aboutsummaryrefslogtreecommitdiffstats
path: root/scalding/src/main/scala/sandcrawler/GrobidScorableDumpJob.scala
diff options
context:
space:
mode:
Diffstat (limited to 'scalding/src/main/scala/sandcrawler/GrobidScorableDumpJob.scala')
-rw-r--r--scalding/src/main/scala/sandcrawler/GrobidScorableDumpJob.scala18
1 files changed, 18 insertions, 0 deletions
diff --git a/scalding/src/main/scala/sandcrawler/GrobidScorableDumpJob.scala b/scalding/src/main/scala/sandcrawler/GrobidScorableDumpJob.scala
new file mode 100644
index 0000000..9a8d701
--- /dev/null
+++ b/scalding/src/main/scala/sandcrawler/GrobidScorableDumpJob.scala
@@ -0,0 +1,18 @@
+
+package sandcrawler
+
+import cascading.pipe.Pipe
+import com.twitter.scalding.Args
+import com.twitter.scalding.TypedPipe
+import com.twitter.scalding.TypedTsv
+import parallelai.spyglass.base.JobBase
+
+class GrobidScorableDumpJob(args: Args) extends JobBase(args) {
+
+ val sc1 : Scorable = new GrobidScorable()
+ val pipe1 : TypedPipe[(String, ReduceFeatures)] = sc1.getInputPipe(args)
+
+ pipe1
+ .map { case (slug, features) => (slug, features.json) }
+ .write(TypedTsv[(String, String)](args("output")))
+}