aboutsummaryrefslogtreecommitdiffstats
path: root/scalding/src/main/scala/sandcrawler/HBaseRowCountJob.scala
diff options
context:
space:
mode:
authorbnewbold <bnewbold@archive.org>2018-06-04 22:05:54 +0000
committerbnewbold <bnewbold@archive.org>2018-06-04 22:05:54 +0000
commitb21cc924224202aba412370acafaaa4846bb4a8d (patch)
tree3d0fcc2f2bee55eb038686c566bdab1763e65c74 /scalding/src/main/scala/sandcrawler/HBaseRowCountJob.scala
parent70069a78c8798352b5bef815a3fd4aa9e9b52394 (diff)
parent9a18da82738c16d3066e02e886ee84e04156889f (diff)
downloadsandcrawler-b21cc924224202aba412370acafaaa4846bb4a8d.tar.gz
sandcrawler-b21cc924224202aba412370acafaaa4846bb4a8d.zip
Merge branch 'refactoring' into 'master'
Refactoring to add, use, and test class HBaseBuilder to eliminate duplicated code and facilitate HBaseSource creation See merge request webgroup/sandcrawler!1
Diffstat (limited to 'scalding/src/main/scala/sandcrawler/HBaseRowCountJob.scala')
-rw-r--r--scalding/src/main/scala/sandcrawler/HBaseRowCountJob.scala18
1 files changed, 9 insertions, 9 deletions
diff --git a/scalding/src/main/scala/sandcrawler/HBaseRowCountJob.scala b/scalding/src/main/scala/sandcrawler/HBaseRowCountJob.scala
index d47fe60..9d074ab 100644
--- a/scalding/src/main/scala/sandcrawler/HBaseRowCountJob.scala
+++ b/scalding/src/main/scala/sandcrawler/HBaseRowCountJob.scala
@@ -16,17 +16,17 @@ class HBaseRowCountJob(args: Args) extends JobBase(args) with HBasePipeConversio
val output = args("output")
- val hbs = new HBaseSource(
- //"table_name",
- //"quorum_name:2181",
- "wbgrp-journal-extract-0-qa", // HBase Table Name
- "mtrcs-zk1.us.archive.org:2181", // HBase Zookeeper server (to get runtime config info; can be array?)
- new Fields("key"),
- List("file"),
- List(new Fields("size", "mimetype")),
- sourceMode = SourceMode.SCAN_ALL)
+ HBaseRowCountJob.getHBaseSource
.read
.debug
.groupAll { _.size('count) }
.write(Tsv(output))
}
+
+object HBaseRowCountJob {
+ def getHBaseSource = HBaseBuilder.build(
+ "wbgrp-journal-extract-0-qa", // HBase Table Name
+ "mtrcs-zk1.us.archive.org:2181", // HBase Zookeeper server (to get runtime config info; can be array?)
+ List("file:size", "file:mime"),
+ SourceMode.SCAN_ALL)
+}