diff options
Diffstat (limited to 'scalding')
3 files changed, 32 insertions, 18 deletions
diff --git a/scalding/src/main/scala/sandcrawler/HBaseCountJob.scala b/scalding/src/main/scala/sandcrawler/HBaseCountJob.scala new file mode 100644 index 0000000..250f175 --- /dev/null +++ b/scalding/src/main/scala/sandcrawler/HBaseCountJob.scala @@ -0,0 +1,30 @@ +package sandcrawler + +import cascading.property.AppProps +import cascading.tuple.Fields +import com.twitter.scalding._ +import java.util.Properties +import parallelai.spyglass.base.JobBase +import parallelai.spyglass.hbase.{HBaseSource, HBasePipeConversions} +import parallelai.spyglass.hbase.HBaseConstants.SourceMode + +class HBaseCountJob(args: Args, colSpec: String) extends JobBase(args) with HBasePipeConversions { + val output = args("output") + // TODO: Add error checking. + val Col: String = colSpec.split(":")(1) + + HBaseCountJob.getHBaseSource(colSpec) + .read + .fromBytesWritable(Symbol(Col)) + .debug + .groupBy(Col){group => group.size('count)} + .write(Tsv(output)) +} + +object HBaseCountJob { + def getHBaseSource(colSpec: String) = 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(colSpec), + SourceMode.SCAN_ALL) +} diff --git a/scalding/src/main/scala/sandcrawler/HBaseMimeCountJob.scala b/scalding/src/main/scala/sandcrawler/HBaseMimeCountJob.scala index 819a652..911ccb8 100644 --- a/scalding/src/main/scala/sandcrawler/HBaseMimeCountJob.scala +++ b/scalding/src/main/scala/sandcrawler/HBaseMimeCountJob.scala @@ -8,21 +8,5 @@ import parallelai.spyglass.base.JobBase import parallelai.spyglass.hbase.{HBaseSource, HBasePipeConversions} import parallelai.spyglass.hbase.HBaseConstants.SourceMode -class HBaseMimeCountJob(args: Args) extends JobBase(args) with HBasePipeConversions { - val output = args("output") +class HBaseMimeCountJob(args: Args) extends HBaseCountJob(args, "file:mime") {} - HBaseMimeCountJob.getHBaseSource - .read - .fromBytesWritable(List('mime)) - .debug - .groupBy('mime){group => group.size('count)} - .write(Tsv(output)) -} - -object HBaseMimeCountJob { - 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:mime"), - SourceMode.SCAN_ALL) -} diff --git a/scalding/src/test/scala/sandcrawler/HBaseMimeCountTest.scala b/scalding/src/test/scala/sandcrawler/HBaseMimeCountTest.scala index eb6f4ff..cc400c5 100644 --- a/scalding/src/test/scala/sandcrawler/HBaseMimeCountTest.scala +++ b/scalding/src/test/scala/sandcrawler/HBaseMimeCountTest.scala @@ -41,7 +41,7 @@ class HBaseMimeCountTest extends FunSpec with TupleConversions { .arg("app.conf.path", "app.conf") .arg("output", output) .arg("debug", "true") - .source[Tuple](HBaseMimeCountJob.getHBaseSource, + .source[Tuple](HBaseCountJob.getHBaseSource("file:mime"), sampleData.map(l => new Tuple(l.map(s => {new ImmutableBytesWritable(Bytes.toBytes(s))}):_*))) .sink[Tuple](Tsv(output)) { outputBuffer => |