aboutsummaryrefslogtreecommitdiffstats
path: root/scalding/src
diff options
context:
space:
mode:
authorEllen Spertus <ellen.spertus@gmail.com>2018-07-25 11:23:16 -0700
committerEllen Spertus <ellen.spertus@gmail.com>2018-07-25 11:23:16 -0700
commit980c4af4fbc9d0c62fc75396f2237e5c58863ebf (patch)
treee177e8a4a38247e58b9eb9e0f48e09c9fd1f8c5c /scalding/src
parent773d5c28e2ac6085172aaebf86031358261a7014 (diff)
downloadsandcrawler-980c4af4fbc9d0c62fc75396f2237e5c58863ebf.tar.gz
sandcrawler-980c4af4fbc9d0c62fc75396f2237e5c58863ebf.zip
Checked all fields of first entry in grobid pipe.
Diffstat (limited to 'scalding/src')
-rw-r--r--scalding/src/test/scala/sandcrawler/HBaseCrossrefScoreTest.scala6
1 files changed, 4 insertions, 2 deletions
diff --git a/scalding/src/test/scala/sandcrawler/HBaseCrossrefScoreTest.scala b/scalding/src/test/scala/sandcrawler/HBaseCrossrefScoreTest.scala
index d70c8f2..9402c0a 100644
--- a/scalding/src/test/scala/sandcrawler/HBaseCrossrefScoreTest.scala
+++ b/scalding/src/test/scala/sandcrawler/HBaseCrossrefScoreTest.scala
@@ -183,8 +183,10 @@ class HBaseCrossrefScoreTest extends FunSpec with TupleConversions {
it("should return a 3-element list.") {
assert(outputBuffer.size === 3)
}
- it("should return the right first slug.") {
- val (_, _, slug0) = outputBuffer(0)
+ it("should return the right first entry.") {
+ val (sha1, json, slug0) = outputBuffer(0)
+ assert(sha1 == new String(grobidSampleData(0)(0), "UTF-8"))
+ assert(json == new String(grobidSampleData(0)(1), "UTF-8"))
assert(slug0 == "title1")
}
/*