aboutsummaryrefslogtreecommitdiffstats
path: root/scalding/src/test/scala/example/SimpleHBaseSourceExampleTest.scala
diff options
context:
space:
mode:
authorBryan Newbold <bnewbold@archive.org>2018-05-23 12:27:59 -0700
committerBryan Newbold <bnewbold@archive.org>2018-05-24 00:02:36 -0700
commit4ba428db30593b67283dd90b92141f99840dc78e (patch)
treef63c8e146e7f90a530abfebdb993ab45d57426d5 /scalding/src/test/scala/example/SimpleHBaseSourceExampleTest.scala
parent29e4a83ff76da07bc6ad5d3f49d746ee0bc72023 (diff)
downloadsandcrawler-4ba428db30593b67283dd90b92141f99840dc78e.tar.gz
sandcrawler-4ba428db30593b67283dd90b92141f99840dc78e.zip
rename jvm/scalding directories
Diffstat (limited to 'scalding/src/test/scala/example/SimpleHBaseSourceExampleTest.scala')
-rw-r--r--scalding/src/test/scala/example/SimpleHBaseSourceExampleTest.scala58
1 files changed, 58 insertions, 0 deletions
diff --git a/scalding/src/test/scala/example/SimpleHBaseSourceExampleTest.scala b/scalding/src/test/scala/example/SimpleHBaseSourceExampleTest.scala
new file mode 100644
index 0000000..cf068c1
--- /dev/null
+++ b/scalding/src/test/scala/example/SimpleHBaseSourceExampleTest.scala
@@ -0,0 +1,58 @@
+package example
+
+import org.junit.runner.RunWith
+import com.twitter.scalding.{JobTest, TupleConversions}
+import org.scalatest.FunSpec
+import org.scalatest.junit.JUnitRunner
+import org.slf4j.LoggerFactory
+import org.apache.hadoop.hbase.io.ImmutableBytesWritable
+import cascading.tuple.{Tuple, Fields}
+import org.apache.hadoop.hbase.util.Bytes
+import scala._
+import com.twitter.scalding.Tsv
+import parallelai.spyglass.hbase.HBaseSource
+import parallelai.spyglass.hbase.HBaseConstants.SourceMode
+
+/**
+ * Example of how to define tests for HBaseSource
+ */
+@RunWith(classOf[JUnitRunner])
+class SimpleHBaseSourceExampleTest extends FunSpec with TupleConversions {
+
+ val output = "/tmp/testOutput"
+
+ val log = LoggerFactory.getLogger(this.getClass.getName)
+
+ val sampleData = List(
+ List("1", "kk1", "pp1"),
+ List("2", "kk2", "pp2"),
+ List("3", "kk3", "pp3")
+ )
+
+ JobTest("example.SimpleHBaseSourceExample")
+ .arg("test", "")
+ .arg("app.conf.path", "app.conf")
+ .arg("output", output)
+ .arg("debug", "true")
+ .source[Tuple](
+ new HBaseSource(
+ "table_name",
+ "mtrcs-zk1.us.archive.org:2181",
+ new Fields("key"),
+ List("column_family"),
+ List(new Fields("column_name1", "column_name2")),
+ sourceMode = SourceMode.GET_LIST, keyList = List("1", "2", "3")),
+ sampleData.map(l => new Tuple(l.map(s => {new ImmutableBytesWritable(Bytes.toBytes(s))}):_*)))
+ .sink[Tuple](Tsv(output format "get_list")) {
+ outputBuffer =>
+ log.debug("Output => " + outputBuffer)
+
+ it("should return the test data provided.") {
+ println("outputBuffer.size => " + outputBuffer.size)
+ assert(outputBuffer.size === 3)
+ }
+ }
+ .run
+ .finish
+
+}