InterLanguageLinksExtractor directly extracts owl:same

sven-h · Jan 21, 2018 · f83fa01 · f83fa01
1 parent e6f396d
commit f83fa01
Show file tree

Hide file tree

Showing 2 changed files with 9 additions and 4 deletions.
diff --git a/core/src/main/scala/org/dbpedia/extraction/mappings/InterLanguageLinksExtractor.scala b/core/src/main/scala/org/dbpedia/extraction/mappings/InterLanguageLinksExtractor.scala
@@ -14,14 +14,14 @@ import scala.language.reflectiveCalls
  */
 class InterLanguageLinksExtractor(context: { def ontology : Ontology; def language : Language }) extends PageNodeExtractor
 {
-  private val interLanguageLinksProperty = context.ontology.properties("wikiPageInterLanguageLink")
+  private val sameAsProperty = context.ontology.properties("owl:sameAs")
 
   override val datasets = Set(DBpediaDatasets.InterLanguageLinks)
 
   private val namespaces = if (context.language == Language.Commons) ExtractorUtils.commonsNamespacesContainingMetadata
     else Set(Namespace.Main, Namespace.Template, Namespace.Category)
 
-  private val quad = QuadBuilder.apply(context.language, DBpediaDatasets.InterLanguageLinks, interLanguageLinksProperty, null) _
+  private val quad = QuadBuilder.apply(context.language, DBpediaDatasets.InterLanguageLinks, sameAsProperty, null) _
 
   override def extract(page : PageNode, subjectUri : String) : Seq[Quad] =
   {

diff --git a/scripts/src/main/scala/org/dbpedia/extraction/scripts/UnmodifiedFeederCacheGenerator.scala b/scripts/src/main/scala/org/dbpedia/extraction/scripts/UnmodifiedFeederCacheGenerator.scala
@@ -2,7 +2,7 @@ package org.dbpedia.extraction.scripts
 
 import java.io.File
 
-import org.apache.commons.lang.StringEscapeUtils
+import org.apache.commons.lang3.StringUtils
 import org.dbpedia.extraction.config.ConfigUtils.parseLanguages
 import org.dbpedia.extraction.util.RichFile.wrapFile
 import org.dbpedia.extraction.util.{DateFinder, IOUtils}
@@ -50,7 +50,7 @@ object UnmodifiedFeederCacheGenerator {
         new QuadMapper().readQuads(finder, "page-ids" + suffix, auto = true) {
           quad =>
             val pageID = quad.value.toInt
-            var pageTitle = StringEscapeUtils.escapeSql(quad.subject.split("dbpedia.org/resource/", 2)(1))
+            var pageTitle = escapeSql(quad.subject.split("dbpedia.org/resource/", 2)(1))
             if (pageTitle.contains(':')) {
               val splitPageTitle = pageTitle.split(":", 2)
               val namespace = Namespace.get(language, splitPageTitle(0).replace("_", " "))
@@ -69,4 +69,9 @@ object UnmodifiedFeederCacheGenerator {
 
     }
   }
+
+  def escapeSql(str: String): String = {
+    if (str == null) return null
+    StringUtils.replace(str, "'", "''")
+  }
 }