diff options
Diffstat (limited to 'scalding')
5 files changed, 46 insertions, 36 deletions
diff --git a/scalding/src/main/scala/sandcrawler/CrossrefScorable.scala b/scalding/src/main/scala/sandcrawler/CrossrefScorable.scala index 667a5cc..e257152 100644 --- a/scalding/src/main/scala/sandcrawler/CrossrefScorable.scala +++ b/scalding/src/main/scala/sandcrawler/CrossrefScorable.scala @@ -41,26 +41,31 @@ class CrossrefScorable extends Scorable with HBasePipeConversions { def getFeaturesPipe(args : Args)(implicit mode : Mode, flowDef : FlowDef) : TypedPipe[MapFeatures] = { getSource(args).read .toTypedPipe[String](new Fields("line")) - .map{ json : String => - Scorable.jsonToMap(json) match { - case None => MapFeatures(Scorable.NoSlug, json) - case Some(map) => { - if ((map contains "title") && (map contains "DOI")) { - val titles = map("title").asInstanceOf[List[String]] - if (titles.isEmpty) { - new MapFeatures(Scorable.NoSlug, json) - } else { - val title = titles(0) - val map2 = Scorable.toScorableMap(title=titles(0), doi=map("DOI").asInstanceOf[String]) - new MapFeatures( - Scorable.mapToSlug(map2), - JSONObject(map2).toString) - } - } else { - new MapFeatures(Scorable.NoSlug, json) - } + .map { CrossrefScorable.jsonToMapFeatures(_) } + } +} + +object CrossrefScorable { + def jsonToMapFeatures(json : String) : MapFeatures = { + Scorable.jsonToMap(json) match { + case None => MapFeatures(Scorable.NoSlug, json) + case Some(map) => { + if ((map contains "titles") && (map contains "DOI")) { + val titles = map("titles").asInstanceOf[List[String]] + val doi = Scorable.getString(map, "DOI") + if (titles.isEmpty || titles == null || doi.isEmpty || doi == null) { + new MapFeatures(Scorable.NoSlug, json) + } else { + val title = titles(0) + val map2 = Scorable.toScorableMap(title=title, doi=doi) + new MapFeatures( + Scorable.mapToSlug(map2), + JSONObject(map2).toString) } + } else { + new MapFeatures(Scorable.NoSlug, json) } } + } } } diff --git a/scalding/src/main/scala/sandcrawler/Scorable.scala b/scalding/src/main/scala/sandcrawler/Scorable.scala index 929461b..a256fa4 100644 --- a/scalding/src/main/scala/sandcrawler/Scorable.scala +++ b/scalding/src/main/scala/sandcrawler/Scorable.scala @@ -7,7 +7,6 @@ import scala.util.parsing.json.JSONObject import cascading.flow.FlowDef import com.twitter.scalding._ import com.twitter.scalding.typed.TDsl._ -//import TDsl._ case class MapFeatures(slug : String, json : String) case class ReduceFeatures(json : String) diff --git a/scalding/src/test/scala/sandcrawler/CrossrefScorableTest.scala b/scalding/src/test/scala/sandcrawler/CrossrefScorableTest.scala index 1c35d66..dc6f347 100644 --- a/scalding/src/test/scala/sandcrawler/CrossrefScorableTest.scala +++ b/scalding/src/test/scala/sandcrawler/CrossrefScorableTest.scala @@ -66,23 +66,23 @@ class CrossrefScorableTest extends FlatSpec with Matchers { val MalformedCrossrefString = CrossrefString.replace("}", "") // Unit tests - "simplifyJson()" should "return None for bad JSON" in { - CrossrefScorable.simplifyJson("") shouldBe None - CrossrefScorable.simplifyJson(MalformedCrossrefString) shouldBe None + "CrossrefScorable.jsonToMapFeatures()" should "handle invalid JSON" in { + val result = CrossrefScorable.jsonToMapFeatures(MalformedCrossrefString) + result.slug shouldBe Scorable.NoSlug } - it should "return None for JSON lacking title" in { - CrossrefScorable.simplifyJson(CrossrefStringWithoutTitle) shouldBe None + it should "handle missing title" in { + val result = CrossrefScorable.jsonToMapFeatures(CrossrefStringWithoutTitle) + result.slug shouldBe Scorable.NoSlug } - it should "return appropriate result for valid JSON" in { - CrossrefScorable.simplifyJson(CrossrefStringWithTitle) match { - case None => fail("None unexpectedly returned by simplifyJson") + it should "handle valid input" in { + val result = CrossrefScorable.jsonToMapFeatures(CrossrefStringWithTitle) + result.slug shouldBe "dummyexamplefile" + Scorable.jsonToMap(result.json) match { + case None => fail() case Some(map) => { - Scorable.isScorableMap(map) shouldBe true - map.size shouldBe 1 - map.keys should contain ("title") - map("title") shouldBe "SomeTitle" + map("title").asInstanceOf[String] shouldBe "Dummy Example File" } } } diff --git a/scalding/src/test/scala/sandcrawler/GrobidScorableTest.scala b/scalding/src/test/scala/sandcrawler/GrobidScorableTest.scala index 3fcd856..4b958b9 100644 --- a/scalding/src/test/scala/sandcrawler/GrobidScorableTest.scala +++ b/scalding/src/test/scala/sandcrawler/GrobidScorableTest.scala @@ -77,6 +77,7 @@ class GrobidScorableTest extends FlatSpec with Matchers { Scorable.jsonToMap(result.json) match { case None => fail() case Some(map) => { + map should contain key "title" map("title").asInstanceOf[String] shouldBe "Dummy Example File" } } diff --git a/scalding/src/test/scala/sandcrawler/ScoreJobTest.scala b/scalding/src/test/scala/sandcrawler/ScoreJobTest.scala index 8acb454..8436817 100644 --- a/scalding/src/test/scala/sandcrawler/ScoreJobTest.scala +++ b/scalding/src/test/scala/sandcrawler/ScoreJobTest.scala @@ -149,11 +149,16 @@ class ScoreJobTest extends FlatSpec with Matchers { 2 -> CrossrefString.replace("<<TITLE>>", "Title 1: TNG 3").replace("<<DOI>>", "DOI-0.75"), 3 -> CrossrefString.replace("<<TITLE>>", "Title 2: Rebooted").replace("<<DOI>>", "DOI-1"))) .sink[(String, Int, String, String)](TypedTsv[(String, Int, String, String)](output)) { - // Grobid titles: - // "Title 1", "Title 2: TNG", "Title 3: The Sequel" - // crossref slugs: - // "Title 1: TNG", "Title 1: TNG 2", "Title 1: TNG 3", "Title 2 Rebooted" - // Join should have 3 "Title 1" slugs and 1 "Title 2" slug + // Grobid titles and slugs (in parentheses): + // Title 1 (title1) + // Title 2: TNG (title2) + // Title 3: The Sequel (title3) + // crossref titles and slugs (in parentheses): + // Title 1: TNG (title1) + // Title 1: TNG 2 (title1) + // Title 1: TNG 3 (title1) + // Title 2 Rebooted (title2rebooted) + // Join should have 3 "title1" slugs and 1 "title2" slug outputBuffer => "The pipeline" should "return a 4-element list" in { outputBuffer should have length 4 |