //"quorum_name:2181", "wbgrp-journal-extract-0-qa", // HBase Table Name @@ -25,8 +33,4 @@ class HBaseRowCountJob(args: Args) extends JobBase(args) with HBasePipeConversio List("file"), List(new Fields("size", "mimetype")), sourceMode = SourceMode.SCAN_ALL) - .read - .debug - .groupAll { _.size('count) } - .write(Tsv(output)) } -- cgit v1.2.3