diff options
Diffstat (limited to 'scalding')
6 files changed, 43 insertions, 41 deletions
diff --git a/scalding/src/main/scala/sandcrawler/CrossrefScorable.scala b/scalding/src/main/scala/sandcrawler/CrossrefScorable.scala index ff8201a..5d1eaf5 100644 --- a/scalding/src/main/scala/sandcrawler/CrossrefScorable.scala +++ b/scalding/src/main/scala/sandcrawler/CrossrefScorable.scala @@ -35,7 +35,7 @@ object CrossrefScorable { new MapFeatures(Scorable.NoSlug, json) } else { // bnewbold: not checking that titles(0) is non-null/non-empty; case would be, in JSON, "title": [ null ] - val sf : ScorableFeatures = new ScorableFeatures(title=titles(0), doi=doi) + val sf : ScorableFeatures = ScorableFeatures.create(title=titles(0), doi=doi) new MapFeatures(sf.toSlug, sf.toString) } } else { diff --git a/scalding/src/main/scala/sandcrawler/GrobidScorable.scala b/scalding/src/main/scala/sandcrawler/GrobidScorable.scala index 9a09e05..d7a1eea 100644 --- a/scalding/src/main/scala/sandcrawler/GrobidScorable.scala +++ b/scalding/src/main/scala/sandcrawler/GrobidScorable.scala @@ -45,7 +45,7 @@ object GrobidScorable { case None => MapFeatures(Scorable.NoSlug, json) case Some(map) => { if (map contains "title") { - new ScorableFeatures(Scorable.getString(map, "title"), sha1=key).toMapFeatures + ScorableFeatures.create(title=Scorable.getString(map, "title"), sha1=key).toMapFeatures } else { MapFeatures(Scorable.NoSlug, json) } diff --git a/scalding/src/main/scala/sandcrawler/ScorableFeatures.scala b/scalding/src/main/scala/sandcrawler/ScorableFeatures.scala index 610f1a4..0b9868a 100644 --- a/scalding/src/main/scala/sandcrawler/ScorableFeatures.scala +++ b/scalding/src/main/scala/sandcrawler/ScorableFeatures.scala @@ -5,19 +5,31 @@ import java.io.InputStream import scala.io.Source import scala.util.parsing.json.JSONObject -// Contains features needed to make slug and to score (in combination -// with a second ScorableFeatures). -class ScorableFeatures(title : String, year: Int = 0, doi : String = "", sha1: String = "") { - def toMap() : Map[String, Any] = { Map( - "title" -> (if (title == null) "" else title), - "year" -> year, - "doi" -> (if (doi == null) "" else doi), - "sha1" -> (if (sha1 == null) "" else sha1)) - } +object ScorableFeatures { + // TODO: Add exception handling. + val fileStream : InputStream = getClass.getResourceAsStream("/slug-blacklist.txt") + val SlugBlacklist : Set[String] = Source.fromInputStream(fileStream).getLines.toSet + fileStream.close - override def toString() : String = { - JSONObject(toMap()).toString + // Static factory method + def create(title : String, year : Int = 0, doi : String = "", sha1 : String = "") : ScorableFeatures = { + new ScorableFeatures( + title=if (title == null) "" else title, + year=year, + doi=if (doi == null) "" else doi, + sha1=if (sha1 == null) "" else sha1) } +} + +// Contains features needed to make slug and to score (in combination +// with a second ScorableFeatures). Create with above static factory method. +class ScorableFeatures private(title : String, year: Int = 0, doi : String = "", sha1: String = "") { + + def toMap() : Map[String, Any] = + Map("title" -> title, "year" -> year, "doi" -> doi, "sha1" -> sha1) + + override def toString() : String = + JSONObject(toMap).toString def toSlug() : String = { if (title == null) { @@ -26,22 +38,10 @@ class ScorableFeatures(title : String, year: Int = 0, doi : String = "", sha1: S val unaccented = StringUtilities.removeAccents(title) // Remove punctuation val slug = StringUtilities.removePunctuation((unaccented.toLowerCase())).replaceAll("\\s", "") - // scalastyle:off if.brace - if (slug.isEmpty || slug == null || (ScorableFeatures.SlugBlacklist contains slug)) - Scorable.NoSlug - else - slug + if (slug.isEmpty || slug == null || (ScorableFeatures.SlugBlacklist contains slug)) Scorable.NoSlug else slug } } - def toMapFeatures : MapFeatures = { + def toMapFeatures : MapFeatures = MapFeatures(toSlug, toString) - } -} - -object ScorableFeatures { - // TODO: Add exception handling. - val fileStream : InputStream = getClass.getResourceAsStream("/slug-blacklist.txt") - val SlugBlacklist : Set[String] = Source.fromInputStream(fileStream).getLines.toSet - fileStream.close } diff --git a/scalding/src/test/scala/sandcrawler/HBaseStatusCountTest.scala b/scalding/src/test/scala/sandcrawler/HBaseStatusCountTest.scala index 0da0b9c..3291670 100644 --- a/scalding/src/test/scala/sandcrawler/HBaseStatusCountTest.scala +++ b/scalding/src/test/scala/sandcrawler/HBaseStatusCountTest.scala @@ -29,17 +29,17 @@ class HBaseStatusCountTest extends FunSpec with TupleConversions { val statusType1Bytes = Bytes.toBytes(statusType1) val statusType2Bytes = Bytes.toBytes(statusType2) - val sampleData : List[List[Array[Byte]]] = List( - // TODO(bnewbold): now to express a null (empty value) in this list? - List(Bytes.toBytes("sha1:K2DKSSVTXWPRMFDTWSTCQW3RVWRIOV3Q"), statusType1Bytes), - List(Bytes.toBytes("sha1:C3YNNEGH5WAG5ZAAXWAEBNXJWT6CZ3WU"), statusType1Bytes), - List(Bytes.toBytes("sha1:SDKUVHC3YNNEGH5WAG5ZAAXWAEBNX4WT"), statusType2Bytes), - List(Bytes.toBytes("sha1:35985C3YNNEGH5WAG5ZAAXWAEBNXJW56"), statusType2Bytes), - List(Bytes.toBytes("sha1:885C3YNNEGH5WAG5ZAAXWA8BNXJWT6CZ"), statusType2Bytes), - List(Bytes.toBytes("sha1:00904C3YNNEGH5WAG5ZA9XWAEBNXJWT6"), statusType2Bytes), - List(Bytes.toBytes("sha1:249C3YNNEGH5WAG5ZAAXWAEBNXJWT6CZ"), statusType1Bytes), - List(Bytes.toBytes("sha1:095893C3YNNEGH5WAG5ZAAXWAEBNXJWT"), statusType2Bytes) - ) + // TODO(bnewbold): now to express a null (empty value) in this list? + val sampleData : List[List[Array[Byte]]] = List( + ("sha1:K2DKSSVTXWPRMFDTWSTCQW3RVWRIOV3Q", statusType1Bytes), + ("sha1:C3YNNEGH5WAG5ZAAXWAEBNXJWT6CZ3WU", statusType1Bytes), + ("sha1:SDKUVHC3YNNEGH5WAG5ZAAXWAEBNX4WT", statusType2Bytes), + ("sha1:35985C3YNNEGH5WAG5ZAAXWAEBNXJW56", statusType2Bytes), + ("sha1:885C3YNNEGH5WAG5ZAAXWA8BNXJWT6CZ", statusType2Bytes), + ("sha1:00904C3YNNEGH5WAG5ZA9XWAEBNXJWT6", statusType2Bytes), + ("sha1:249C3YNNEGH5WAG5ZAAXWAEBNXJWT6CZ", statusType1Bytes), + ("sha1:095893C3YNNEGH5WAG5ZAAXWAEBNXJWT", statusType2Bytes)) + .map(pair => List(Bytes.toBytes(pair._1), pair._2)) val statusType1Count = sampleData.count(lst => lst(1) == statusType1Bytes) val statusType2Count = sampleData.count(lst => lst(1) == statusType2Bytes) diff --git a/scalding/src/test/scala/sandcrawler/ScorableFeaturesTest.scala b/scalding/src/test/scala/sandcrawler/ScorableFeaturesTest.scala index a9a90ec..5a22ef8 100644 --- a/scalding/src/test/scala/sandcrawler/ScorableFeaturesTest.scala +++ b/scalding/src/test/scala/sandcrawler/ScorableFeaturesTest.scala @@ -26,12 +26,12 @@ class ScorableFeaturesTest extends FlatSpec with Matchers { } private def titleToSlug(s : String) : String = { - new ScorableFeatures(title = s).toSlug + ScorableFeatures.create(title = s).toSlug } "toMapFeatures()" should "work with gnarly inputs" in { - new ScorableFeatures(title = null).toMapFeatures - new ScorableFeatures(title = "something", doi = null, sha1 = null, year = 123).toMapFeatures + ScorableFeatures.create(title = null).toMapFeatures + ScorableFeatures.create(title = "something", doi = null, sha1 = null, year = 123).toMapFeatures } "mapToSlug()" should "extract the parts of titles before a colon" in { diff --git a/scalding/src/test/scala/sandcrawler/ScoreJobTest.scala b/scalding/src/test/scala/sandcrawler/ScoreJobTest.scala index 5516869..00e4659 100644 --- a/scalding/src/test/scala/sandcrawler/ScoreJobTest.scala +++ b/scalding/src/test/scala/sandcrawler/ScoreJobTest.scala @@ -161,10 +161,12 @@ class ScoreJobTest extends FlatSpec with Matchers { .map { case (sha, json, status) => List(Bytes.toBytes(sha), Bytes.toBytes(json), Bytes.toBytes(status)) } .map { l => new Tuple(l.map(s => {new ImmutableBytesWritable(s)}):_*) } + // scalastyle:off null // Add example of lines without GROBID data // scalastyle:off null val SampleData = SampleDataHead :+ new Tuple( new ImmutableBytesWritable(Bytes.toBytes("sha1:35985C3YNNEGH5WAG5ZAA88888888888")), null, null) + // scalastyle:on null JobTest("sandcrawler.ScoreJob") .arg("test", "") |