diff options
Diffstat (limited to 'scalding/src/test/scala/sandcrawler')
15 files changed, 1497 insertions, 35 deletions
diff --git a/scalding/src/test/scala/sandcrawler/CrossrefScorableTest.scala b/scalding/src/test/scala/sandcrawler/CrossrefScorableTest.scala new file mode 100644 index 0000000..8302b8f --- /dev/null +++ b/scalding/src/test/scala/sandcrawler/CrossrefScorableTest.scala @@ -0,0 +1,172 @@ +package sandcrawler + +import cascading.tuple.Fields +import cascading.tuple.Tuple +import com.twitter.scalding.JobTest +import com.twitter.scalding.TextLine +import com.twitter.scalding.TupleConversions +import com.twitter.scalding.TypedTsv +import org.apache.hadoop.hbase.io.ImmutableBytesWritable +import org.apache.hadoop.hbase.util.Bytes +import org.scalatest._ +import parallelai.spyglass.hbase.HBaseConstants.SourceMode + +class CrossrefScorableTest extends FlatSpec with Matchers { + // scalastyle:off + val CrossrefString = +""" +{ "_id" : { "$oid" : "5a553d5988a035a45bf50ed3" }, + "indexed" : { "date-parts" : [ [ 2017, 10, 23 ] ], + "date-time" : "2017-10-23T17:19:16Z", + "timestamp" : { "$numberLong" : "1508779156477" } }, + "reference-count" : 0, + "publisher" : "Elsevier BV", + "issue" : "3", + "license" : [ { "URL" : "http://www.elsevier.com/tdm/userlicense/1.0/", + "start" : { "date-parts" : [ [ 1996, 1, 1 ] ], + "date-time" : "1996-01-01T00:00:00Z", + "timestamp" : { "$numberLong" : "820454400000" } }, + "delay-in-days" : 0, "content-version" : "tdm" }], + "content-domain" : { "domain" : [], "crossmark-restriction" : false }, + "published-print" : { "date-parts" : [ [ 1996 ] ] }, + "DOI" : "<<DOI>>", + "type" : "journal-article", + "created" : { "date-parts" : [ [ 2002, 7, 25 ] ], + "date-time" : "2002-07-25T15:09:41Z", + "timestamp" : { "$numberLong" : "1027609781000" } }, + "page" : "186-187", + "source" : "Crossref", + "is-referenced-by-count" : 0, + "title" : [ "<<TITLE>>" ], + "prefix" : "10.1016", + "volume" : "9", + "author" : [ { "given" : "W", "family" : "Gaier", "affiliation" : [] } ], + "member" : "78", + "container-title" : [ "Journal de Pédiatrie et de Puériculture" ], + "link" : [ { "URL" : "http://api.elsevier.com/content/article/PII:0987-7983(96)87729-2?httpAccept=text/xml", + "content-type" : "text/xml", + "content-version" : "vor", + "intended-application" : "text-mining" }, + { "URL" : + "http://api.elsevier.com/content/article/PII:0987-7983(96)87729-2?httpAccept=text/plain", + "content-type" : "text/plain", + "content-version" : "vor", + "intended-application" : "text-mining" } ], + "deposited" : { "date-parts" : [ [ 2015, 9, 3 ] ], + "date-time" : "2015-09-03T10:03:43Z", + "timestamp" : { "$numberLong" : "1441274623000" } }, + "score" : 1, + "issued" : { "date-parts" : [ [ 1996 ] ] }, + "references-count" : 0, + "alternative-id" : [ "0987-7983(96)87729-2" ], + "URL" : "http://dx.doi.org/10.1016/0987-7983(96)87729-2", + "ISSN" : [ "0987-7983" ], + "issn-type" : [ { "value" : "0987-7983", "type" : "print" } ], + "subject" : [ "Pediatrics, Perinatology, and Child Health" ] +} +""".replace("<<DOI>>", "10.123/aBc") + // scalastyle:on + val CrossrefStringWithGoodTitle = CrossrefString.replace("<<TITLE>>", "Some Title") + val CrossrefStringWithMaximumTitle = CrossrefString.replace("<<TITLE>>", "T" * Scorable.MaxTitleLength) + val CrossrefStringWithExcessiveTitle = CrossrefString.replace("<<TITLE>>", "T" * Scorable.MaxTitleLength + "0") + val CrossrefStringWithNullTitle = CrossrefString.replace("\"<<TITLE>>\"", "null") + val CrossrefStringWithEmptyTitle = CrossrefString.replace("<<TITLE>>", "") + val CrossrefStringWithoutTitle = CrossrefString.replace("title", "nottitle") + val MalformedCrossrefString = CrossrefString.replace("}", "") + val CrossrefStringWithNoAuthors = CrossrefString.replace("<<TITLE>>", "Some Valid Title").replace("author", "no-author") + val CrossrefStringWrongType = CrossrefString.replace("<<TITLE>>", "Some Valid Title").replace("journal-article", "other") + val CrossrefStringNoType = CrossrefString.replace("<<TITLE>>", "Some Valid Title").replace("type", "not-type") + + // Unit tests + "CrossrefScorable.jsonToMapFeatures()" should "handle invalid JSON" in { + CrossrefScorable.jsonToMapFeatures(MalformedCrossrefString) should be (None) + } + + it should "handle missing title" in { + CrossrefScorable.jsonToMapFeatures(CrossrefStringWithoutTitle) should be (None) + } + + it should "handle null title" in { + CrossrefScorable.jsonToMapFeatures(CrossrefStringWithNullTitle) should be (None) + } + + it should "handle empty title" in { + CrossrefScorable.jsonToMapFeatures(CrossrefStringWithEmptyTitle) should be (None) + } + + it should "handle subtitle" in { + CrossrefScorable.jsonToMapFeatures( + """{"title": ["short but not too short"], "subtitle": ["just right!"], "DOI": "10.123/asdf", "type":"journal-article","author":[{ "given" : "W", "family" : "Gaier"}]}""") match { + case None => fail() + case Some(result) => result.slug shouldBe "shortbutnottooshortjustright" + } + } + + it should "handle empty subtitle" in { + CrossrefScorable.jsonToMapFeatures( + """{"title": ["short but not too short"], "subtitle": [""], "DOI": "10.123/asdf", "type":"journal-article", "author":[{ "given" : "W", "family" : "Gaier"}]}""") match { + case None => fail() + case Some(result) => result.slug shouldBe "shortbutnottooshort" + } + } + + it should "handle null subtitle" in { + CrossrefScorable.jsonToMapFeatures( + """{"title": ["short but not too short"], "subtitle": [null], "DOI": "10.123/asdf", "type":"journal-article", "author":[{ "given" : "W", "family" : "Gaier"}]}""") match { + case None => fail() + case Some(result) => result.slug shouldBe "shortbutnottooshort" + } + } + + it should "handle missing authors" in { + CrossrefScorable.jsonToMapFeatures(CrossrefStringWithNoAuthors) should be (None) + } + + it should "handle valid input" in { + CrossrefScorable.jsonToMapFeatures(CrossrefStringWithGoodTitle) match { + case None => fail() + case Some(result) => { + result.slug shouldBe "sometitle" + Scorable.jsonToMap(result.json) match { + case None => fail() + case Some(map) => { + map("title").asInstanceOf[String] shouldBe "Some Title" + map("doi").asInstanceOf[String] shouldBe "10.123/abc" + // TODO: full name? not just a string? + map("authors").asInstanceOf[List[String]] shouldBe List("Gaier") + map("year").asInstanceOf[Double].toInt shouldBe 2002 + } + } + } + } + } + + "CrossrefScorable.keepRecord()" should "return true for valid JSON with title" in { + CrossrefScorable.keepRecord(CrossrefStringWithGoodTitle) shouldBe true + } + + it should "return true for valid JSON with a title of maximum permitted length" in { + CrossrefScorable.keepRecord(CrossrefStringWithMaximumTitle) shouldBe true + } + + it should "return false for valid JSON with excessively long title" in { + CrossrefScorable.keepRecord(CrossrefStringWithExcessiveTitle) shouldBe false + } + + it should "return false for valid JSON with null title" in { + CrossrefScorable.keepRecord(CrossrefStringWithNullTitle) shouldBe false + } + + it should "return false for valid JSON with no title" in { + CrossrefScorable.keepRecord(CrossrefStringWithoutTitle) shouldBe false + } + + it should "return false for invalid JSON" in { + CrossrefScorable.keepRecord(CrossrefStringWithoutTitle) shouldBe false + } + + it should "handle content types" in { + CrossrefScorable.jsonToMapFeatures(CrossrefStringWrongType) should be (None) + CrossrefScorable.jsonToMapFeatures(CrossrefStringNoType) should be (None) + } +} diff --git a/scalding/src/test/scala/sandcrawler/DumpUnGrobidedJobTest.scala b/scalding/src/test/scala/sandcrawler/DumpUnGrobidedJobTest.scala new file mode 100644 index 0000000..8dda5c8 --- /dev/null +++ b/scalding/src/test/scala/sandcrawler/DumpUnGrobidedJobTest.scala @@ -0,0 +1,72 @@ +package sandcrawler + +import cascading.tuple.Fields +import cascading.tuple.Tuple +import com.twitter.scalding.JobTest +import com.twitter.scalding.Tsv +import com.twitter.scalding.TupleConversions +import com.twitter.scalding.TypedTsv +import org.apache.hadoop.hbase.io.ImmutableBytesWritable +import org.apache.hadoop.hbase.util.Bytes +import org.junit.runner.RunWith +import org.scalatest.FunSpec +import org.scalatest.junit.JUnitRunner +import org.slf4j.LoggerFactory +import parallelai.spyglass.hbase.HBaseConstants.SourceMode +import parallelai.spyglass.hbase.HBaseSource +import scala._ + +@RunWith(classOf[JUnitRunner]) +class DumpUnGrobidedJobTest extends FunSpec with TupleConversions { + + val output = "/tmp/testOutput" + val (testTable, testHost) = ("test-table", "dummy-host:2181") + + val log = LoggerFactory.getLogger(this.getClass.getName) + + val statusCode: Long = 200 + val statusBytes = Bytes.toBytes(statusCode) + + val sampleDataGrobid : List[List[Array[Byte]]] = List( + ("sha1:SDKUVHC3YNNEGH5WAG5ZAAXWAEBNX4WT", statusBytes), + ("sha1:35985C3YNNEGH5WAG5ZAAXWAEBNXJW56", statusBytes), + ("sha1:885C3YNNEGH5WAG5ZAAXWA8BNXJWT6CZ", statusBytes), + ("sha1:00904C3YNNEGH5WAG5ZA9XWAEBNXJWT6", statusBytes), + ("sha1:249C3YNNEGH5WAG5ZAAXWAEBNXJWT6CZ", statusBytes), + ("sha1:095893C3YNNEGH5WAG5ZAAXWAEBNXJWT", statusBytes)) + .map(pair => List(Bytes.toBytes(pair._1), pair._2)) + + val sampleDataFile : List[List[Array[Byte]]] = List( + ("sha1:K2DKSSVTXWPRMFDTWSTCQW3RVWRIOV3Q", """{c-json-data}""", "application/pdf", """{cdx-json-data}"""), + ("sha1:C3YNNEGH5WAG5ZAAXWAEBNXJWT6CZ3WU", """{c-json-data}""", "application/pdf", """{cdx-json-data}"""), + ("sha1:SDKUVHC3YNNEGH5WAG5ZAAXWAEBNX4WT", """{c-json-data}""", "application/pdf", """{cdx-json-data}"""), + ("sha1:35985C3YNNEGH5WAG5ZAAXWAEBNXJW56", """{c-json-data}""", "application/pdf", """{cdx-json-data}"""), + ("sha1:885C3YNNEGH5WAG5ZAAXWA8BNXJWT6CZ", """{c-json-data}""", "application/pdf", """{cdx-json-data}"""), + ("sha1:00904C3YNNEGH5WAG5ZA9XWAEBNXJWT6", """{c-json-data}""", "application/pdf", """{cdx-json-data}"""), + ("sha1:249C3YNNEGH5WAG5ZAAXWAEBNXJWT6CZ", """{c-json-data}""", "application/pdf", """{cdx-json-data}"""), + ("sha1:095893C3YNNEGH5WAG5ZAAXWAEBNXJWT", """{c-json-data}""", "application/pdf", """{cdx-json-data}""")) + .map(pair => List(Bytes.toBytes(pair._1), + Bytes.toBytes(pair._2), + Bytes.toBytes(pair._3), + Bytes.toBytes(pair._4))) + + JobTest("sandcrawler.DumpUnGrobidedJob") + .arg("test", "") + .arg("app.conf.path", "app.conf") + .arg("output", output) + .arg("hbase-table", testTable) + .arg("zookeeper-hosts", testHost) + .arg("debug", "true") + .source[Tuple](DumpUnGrobidedJob.getHBaseColSource(testTable, testHost), + sampleDataGrobid.map(l => new Tuple(l.map(s => {new ImmutableBytesWritable(s)}):_*))) + .source[Tuple](DumpUnGrobidedJob.getHBaseKeySource(testTable, testHost), + sampleDataFile.map(l => new Tuple(l.map(s => {new ImmutableBytesWritable(s)}):_*))) + .sink[Tuple](TypedTsv[(String,String,String,String)](output)) { + outputBuffer => + it("should return correct-length list.") { + assert(outputBuffer.size === 2) + } + } + .run + .finish +} diff --git a/scalding/src/test/scala/sandcrawler/FatcatScorableTest.scala b/scalding/src/test/scala/sandcrawler/FatcatScorableTest.scala new file mode 100644 index 0000000..823e14a --- /dev/null +++ b/scalding/src/test/scala/sandcrawler/FatcatScorableTest.scala @@ -0,0 +1,160 @@ +package sandcrawler + +import cascading.tuple.Fields +import cascading.tuple.Tuple +import com.twitter.scalding.JobTest +import com.twitter.scalding.TextLine +import com.twitter.scalding.TupleConversions +import com.twitter.scalding.TypedTsv +import org.apache.hadoop.hbase.io.ImmutableBytesWritable +import org.apache.hadoop.hbase.util.Bytes +import org.scalatest._ +import parallelai.spyglass.hbase.HBaseConstants.SourceMode + +class FatcatScorableTest extends FlatSpec with Matchers { + // scalastyle:off + val FatcatString = +""" +{ + "abstracts": [], + "refs": [], + "contribs": [ + { + "index": 0, + "raw_name": "W Gaier", + "surname": "Gaier", + "role": "author", + "extra": { + "seq": "first" + } + } + ], + "publisher": "Elsevier BV", + "pages": "186-187", + "ext_ids": { + "doi": "<<DOI>>" + }, + "release_year": 1996, + "release_stage": "published", + "release_type": "article-journal", + "container_id": "3nccslsn5jez3ixrp5skjyjxu4", + "title": "<<TITLE>>", + "state": "active", + "ident": "pnri57u66ffytigdmyybbmouni", + "work_id": "tdmqnfzm2nggrhfwzasyegvpyu", + "revision": "e50bd04e-d0d4-4ee7-b7a4-6b4f079de154", + "extra": { + "crossref": { + "alternative-id": [ + "0987-7983(96)87729-2" + ], + "type": "journal-article" + } + } +} +""".replace("<<DOI>>", "10.123/aBc") + // scalastyle:on + val FatcatStringWithGoodTitle = FatcatString.replace("<<TITLE>>", "Some Title") + val FatcatStringWithMaximumTitle = FatcatString.replace("<<TITLE>>", "T" * Scorable.MaxTitleLength) + val FatcatStringWithExcessiveTitle = FatcatString.replace("<<TITLE>>", "T" * Scorable.MaxTitleLength + "0") + val FatcatStringWithNullTitle = FatcatString.replace("\"<<TITLE>>\"", "null") + val FatcatStringWithEmptyTitle = FatcatString.replace("<<TITLE>>", "") + val FatcatStringWithoutTitle = FatcatString.replace("title", "nottitle") + val MalformedFatcatString = FatcatString.replace("}", "") + val FatcatStringWithNoAuthors = FatcatString.replace("<<TITLE>>", "Some Valid Title").replace("contribs", "no-contribs") + //val FatcatStringWrongType = FatcatString.replace("<<TITLE>>", "Some Valid Title").replace("journal-article", "other") + //val FatcatStringNoType = FatcatString.replace("<<TITLE>>", "Some Valid Title").replace("type", "not-type") + + // Unit tests + "FatcatScorable.jsonToMapFeatures()" should "handle invalid JSON" in { + FatcatScorable.jsonToMapFeatures(MalformedFatcatString) should be (None) + } + + it should "handle missing title" in { + FatcatScorable.jsonToMapFeatures(FatcatStringWithoutTitle) should be (None) + } + + it should "handle null title" in { + FatcatScorable.jsonToMapFeatures(FatcatStringWithNullTitle) should be (None) + } + + it should "handle empty title" in { + FatcatScorable.jsonToMapFeatures(FatcatStringWithEmptyTitle) should be (None) + } + + it should "handle subtitle" in { + FatcatScorable.jsonToMapFeatures( + """{"title": "short but not too short", "subtitle": "just right!", "ident": "pnri57u66ffytigdmyybbmouni", "work_id": "tdmqnfzm2nggrhfwzasyegvpyu", "DOI": "10.123/asdf", "type":"journal-article","contribs":[{ "raw_name" : "W Gaier", "surname" : "Gaier"}]}""") match { + case None => fail() + case Some(result) => result.slug shouldBe "shortbutnottooshortjustright" + } + } + + it should "handle empty subtitle" in { + FatcatScorable.jsonToMapFeatures( + """{"title": "short but not too short", "subtitle": "", "ident": "pnri57u66ffytigdmyybbmouni", "work_id": "tdmqnfzm2nggrhfwzasyegvpyu", "DOI": "10.123/asdf", "type":"journal-article", "contribs":[{ "raw_name" : "W Gaier", "surname" : "Gaier"}]}""") match { + case None => fail() + case Some(result) => result.slug shouldBe "shortbutnottooshort" + } + } + + it should "handle null subtitle" in { + FatcatScorable.jsonToMapFeatures( + """{"title": "short but not too short", "subtitle": null, "ident": "pnri57u66ffytigdmyybbmouni", "work_id": "tdmqnfzm2nggrhfwzasyegvpyu", "DOI": "10.123/asdf", "type":"journal-article", "contribs":[{ "raw_name" : "W Gaier", "surname" : "Gaier"}]}""") match { + case None => fail() + case Some(result) => result.slug shouldBe "shortbutnottooshort" + } + } + + it should "handle missing authors" in { + // TODO: not actually removing these + //FatcatScorable.jsonToMapFeatures(FatcatStringWithNoAuthors) should be (None) + FatcatScorable.jsonToMapFeatures(FatcatStringWithNoAuthors) + } + + it should "handle valid input" in { + FatcatScorable.jsonToMapFeatures(FatcatStringWithGoodTitle) match { + case None => fail() + case Some(result) => { + result.slug shouldBe "sometitle" + Scorable.jsonToMap(result.json) match { + case None => fail() + case Some(map) => { + map("title").asInstanceOf[String] shouldBe "Some Title" + //map("doi").asInstanceOf[String] shouldBe "10.123/abc" + map("fatcat_release").asInstanceOf[String] shouldBe "pnri57u66ffytigdmyybbmouni" + map("fatcat_work").asInstanceOf[String] shouldBe "tdmqnfzm2nggrhfwzasyegvpyu" + // TODO: full name? not just a string? + map("authors").asInstanceOf[List[String]] shouldBe List("W Gaier") + map("year").asInstanceOf[Double].toInt shouldBe 1996 + } + } + } + } + } + + "FatcatScorable.keepRecord()" should "return true for valid JSON with title" in { + FatcatScorable.keepRecord(FatcatStringWithGoodTitle) shouldBe true + } + + it should "return true for valid JSON with a title of maximum permitted length" in { + FatcatScorable.keepRecord(FatcatStringWithMaximumTitle) shouldBe true + } + + it should "return false for valid JSON with excessively long title" in { + FatcatScorable.keepRecord(FatcatStringWithExcessiveTitle) shouldBe false + } + + it should "return false for valid JSON with null title" in { + FatcatScorable.keepRecord(FatcatStringWithNullTitle) shouldBe false + } + + it should "return false for valid JSON with no title" in { + FatcatScorable.keepRecord(FatcatStringWithoutTitle) shouldBe false + } + + it should "return false for invalid JSON" in { + FatcatScorable.keepRecord(FatcatStringWithoutTitle) shouldBe false + } + +} diff --git a/scalding/src/test/scala/sandcrawler/GrobidScorableDumpJobTest.scala b/scalding/src/test/scala/sandcrawler/GrobidScorableDumpJobTest.scala new file mode 100644 index 0000000..bf9343b --- /dev/null +++ b/scalding/src/test/scala/sandcrawler/GrobidScorableDumpJobTest.scala @@ -0,0 +1,124 @@ + +package sandcrawler + +import cascading.tuple.Fields +import cascading.tuple.Tuple +import com.twitter.scalding.JobTest +import com.twitter.scalding.TextLine +import com.twitter.scalding.TupleConversions +import com.twitter.scalding.TypedTsv +import org.apache.hadoop.hbase.io.ImmutableBytesWritable +import org.apache.hadoop.hbase.util.Bytes +import org.scalatest._ +import parallelai.spyglass.hbase.HBaseConstants.SourceMode + +class GrobidScorableDumpJobTest extends FlatSpec with Matchers { + //scalastyle:off + val JsonString = """ +{ + "title": "<<TITLE>>", + "authors": [ + {"name": "Brewster Kahle"}, + {"name": "J Doe"} + ], + "journal": { + "name": "Dummy Example File. Journal of Fake News. pp. 1-2. ISSN 1234-5678", + "eissn": null, + "issn": null, + "issue": null, + "publisher": null, + "volume": null + }, + "date": "2000", + "doi": null, + "citations": [ + { "authors": [{"name": "A Seaperson"}], + "date": "2001", + "id": "b0", + "index": 0, + "issue": null, + "journal": "Letters in the Alphabet", + "publisher": null, + "title": "Everything is Wonderful", + "url": null, + "volume": "20"}, + { "authors": [], + "date": "2011-03-28", + "id": "b1", + "index": 1, + "issue": null, + "journal": "The Dictionary", + "publisher": null, + "title": "All about Facts", + "url": null, + "volume": "14"} + ], + "abstract": "Everything you ever wanted to know about nothing", + "body": "Introduction \nEverything starts somewhere, as somebody [1] once said. \n\n In Depth \n Meat \nYou know, for kids. \n Potatos \nQED.", + "acknowledgement": null, + "annex": null +} +""" + // scalastyle:on + val JsonStringWithTitle = JsonString.replace("<<TITLE>>", "Dummy Example File") + val JsonStringWithoutTitle = JsonString.replace("title", "nottitle") + val MalformedJsonString = JsonString.replace("}", "") + + // Pipeline tests + val output = "/tmp/testOutput" + val input = "/tmp/testInput" + val (testTable, testHost) = ("test-table", "dummy-host:2181") + + val Sha1Strings : List[String] = List( + "sha1:K2DKSSVTXWPRMFDTWSTCQW3RVWRIOV3Q", // good + "sha1:C3YNNEGH5WAG5ZAAXWAEBNXJWT6CZ3WU", // good + "sha1:SDKUVHC3YNNEGH5WAG5ZAAXWAEBNX4WT", // good + "sha1:35985C3YNNEGH5WAG5ZAAXWAEBNXJW56", // bad status + "sha1:93187A85273589347598473894839443", // malformed + "sha1:024937534094897039547e9824382943") // bad status + + val JsonStrings : List[String] = List( + JsonString.replace("<<TITLE>>", "Title 1: The Classic"), + JsonString.replace("<<TITLE>>", "Title 2: TNG"), + JsonString.replace("<<TITLE>>", "Title 3: The Sequel"), + // This will have bad status. + JsonString.replace("<<TITLE>>", "Title 1: The Classic"), + MalformedJsonString, + // This will have bad status. + JsonString.replace("<<TITLE>>", "Title 2: Not TNG") + ) + + // bnewbold: status codes aren't strings, they are uint64 + val Ok : Long = 200 + val Bad : Long = 400 + val StatusCodes = List(Ok, Ok, Ok, Bad, Ok, Bad) + + val SampleDataHead : List[Tuple] = (Sha1Strings, JsonStrings, StatusCodes) + .zipped + .toList + .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 + val SampleData = SampleDataHead :+ new Tuple( + new ImmutableBytesWritable(Bytes.toBytes("sha1:35985C3YNNEGH5WAG5ZAA88888888888")), null, null) + // scalastyle:on null + + JobTest("sandcrawler.GrobidScorableDumpJob") + .arg("test", "") + .arg("app.conf.path", "app.conf") + .arg("output", output) + .arg("hbase-table", testTable) + .arg("zookeeper-hosts", testHost) + .arg("debug", "true") + .source[Tuple](GrobidScorable.getHBaseSource(testTable, testHost), SampleData) + .sink[(String, String)](TypedTsv[(String, String)](output)) { + outputBuffer => + "The pipeline" should "return correct-length list" in { + outputBuffer should have length 3 + } + } + .run + .finish +} diff --git a/scalding/src/test/scala/sandcrawler/GrobidScorableTest.scala b/scalding/src/test/scala/sandcrawler/GrobidScorableTest.scala new file mode 100644 index 0000000..b395a64 --- /dev/null +++ b/scalding/src/test/scala/sandcrawler/GrobidScorableTest.scala @@ -0,0 +1,122 @@ +package sandcrawler + +import cascading.tuple.Fields +import cascading.tuple.Tuple +import com.twitter.scalding.JobTest +import com.twitter.scalding.TextLine +import com.twitter.scalding.TupleConversions +import com.twitter.scalding.TypedTsv +import org.apache.hadoop.hbase.io.ImmutableBytesWritable +import org.apache.hadoop.hbase.util.Bytes +import org.scalatest._ +import parallelai.spyglass.hbase.HBaseConstants.SourceMode + +class GrobidScorableTest extends FlatSpec with Matchers { + val GrobidString = """ +{ + "title": "<<TITLE>>", + "authors": [ + {"name": "Brewster Kahle"}, + {"name": "J Doe"} + ], + "journal": { + "name": "Dummy Example File. Journal of Fake News. pp. 1-2. ISSN 1234-5678", + "eissn": null, + "issn": null, + "issue": null, + "publisher": null, + "volume": null + }, + "date": "2000", + "doi": null, + "citations": [ + { "authors": [{"name": "A Seaperson"}], + "date": "2001", + "id": "b0", + "index": 0, + "issue": null, + "journal": "Letters in the Alphabet", + "publisher": null, + "title": "Everything is Wonderful", + "url": null, + "volume": "20"}, + { "authors": [], + "date": "2011-03-28", + "id": "b1", + "index": 1, + "issue": null, + "journal": "The Dictionary", + "publisher": null, + "title": "All about Facts", + "url": null, + "volume": "14"} + ], + "abstract": "Everything you ever wanted to know about nothing", + "body": "Introduction \nEverything starts somewhere, as somebody [1] once said. \n\n In Depth \n Meat \nYou know, for kids. \n Potatos \nQED.", + "acknowledgement": null, + "annex": null +} +""" + val GrobidStringWithGoodTitle = GrobidString.replace("<<TITLE>>", "Dummy Example File") + val GrobidStringWithMaximumTitle = GrobidString.replace("<<TITLE>>", "T" * Scorable.MaxTitleLength) + val GrobidStringWithExcessiveTitle = GrobidString.replace("<<TITLE>>", "T" * Scorable.MaxTitleLength + "0") + val GrobidStringWithNullTitle = GrobidString.replace("\"<<TITLE>>\"", "null") + val GrobidStringWithoutTitle = GrobidString.replace("title", "nottitle") + val MalformedGrobidString = GrobidString.replace("}", "") + val Key = "Dummy Key" + + // Unit tests + + "GrobidScorable.jsonToMapFeatures()" should "handle invalid JSON" in { + GrobidScorable.jsonToMapFeatures(Key, MalformedGrobidString) should be (None) + } + + it should "handle null title" in { + GrobidScorable.jsonToMapFeatures(Key, GrobidStringWithNullTitle) should be (None) + } + + it should "handle missing title" in { + GrobidScorable.jsonToMapFeatures(Key, GrobidStringWithoutTitle) should be (None) + } + + it should "handle valid input" in { + GrobidScorable.jsonToMapFeatures(Key, GrobidStringWithGoodTitle) match { + case None => fail() + case Some(result) => { + result.slug shouldBe "dummyexamplefile" + Scorable.jsonToMap(result.json) match { + case None => fail() + case Some(map) => { + map should contain key "title" + map("title").asInstanceOf[String] shouldBe "Dummy Example File" + map("authors").asInstanceOf[List[String]] shouldBe List("Brewster Kahle", "J Doe") + } + } + } + } + } + + "GrobidScorable.keepRecord()" should "return true for valid JSON with title" in { + GrobidScorable.keepRecord(GrobidStringWithGoodTitle) shouldBe true + } + + it should "return true for valid JSON with a title of maximum permitted length" in { + GrobidScorable.keepRecord(GrobidStringWithMaximumTitle) shouldBe true + } + + it should "return false for valid JSON with excessively long title" in { + GrobidScorable.keepRecord(GrobidStringWithExcessiveTitle) shouldBe false + } + + it should "return false for valid JSON with null title" in { + GrobidScorable.keepRecord(GrobidStringWithNullTitle) shouldBe false + } + + it should "return false for valid JSON with no title" in { + GrobidScorable.keepRecord(GrobidStringWithoutTitle) shouldBe false + } + + it should "return false for invalid JSON" in { + GrobidScorable.keepRecord(GrobidStringWithoutTitle) shouldBe false + } +} diff --git a/scalding/src/test/scala/sandcrawler/HBaseBuilderTest.scala b/scalding/src/test/scala/sandcrawler/HBaseBuilderTest.scala index 603a4c7..c61cb22 100644 --- a/scalding/src/test/scala/sandcrawler/HBaseBuilderTest.scala +++ b/scalding/src/test/scala/sandcrawler/HBaseBuilderTest.scala @@ -22,6 +22,7 @@ class HBaseBuilderTest extends FlatSpec with Matchers { fields should have length 0 } + //scalastyle:off no.whitespace.before.left.bracket it should "throw IllegalArgumentException on malformed input" in { a [IllegalArgumentException] should be thrownBy { HBaseBuilder.parseColSpecs(List("file_size")) diff --git a/scalding/src/test/scala/sandcrawler/HBaseMimeCountTest.scala b/scalding/src/test/scala/sandcrawler/HBaseMimeCountTest.scala index fde2290..d6d283f 100644 --- a/scalding/src/test/scala/sandcrawler/HBaseMimeCountTest.scala +++ b/scalding/src/test/scala/sandcrawler/HBaseMimeCountTest.scala @@ -1,15 +1,18 @@ package sandcrawler -import cascading.tuple.{Tuple, Fields} -import com.twitter.scalding.{JobTest, Tsv, TupleConversions} +import cascading.tuple.Fields +import cascading.tuple.Tuple +import com.twitter.scalding.JobTest +import com.twitter.scalding.Tsv +import com.twitter.scalding.TupleConversions import org.apache.hadoop.hbase.io.ImmutableBytesWritable import org.apache.hadoop.hbase.util.Bytes import org.junit.runner.RunWith import org.scalatest.FunSpec import org.scalatest.junit.JUnitRunner import org.slf4j.LoggerFactory -import parallelai.spyglass.hbase.HBaseSource import parallelai.spyglass.hbase.HBaseConstants.SourceMode +import parallelai.spyglass.hbase.HBaseSource import scala._ @RunWith(classOf[JUnitRunner]) diff --git a/scalding/src/test/scala/sandcrawler/HBaseRowCountTest.scala b/scalding/src/test/scala/sandcrawler/HBaseRowCountTest.scala index 3424a36..c4ca5aa 100644 --- a/scalding/src/test/scala/sandcrawler/HBaseRowCountTest.scala +++ b/scalding/src/test/scala/sandcrawler/HBaseRowCountTest.scala @@ -1,15 +1,18 @@ package sandcrawler -import cascading.tuple.{Tuple, Fields} -import com.twitter.scalding.{JobTest, Tsv, TupleConversions} +import cascading.tuple.Fields +import cascading.tuple.Tuple +import com.twitter.scalding.JobTest +import com.twitter.scalding.Tsv +import com.twitter.scalding.TupleConversions import org.apache.hadoop.hbase.io.ImmutableBytesWritable import org.apache.hadoop.hbase.util.Bytes import org.junit.runner.RunWith import org.scalatest.FunSpec import org.scalatest.junit.JUnitRunner import org.slf4j.LoggerFactory -import parallelai.spyglass.hbase.HBaseSource import parallelai.spyglass.hbase.HBaseConstants.SourceMode +import parallelai.spyglass.hbase.HBaseSource import scala._ /** @@ -47,12 +50,10 @@ class HBaseRowCountTest extends FunSpec with TupleConversions { outputBuffer => it("should return the test data provided.") { - println("outputBuffer.size => " + outputBuffer.size) assert(outputBuffer.size === 1) } it("should return the correct count") { - println("raw output => " + outputBuffer) assert(outputBuffer(0).getObject(0) === 8) } } diff --git a/scalding/src/test/scala/sandcrawler/HBaseStatusCodeCountTest.scala b/scalding/src/test/scala/sandcrawler/HBaseStatusCodeCountTest.scala new file mode 100644 index 0000000..d2cf9de --- /dev/null +++ b/scalding/src/test/scala/sandcrawler/HBaseStatusCodeCountTest.scala @@ -0,0 +1,71 @@ +package sandcrawler + +import cascading.tuple.Fields +import cascading.tuple.Tuple +import com.twitter.scalding.JobTest +import com.twitter.scalding.Tsv +import com.twitter.scalding.TupleConversions +import com.twitter.scalding.TypedTsv +import org.apache.hadoop.hbase.io.ImmutableBytesWritable +import org.apache.hadoop.hbase.util.Bytes +import org.junit.runner.RunWith +import org.scalatest.FunSpec +import org.scalatest.junit.JUnitRunner +import org.slf4j.LoggerFactory +import parallelai.spyglass.hbase.HBaseConstants.SourceMode +import parallelai.spyglass.hbase.HBaseSource +import scala._ + +@RunWith(classOf[JUnitRunner]) +class HBaseStatusCodeCountTest extends FunSpec with TupleConversions { + + val output = "/tmp/testOutput" + val (testTable, testHost) = ("test-table", "dummy-host:2181") + + val log = LoggerFactory.getLogger(this.getClass.getName) + + val statusType1 : Long = 200 + val statusType2 : Long = 404 + val statusType1Bytes = Bytes.toBytes(statusType1) + val statusType2Bytes = Bytes.toBytes(statusType2) + + // 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) + + JobTest("sandcrawler.HBaseStatusCodeCountJob") + .arg("test", "") + .arg("app.conf.path", "app.conf") + .arg("output", output) + .arg("hbase-table", testTable) + .arg("zookeeper-hosts", testHost) + .arg("debug", "true") + .source[Tuple](HBaseCountJob.getHBaseSource(testTable, testHost, "grobid0:status_code"), + sampleData.map(l => new Tuple(l.map(s => {new ImmutableBytesWritable(s)}):_*))) + .sink[Tuple](TypedTsv[(Long, Long)](output)) { + outputBuffer => + it("should return a correct number of elements.") { + assert(outputBuffer.size === 2) + } + + // Convert List[Tuple] to Map[Long, Long]. + val counts = outputBuffer.map(t => (t.getLong(0), t.getLong(1))).toMap + it("should have the appropriate number of each status type") { + assert(counts(statusType1) == statusType1Count) + assert(counts(statusType2) == statusType2Count) + } + } + .run + .finish +} diff --git a/scalding/src/test/scala/sandcrawler/HBaseStatusCountTest.scala b/scalding/src/test/scala/sandcrawler/HBaseStatusCountTest.scala index d7689cd..7e91af3 100644 --- a/scalding/src/test/scala/sandcrawler/HBaseStatusCountTest.scala +++ b/scalding/src/test/scala/sandcrawler/HBaseStatusCountTest.scala @@ -1,15 +1,19 @@ package sandcrawler -import cascading.tuple.{Tuple, Fields} -import com.twitter.scalding.{JobTest, Tsv, TypedTsv, TupleConversions} +import cascading.tuple.Fields +import cascading.tuple.Tuple +import com.twitter.scalding.JobTest +import com.twitter.scalding.Tsv +import com.twitter.scalding.TupleConversions +import com.twitter.scalding.TypedTsv import org.apache.hadoop.hbase.io.ImmutableBytesWritable import org.apache.hadoop.hbase.util.Bytes import org.junit.runner.RunWith import org.scalatest.FunSpec import org.scalatest.junit.JUnitRunner import org.slf4j.LoggerFactory -import parallelai.spyglass.hbase.HBaseSource import parallelai.spyglass.hbase.HBaseConstants.SourceMode +import parallelai.spyglass.hbase.HBaseSource import scala._ @RunWith(classOf[JUnitRunner]) @@ -20,21 +24,20 @@ class HBaseStatusCountTest extends FunSpec with TupleConversions { val log = LoggerFactory.getLogger(this.getClass.getName) - val statusType1 : Long = 200 - val statusType2 : Long = 404 - val statusType1Bytes = Bytes.toBytes(statusType1) - val statusType2Bytes = Bytes.toBytes(statusType2) - - val sampleData = 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) - ) + val statusType1Bytes = Bytes.toBytes("""{"status": "success"}""") + val statusType2Bytes = Bytes.toBytes("""{"status": "partial"}""") + + // 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) @@ -46,20 +49,13 @@ class HBaseStatusCountTest extends FunSpec with TupleConversions { .arg("hbase-table", testTable) .arg("zookeeper-hosts", testHost) .arg("debug", "true") - .source[Tuple](HBaseCountJob.getHBaseSource(testTable, testHost, "grobid0:status_code"), + .source[Tuple](HBaseCountJob.getHBaseSource(testTable, testHost, "grobid0:status"), sampleData.map(l => new Tuple(l.map(s => {new ImmutableBytesWritable(s)}):_*))) - .sink[Tuple](TypedTsv[(Long, Long)](output)) { + .sink[Tuple](TypedTsv[(String, Long)](output)) { outputBuffer => it("should return a 2-element list.") { assert(outputBuffer.size === 2) } - - // Convert List[Tuple] to Map[Long, Long]. - val counts = outputBuffer.map(t => (t.getLong(0), t.getLong(1))).toMap - it("should have the appropriate number of each status type") { - assert(counts(statusType1) == statusType1Count) - assert(counts(statusType2) == statusType2Count) - } } .run .finish diff --git a/scalding/src/test/scala/sandcrawler/ScorableFeaturesTest.scala b/scalding/src/test/scala/sandcrawler/ScorableFeaturesTest.scala new file mode 100644 index 0000000..c847296 --- /dev/null +++ b/scalding/src/test/scala/sandcrawler/ScorableFeaturesTest.scala @@ -0,0 +1,64 @@ +package sandcrawler + +import java.io.InputStream + +import scala.io.Source + +import org.scalatest._ + +// scalastyle:off null +class ScorableFeaturesTest extends FlatSpec with Matchers { + "toMapFeatures()" should "work with gnarly inputs" in { + ScorableFeatures.create(title = null).toMapFeatures + ScorableFeatures.create(title = "something", doi = null, sha1 = null, year = 123).toMapFeatures + } + + private def titleToSlug(s : String) : Option[String] = ScorableFeatures.create(title = s).toSlug + + "mapToSlug()" should "extract the parts of titles before a colon" in { + titleToSlug("HELLO:there") shouldBe Some("hellothere") + } + + it should "extract an entire colon-less string" in { + titleToSlug("hello THERE") shouldBe Some("hellothere") + } + + it should "return Scorable.NoSlug if given empty string" in { + titleToSlug("") shouldBe (None) + } + + it should "return Scorable.NoSlug if given null" in { + titleToSlug(null) shouldBe (None) + } + + it should "strip punctuation" in { + titleToSlug("HELLO!:the:re") shouldBe Some("hellothere") + titleToSlug("a:b:cdefgh") shouldBe Some("abcdefgh") + titleToSlug( + "If you're happy and you know it, clap your hands!") shouldBe Some("ifyourehappyandyouknowitclapyourhands") + titleToSlug(":;\"\'") shouldBe (None) + } + + it should "filter stub titles" in { + titleToSlug("abstract") shouldBe (None) + titleToSlug("title!") shouldBe (None) + titleToSlug("a real title which is not on denylist") shouldBe Some("arealtitlewhichisnotondenylist") + } + + it should "strip special characters" in { + titleToSlug(":;!',|\"\'`.#?!-@*/\\=+~%$^{}()[]<>-_’·“”‘’“”«»「」¿–±§ʿ") shouldBe (None) + // TODO: titleToSlug("©™₨№…") shouldBe (None) + // TODO: titleToSlug("πµΣσ") shouldBe (None) + } + + it should "remove whitespace" in { + titleToSlug("foo bar : baz ::") shouldBe Some("foobarbaz") + titleToSlug("\na\t:b:cdefghi") shouldBe Some("abcdefghi") + titleToSlug("\n \t \r ") shouldBe (None) + } + + it should "skip very short slugs" in { + titleToSlug("short") shouldBe (None) + titleToSlug("a longer, more in depth title") shouldBe Some("alongermoreindepthtitle") + } +} diff --git a/scalding/src/test/scala/sandcrawler/ScorableTest.scala b/scalding/src/test/scala/sandcrawler/ScorableTest.scala new file mode 100644 index 0000000..2094543 --- /dev/null +++ b/scalding/src/test/scala/sandcrawler/ScorableTest.scala @@ -0,0 +1,81 @@ +package sandcrawler + +import cascading.tuple.Fields +import cascading.tuple.Tuple +import com.twitter.scalding.JobTest +import com.twitter.scalding.TextLine +import com.twitter.scalding.TupleConversions +import com.twitter.scalding.TypedTsv +import org.apache.hadoop.hbase.io.ImmutableBytesWritable +import org.apache.hadoop.hbase.util.Bytes +import org.scalatest._ +import parallelai.spyglass.hbase.HBaseConstants.SourceMode + +class ScorableTest extends FlatSpec with Matchers { + val JsonString = """ +{ + "title": "<<TITLE>>", + "authors": [ + {"name": "Brewster Kahle"}, + {"name": "J Doe"} + ], + "journal": { + "name": "Dummy Example File. Journal of Fake News. pp. 1-2. ISSN 1234-5678", + "eissn": null, + "issn": null, + "issue": null, + "publisher": null, + "volume": null + }, + "date": "2000", + "doi": null, + "citations": [ + { "authors": [{"name": "A Seaperson"}], + "date": "2001", + "id": "b0", + "index": 0, + "issue": null, + "journal": "Letters in the Alphabet", + "publisher": null, + "title": "Everything is Wonderful", + "url": null, + "volume": "20"}, + { "authors": [], + "date": "2011-03-28", + "id": "b1", + "index": 1, + "issue": null, + "journal": "The Dictionary", + "publisher": null, + "title": "All about Facts", + "url": null, + "volume": "14"} + ], + "abstract": "Everything you ever wanted to know about nothing", + "body": "Introduction \nEverything starts somewhere, as somebody [1] once said. \n\n In Depth \n Meat \nYou know, for kids. \n Potatos \nQED.", + "acknowledgement": null, + "annex": null +} +""" + "jsonToMap()" should "return a map, given a legal JSON string" in { + Scorable.jsonToMap(JsonString) should not be (None) + } + + it should "return None, given illegal JSON" in { + Scorable.jsonToMap("illegal{,json{{") should be (None) + } + + "computeOutput()" should "return Scorable.MaxScore if given identical ReduceFeatures" in { + val score = Scorable.computeSimilarity( + new ReduceFeatures(JsonString), new ReduceFeatures(JsonString)) + score shouldBe Scorable.MaxScore + } + + "computeOutput()" should "be case-insensitive" in { + val left = JsonString.replace("<<TITLE>>", "A TITLE UPPER CASE") + val right = JsonString.replace("<<TITLE>>", "a title upper case") + val score = Scorable.computeSimilarity( + new ReduceFeatures(left), new ReduceFeatures(right)) + score shouldBe Scorable.MaxScore + } +} diff --git a/scalding/src/test/scala/sandcrawler/ScoreInsertableJobTest.scala b/scalding/src/test/scala/sandcrawler/ScoreInsertableJobTest.scala new file mode 100644 index 0000000..5393f10 --- /dev/null +++ b/scalding/src/test/scala/sandcrawler/ScoreInsertableJobTest.scala @@ -0,0 +1,262 @@ +package sandcrawler + +import cascading.tuple.Fields +import cascading.tuple.Tuple +import com.twitter.scalding.JobTest +import com.twitter.scalding.TextLine +import com.twitter.scalding.TupleConversions +import com.twitter.scalding.TypedTsv +import org.apache.hadoop.hbase.io.ImmutableBytesWritable +import org.apache.hadoop.hbase.util.Bytes +import org.scalatest._ +import parallelai.spyglass.hbase.HBaseConstants.SourceMode + +class ScoreInsertableJobTest extends FlatSpec with Matchers { + //scalastyle:off + val JsonString = """ +{ + "title": "<<TITLE>>", + "authors": [ + {"name": "Brewster Kahle"}, + {"name": "J Doe"} + ], + "journal": { + "name": "Dummy Example File. Journal of Fake News. pp. 1-2. ISSN 1234-5678", + "eissn": null, + "issn": null, + "issue": null, + "publisher": null, + "volume": null + }, + "date": "2000", + "doi": null, + "citations": [ + { "authors": [{"name": "A Seaperson"}], + "date": "2001", + "id": "b0", + "index": 0, + "issue": null, + "journal": "Letters in the Alphabet", + "publisher": null, + "title": "Everything is Wonderful", + "url": null, + "volume": "20"}, + { "authors": [], + "date": "2011-03-28", + "id": "b1", + "index": 1, + "issue": null, + "journal": "The Dictionary", + "publisher": null, + "title": "All about Facts", + "url": null, + "volume": "14"} + ], + "abstract": "Everything you ever wanted to know about nothing", + "body": "Introduction \nEverything starts somewhere, as somebody [1] once said. \n\n In Depth \n Meat \nYou know, for kids. \n Potatos \nQED.", + "acknowledgement": null, + "annex": null +} +""" + // scalastyle:on + val JsonStringWithTitle = JsonString.replace("<<TITLE>>", "Dummy Example File") + val JsonStringWithoutTitle = JsonString.replace("title", "nottitle") + val MalformedJsonString = JsonString.replace("}", "") + + // scalastyle:off + val CrossrefString = +""" +{ "_id" : { "$oid" : "5a553d5988a035a45bf50ed3" }, + "indexed" : { "date-parts" : [ [ 2017, 10, 23 ] ], + "date-time" : "2017-10-23T17:19:16Z", + "timestamp" : { "$numberLong" : "1508779156477" } }, + "reference-count" : 0, + "publisher" : "Elsevier BV", + "issue" : "3", + "license" : [ { "URL" : "http://www.elsevier.com/tdm/userlicense/1.0/", + "start" : { "date-parts" : [ [ 1996, 1, 1 ] ], + "date-time" : "1996-01-01T00:00:00Z", + "timestamp" : { "$numberLong" : "820454400000" } }, + "delay-in-days" : 0, "content-version" : "tdm" }], + "content-domain" : { "domain" : [], "crossmark-restriction" : false }, + "published-print" : { "date-parts" : [ [ 1996 ] ] }, + "DOI" : "<<DOI>>", + "type" : "journal-article", + "created" : { "date-parts" : [ [ 2002, 7, 25 ] ], + "date-time" : "2002-07-25T15:09:41Z", + "timestamp" : { "$numberLong" : "1027609781000" } }, + "page" : "186-187", + "source" : "Crossref", + "is-referenced-by-count" : 0, + "title" : [ "<<TITLE>>" ], + "prefix" : "10.1016", + "volume" : "9", + "author" : [ { "given" : "W", "family" : "Gaier", "affiliation" : [] } ], + "member" : "78", + "container-title" : [ "Journal de Pédiatrie et de Puériculture" ], + "link" : [ { "URL" : "http://api.elsevier.com/content/article/PII:0987-7983(96)87729-2?httpAccept=text/xml", + "content-type" : "text/xml", + "content-version" : "vor", + "intended-application" : "text-mining" }, + { "URL" : + "http://api.elsevier.com/content/article/PII:0987-7983(96)87729-2?httpAccept=text/plain", + "content-type" : "text/plain", + "content-version" : "vor", + "intended-application" : "text-mining" } ], + "deposited" : { "date-parts" : [ [ 2015, 9, 3 ] ], + "date-time" : "2015-09-03T10:03:43Z", + "timestamp" : { "$numberLong" : "1441274623000" } }, + "score" : 1, + "issued" : { "date-parts" : [ [ 1996 ] ] }, + "references-count" : 0, + "alternative-id" : [ "0987-7983(96)87729-2" ], + "URL" : "http://dx.doi.org/10.1016/0987-7983(96)87729-2", + "ISSN" : [ "0987-7983" ], + "issn-type" : [ { "value" : "0987-7983", "type" : "print" } ], + "subject" : [ "Pediatrics, Perinatology, and Child Health" ] +} +""" + // scalastyle:on + val TooLongOfTitle = "X" * Scorable.MaxTitleLength + "Y" // arbitrary long string + val TooShortOfTitle = "X" * (ScorableFeatures.MinSlugLength - 1) + val CrossrefStringWithTitle = CrossrefString.replace("<<TITLE>>", "SomeTitle") + val CrossrefStringWithoutTitle = CrossrefString.replace("title", "nottitle") + val MalformedCrossrefString = CrossrefString.replace("}", "") + val CrossrefStrings = List( + CrossrefString.replace("<<TITLE>>", "Title 2: TNG").replace("<<DOI>>", "DOI-0"), + CrossrefString.replace("<<TITLE>>", "Title 1: TNG 2A").replace("<<DOI>>", "DOI-0.5"), + CrossrefString.replace("<<TITLE>>", "Title 1: TNG 3").replace("<<DOI>>", "DOI-0.75"), + CrossrefString.replace("<<TITLE>>", "Title 2: Rebooted").replace("<<DOI>>", "DOI-1"), + CrossrefString.replace("<<TITLE>>", TooLongOfTitle).replace("<<DOI>>", "DOI-1"), + CrossrefString.replace("<<TITLE>>", TooShortOfTitle).replace("<<DOI>>", "DOI-1")) + + // Pipeline tests + val output = "/tmp/testOutput" + val input = "/tmp/testInput" + val (testTable, testHost) = ("test-table", "dummy-host:2181") + + val Sha1Strings : List[String] = List( + "sha1:K2DKSSVTXWPRMFDTWSTCQW3RVWRIOV3Q", + "sha1:C3YNNEGH5WAG5ZAAXWAEBNXJWT6CZ3WU", + "sha1:SDKUVHC3YNNEGH5WAG5ZAAXWAEBNX4WT", + "sha1:35985C3YNNEGH5WAG5ZAAXWAEBNXJW56", + "sha1:93187A85273589347598473894839443", + "sha1:024937534094897039547e9824382943", + "sha1:93229759932857982837892347893892", + "sha1:83229759932857982837892347893892") + + val JsonStrings : List[String] = List( + JsonString.replace("<<TITLE>>", "Title 1: The Original"), + JsonString.replace("<<TITLE>>", "Title 2: TNG"), + JsonString.replace("<<TITLE>>", "Title 3: The Sequel"), + // This will have bad status. + JsonString.replace("<<TITLE>>", "Title 1: The Original"), + MalformedJsonString, + // This will have bad status. + JsonString.replace("<<TITLE>>", "Title 2: Not TNG"), + // These are in both sources but have bad titles + JsonString.replace("<<TITLE>>", TooLongOfTitle), + JsonString.replace("<<TITLE>>", TooShortOfTitle) + ) + + // bnewbold: status codes aren't strings, they are uint64 + val Ok : Long = 200 + val Bad : Long = 400 + val StatusCodes = List(Ok, Ok, Ok, Bad, Ok, Bad, Ok, Ok) + + val SampleDataHead : List[Tuple] = (Sha1Strings, JsonStrings, StatusCodes) + .zipped + .toList + .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 + + val CdxList: List[String] = List("{}", "{}", "{}", "{}", "{}", "{}", "{}", "{}" ) + val MimeList: List[String] = List("application/pdf", "application/pdf", "application/pdf", + "application/pdf", "application/pdf", "application/pdf", "application/pdf", + "application/pdf") + val SizeList: List[Long] = List(1,2,3,4,5,6,7,8) + + // Can zip 3 lists, but not 4... so we recursively zip + val SampleCdxData : List[Tuple] = ((Sha1Strings, CdxList).zipped.toList, (MimeList, SizeList).zipped.toList) + .zipped + .toList + .map { case ((sha: String, cdx: String), (mime: String, size: Long)) => List(Bytes.toBytes(sha), Bytes.toBytes(cdx), Bytes.toBytes(mime), Bytes.toBytes(size)) } + .map { l => new Tuple(l.map(s => {new ImmutableBytesWritable(s)}):_*) } + + JobTest("sandcrawler.ScoreInsertableJob") + .arg("test", "") + .arg("app.conf.path", "app.conf") + .arg("output", output) + .arg("hbase-table", testTable) + .arg("zookeeper-hosts", testHost) + .arg("crossref-input", input) + .arg("debug", "true") + .source[Tuple](GrobidScorable.getHBaseSource(testTable, testHost), SampleData) + .source[Tuple](ScoreInsertableJob.getHBaseCdxSource(testTable, testHost), SampleCdxData) + .source(TextLine(input), List( + 0 -> CrossrefStrings(0), + 1 -> CrossrefStrings(1), + 2 -> CrossrefStrings(2), + 3 -> CrossrefStrings(3), + 4 -> CrossrefStrings(4), + 4 -> CrossrefStrings(5))) + .sink[(String, ReduceFeatures)](TypedTsv[(String, ReduceFeatures)](output + ".trapped")) { _ => () } + .sink[(String, String, Int, String, String, String, String, Long)](TypedTsv[(String, String, Int, String, String, String, String, Long)](output)) { + // Grobid titles and slugs (in parentheses): + // Title 1 (title1) + // Title 2: TNG (title2tng) + // Title 3: The Sequel (title3thesequel) + // <too long of a title> + // <too short of a title> + // crossref titles and slugs (in parentheses): + // Title 2: TNG (title2tng) + // Title 1: TNG 2A (title1tng2a) + // Title 1: TNG 3 (title1tng3) + // Title 2: Rebooted (title2rebooted) + // <too long of a title> + // <too short of a title> + // XXX: Join should have 3 "title1" slugs and 1 "title2tng" slug + outputBuffer => + "The pipeline" should "return a 1-element list" in { + outputBuffer should have length 1 + } + + it should "has right # of entries with each slug" in { + val slugs = outputBuffer.map(_._2) + val countMap : Map[String, Int] = slugs.groupBy(identity).mapValues(_.size) + // XXX: countMap("title1") shouldBe 3 + countMap("title2tng") shouldBe 1 + } + + def bundle(slug : String, grobidIndex : Int, crossrefIndex : Int) : (String, Int, String, String) = { + val mfg : Option[MapFeatures] = GrobidScorable.jsonToMapFeatures( + Sha1Strings(grobidIndex), + JsonStrings(grobidIndex)) + val mfc : Option[MapFeatures] = CrossrefScorable.jsonToMapFeatures(CrossrefStrings(crossrefIndex)) + if (mfg.isEmpty || mfc.isEmpty) { + fail() + } else { + val score = Scorable.computeSimilarity( + ReduceFeatures(mfg.get.json), + ReduceFeatures(mfc.get.json)) + (slug, score, mfg.get.json, mfc.get.json) + } + } + + it should "have right output values" in { + //outputBuffer.exists(_ == bundle("title1", 0, 0)) + //outputBuffer.exists(_ == bundle("title1", 0, 2)) + //outputBuffer.exists(_ == bundle("title1", 0, 1)) + outputBuffer.exists(_ == bundle("title2tng", 1, 3)) + } + } + .run + .finish +} diff --git a/scalding/src/test/scala/sandcrawler/ScoreJobTest.scala b/scalding/src/test/scala/sandcrawler/ScoreJobTest.scala new file mode 100644 index 0000000..fbc0ee5 --- /dev/null +++ b/scalding/src/test/scala/sandcrawler/ScoreJobTest.scala @@ -0,0 +1,248 @@ +package sandcrawler + +import cascading.tuple.Fields +import cascading.tuple.Tuple +import com.twitter.scalding.JobTest +import com.twitter.scalding.TextLine +import com.twitter.scalding.TupleConversions +import com.twitter.scalding.TypedTsv +import org.apache.hadoop.hbase.io.ImmutableBytesWritable +import org.apache.hadoop.hbase.util.Bytes +import org.scalatest._ +import parallelai.spyglass.hbase.HBaseConstants.SourceMode + +class ScoreJobTest extends FlatSpec with Matchers { + //scalastyle:off + val JsonString = """ +{ + "title": "<<TITLE>>", + "authors": [ + {"name": "Brewster Kahle"}, + {"name": "J Doe"} + ], + "journal": { + "name": "Dummy Example File. Journal of Fake News. pp. 1-2. ISSN 1234-5678", + "eissn": null, + "issn": null, + "issue": null, + "publisher": null, + "volume": null + }, + "date": "2000", + "doi": null, + "citations": [ + { "authors": [{"name": "A Seaperson"}], + "date": "2001", + "id": "b0", + "index": 0, + "issue": null, + "journal": "Letters in the Alphabet", + "publisher": null, + "title": "Everything is Wonderful", + "url": null, + "volume": "20"}, + { "authors": [], + "date": "2011-03-28", + "id": "b1", + "index": 1, + "issue": null, + "journal": "The Dictionary", + "publisher": null, + "title": "All about Facts", + "url": null, + "volume": "14"} + ], + "abstract": "Everything you ever wanted to know about nothing", + "body": "Introduction \nEverything starts somewhere, as somebody [1] once said. \n\n In Depth \n Meat \nYou know, for kids. \n Potatos \nQED.", + "acknowledgement": null, + "annex": null +} +""" + // scalastyle:on + val JsonStringWithTitle = JsonString.replace("<<TITLE>>", "Dummy Example File") + val JsonStringWithoutTitle = JsonString.replace("title", "nottitle") + val MalformedJsonString = JsonString.replace("}", "") + + // scalastyle:off + val CrossrefString = +""" +{ "_id" : { "$oid" : "5a553d5988a035a45bf50ed3" }, + "indexed" : { "date-parts" : [ [ 2017, 10, 23 ] ], + "date-time" : "2017-10-23T17:19:16Z", + "timestamp" : { "$numberLong" : "1508779156477" } }, + "reference-count" : 0, + "publisher" : "Elsevier BV", + "issue" : "3", + "license" : [ { "URL" : "http://www.elsevier.com/tdm/userlicense/1.0/", + "start" : { "date-parts" : [ [ 1996, 1, 1 ] ], + "date-time" : "1996-01-01T00:00:00Z", + "timestamp" : { "$numberLong" : "820454400000" } }, + "delay-in-days" : 0, "content-version" : "tdm" }], + "content-domain" : { "domain" : [], "crossmark-restriction" : false }, + "published-print" : { "date-parts" : [ [ 1996 ] ] }, + "DOI" : "<<DOI>>", + "type" : "journal-article", + "created" : { "date-parts" : [ [ 2002, 7, 25 ] ], + "date-time" : "2002-07-25T15:09:41Z", + "timestamp" : { "$numberLong" : "1027609781000" } }, + "page" : "186-187", + "source" : "Crossref", + "is-referenced-by-count" : 0, + "title" : [ "<<TITLE>>" ], + "prefix" : "10.1016", + "volume" : "9", + "author" : [ { "given" : "W", "family" : "Gaier", "affiliation" : [] } ], + "member" : "78", + "container-title" : [ "Journal de Pédiatrie et de Puériculture" ], + "link" : [ { "URL" : "http://api.elsevier.com/content/article/PII:0987-7983(96)87729-2?httpAccept=text/xml", + "content-type" : "text/xml", + "content-version" : "vor", + "intended-application" : "text-mining" }, + { "URL" : + "http://api.elsevier.com/content/article/PII:0987-7983(96)87729-2?httpAccept=text/plain", + "content-type" : "text/plain", + "content-version" : "vor", + "intended-application" : "text-mining" } ], + "deposited" : { "date-parts" : [ [ 2015, 9, 3 ] ], + "date-time" : "2015-09-03T10:03:43Z", + "timestamp" : { "$numberLong" : "1441274623000" } }, + "score" : 1, + "issued" : { "date-parts" : [ [ 1996 ] ] }, + "references-count" : 0, + "alternative-id" : [ "0987-7983(96)87729-2" ], + "URL" : "http://dx.doi.org/10.1016/0987-7983(96)87729-2", + "ISSN" : [ "0987-7983" ], + "issn-type" : [ { "value" : "0987-7983", "type" : "print" } ], + "subject" : [ "Pediatrics, Perinatology, and Child Health" ] +} +""" + // scalastyle:on + val TooLongOfTitle = "X" * Scorable.MaxTitleLength + "Y" // arbitrary long string + val TooShortOfTitle = "X" * (ScorableFeatures.MinSlugLength - 1) + val CrossrefStringWithTitle = CrossrefString.replace("<<TITLE>>", "SomeTitle") + val CrossrefStringWithoutTitle = CrossrefString.replace("title", "nottitle") + val MalformedCrossrefString = CrossrefString.replace("}", "") + val CrossrefStrings = List( + CrossrefString.replace("<<TITLE>>", "Title 2: TNG").replace("<<DOI>>", "DOI-0"), + CrossrefString.replace("<<TITLE>>", "Title 1: TNG 2A").replace("<<DOI>>", "DOI-0.5"), + CrossrefString.replace("<<TITLE>>", "Title 1: TNG 3").replace("<<DOI>>", "DOI-0.75"), + CrossrefString.replace("<<TITLE>>", "Title 2: Rebooted").replace("<<DOI>>", "DOI-1"), + CrossrefString.replace("<<TITLE>>", TooLongOfTitle).replace("<<DOI>>", "DOI-1"), + CrossrefString.replace("<<TITLE>>", TooShortOfTitle).replace("<<DOI>>", "DOI-1")) + + // Pipeline tests + val output = "/tmp/testOutput" + val input = "/tmp/testInput" + val (testTable, testHost) = ("test-table", "dummy-host:2181") + + val Sha1Strings : List[String] = List( + "sha1:K2DKSSVTXWPRMFDTWSTCQW3RVWRIOV3Q", + "sha1:C3YNNEGH5WAG5ZAAXWAEBNXJWT6CZ3WU", + "sha1:SDKUVHC3YNNEGH5WAG5ZAAXWAEBNX4WT", + "sha1:35985C3YNNEGH5WAG5ZAAXWAEBNXJW56", + "sha1:93187A85273589347598473894839443", + "sha1:024937534094897039547e9824382943", + "sha1:93229759932857982837892347893892", + "sha1:83229759932857982837892347893892") + + val JsonStrings : List[String] = List( + JsonString.replace("<<TITLE>>", "Title 1: The Original"), + JsonString.replace("<<TITLE>>", "Title 2: TNG"), + JsonString.replace("<<TITLE>>", "Title 3: The Sequel"), + // This will have bad status. + JsonString.replace("<<TITLE>>", "Title 1: The Original"), + MalformedJsonString, + // This will have bad status. + JsonString.replace("<<TITLE>>", "Title 2: Not TNG"), + // These are in both sources but have bad titles + JsonString.replace("<<TITLE>>", TooLongOfTitle), + JsonString.replace("<<TITLE>>", TooShortOfTitle) + ) + + // bnewbold: status codes aren't strings, they are uint64 + val Ok : Long = 200 + val Bad : Long = 400 + val StatusCodes = List(Ok, Ok, Ok, Bad, Ok, Bad, Ok, Ok) + + val SampleDataHead : List[Tuple] = (Sha1Strings, JsonStrings, StatusCodes) + .zipped + .toList + .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", "") + .arg("app.conf.path", "app.conf") + .arg("output", output) + .arg("hbase-table", testTable) + .arg("zookeeper-hosts", testHost) + .arg("crossref-input", input) + .arg("debug", "true") + .source[Tuple](GrobidScorable.getHBaseSource(testTable, testHost), SampleData) + .source(TextLine(input), List( + 0 -> CrossrefStrings(0), + 1 -> CrossrefStrings(1), + 2 -> CrossrefStrings(2), + 3 -> CrossrefStrings(3), + 4 -> CrossrefStrings(4), + 4 -> CrossrefStrings(5))) + .sink[(String, ReduceFeatures)](TypedTsv[(String, ReduceFeatures)](output + ".trapped")) { _ => () } + .sink[(String, Int, String, String)](TypedTsv[(String, Int, String, String)](output)) { + // Grobid titles and slugs (in parentheses): + // Title 1 (title1) + // Title 2: TNG (title2tng) + // Title 3: The Sequel (title3thesequel) + // <too long of a title> + // <too short of a title> + // crossref titles and slugs (in parentheses): + // Title 2: TNG (title2tng) + // Title 1: TNG 2A (title1tng2a) + // Title 1: TNG 3 (title1tng3) + // Title 2: Rebooted (title2rebooted) + // <too long of a title> + // <too short of a title> + // XXX: Join should have 3 "title1" slugs and 1 "title2tng" slug + outputBuffer => + "The pipeline" should "return a 1-element list" in { + outputBuffer should have length 1 + } + + it should "has right # of entries with each slug" in { + val slugs = outputBuffer.map(_._1) + val countMap : Map[String, Int] = slugs.groupBy(identity).mapValues(_.size) + // XXX: countMap("title1") shouldBe 3 + countMap("title2tng") shouldBe 1 + } + + def bundle(slug : String, grobidIndex : Int, crossrefIndex : Int) : (String, Int, String, String) = { + val mfg : Option[MapFeatures] = GrobidScorable.jsonToMapFeatures( + Sha1Strings(grobidIndex), + JsonStrings(grobidIndex)) + val mfc : Option[MapFeatures] = CrossrefScorable.jsonToMapFeatures(CrossrefStrings(crossrefIndex)) + if (mfg.isEmpty || mfc.isEmpty) { + fail() + } else { + val score = Scorable.computeSimilarity( + ReduceFeatures(mfg.get.json), + ReduceFeatures(mfc.get.json)) + (slug, score, mfg.get.json, mfc.get.json) + } + } + + it should "have right output values" in { + //outputBuffer.exists(_ == bundle("title1", 0, 0)) + //outputBuffer.exists(_ == bundle("title1", 0, 2)) + //outputBuffer.exists(_ == bundle("title1", 0, 1)) + outputBuffer.exists(_ == bundle("title2tng", 1, 3)) + } + } + .run + .finish +} diff --git a/scalding/src/test/scala/sandcrawler/StringUtilitiesTest.scala b/scalding/src/test/scala/sandcrawler/StringUtilitiesTest.scala new file mode 100644 index 0000000..410819b --- /dev/null +++ b/scalding/src/test/scala/sandcrawler/StringUtilitiesTest.scala @@ -0,0 +1,85 @@ +package sandcrawler + +import org.scalatest._ + +class StringUtilitiesTest extends FlatSpec with Matchers { + "removeAccents()" should "handle the empty string" in { + StringUtilities.removeAccents("") shouldBe "" + } + + it should "not change a string with unaccented characters" in { + StringUtilities.removeAccents("abc123") shouldBe "abc123" + } + + it should "remove accents from Ls" in { + StringUtilities.removeAccents("E\u0141\u0142en") shouldBe "ELlen" + } + + it should "remove accents from Es without changing case" in { + val result = StringUtilities.removeAccents("\u00e9") + result should have length 1 + result shouldBe "e" + } + + it should "convert the ø in Soren" in { + StringUtilities.removeAccents("Søren") shouldBe "Soren" + StringUtilities.removeAccents("SØREN") shouldBe "SOREN" + } + + "removePunctuation" should "work on the empty string" in { + StringUtilities.removePunctuation("") shouldBe "" + } + + it should "work on non-empty text strings" in { + StringUtilities.removePunctuation("Hello, world!") shouldBe "Hello world" + StringUtilities.removePunctuation(":-)") shouldBe "" + StringUtilities.removePunctuation("<<---a---b--->") shouldBe "ab" + } + + // Tests adapted from https://oldfashionedsoftware.com/2009/11/19/string-distance-and-refactoring-in-scala/ + "stringDistance" should "work on empty strings" in { + StringUtilities.stringDistance("", "") shouldBe 0 + StringUtilities.stringDistance("a", "") shouldBe 1 + StringUtilities.stringDistance("", "a") shouldBe 1 + StringUtilities.stringDistance("abc", "") shouldBe 3 + StringUtilities.stringDistance("", "abc") shouldBe 3 + } + + it should "work on equal strings" in { + StringUtilities.stringDistance("", "") shouldBe 0 + StringUtilities.stringDistance("a", "a") shouldBe 0 + StringUtilities.stringDistance("abc", "abc") shouldBe 0 + } + + it should "work where only inserts are needed" in { + StringUtilities.stringDistance("", "a") shouldBe 1 + StringUtilities.stringDistance("a", "ab") shouldBe 1 + StringUtilities.stringDistance("b", "ab") shouldBe 1 + StringUtilities.stringDistance("ac", "abc") shouldBe 1 + StringUtilities.stringDistance("abcdefg", "xabxcdxxefxgx") shouldBe 6 + } + + it should "work where only deletes are needed" in { + StringUtilities.stringDistance( "a", "") shouldBe 1 + StringUtilities.stringDistance( "ab", "a") shouldBe 1 + StringUtilities.stringDistance( "ab", "b") shouldBe 1 + StringUtilities.stringDistance("abc", "ac") shouldBe 1 + StringUtilities.stringDistance("xabxcdxxefxgx", "abcdefg") shouldBe 6 + } + + it should "work where only substitutions are needed" in { + StringUtilities.stringDistance( "a", "b") shouldBe 1 + StringUtilities.stringDistance( "ab", "ac") shouldBe 1 + StringUtilities.stringDistance( "ac", "bc") shouldBe 1 + StringUtilities.stringDistance("abc", "axc") shouldBe 1 + StringUtilities.stringDistance("xabxcdxxefxgx", "1ab2cd34ef5g6") shouldBe 6 + } + + it should "work where many operations are needed" in { + StringUtilities.stringDistance("example", "samples") shouldBe 3 + StringUtilities.stringDistance("sturgeon", "urgently") shouldBe 6 + StringUtilities.stringDistance("levenshtein", "frankenstein") shouldBe 6 + StringUtilities.stringDistance("distance", "difference") shouldBe 5 + StringUtilities.stringDistance("java was neat", "scala is great") shouldBe 7 + } +} |