fix ID consideration in DuplicateCheck

JabRef · Apr 5, 2017 · a1f4101 · a1f4101
1 parent b714206
commit a1f4101
Show file tree

Hide file tree

Showing 2 changed files with 70 additions and 12 deletions.
diff --git a/src/main/java/org/jabref/logic/bibtex/DuplicateCheck.java b/src/main/java/org/jabref/logic/bibtex/DuplicateCheck.java
@@ -59,27 +59,21 @@ private DuplicateCheck() {}
     /**
      * Checks if the two entries represent the same publication.
      *
-     * Requirements:
-     * 1. Equal entry type
-     *
-     * Checks:
-     * 1. Equal identifier (DOI)
-     *
      * @param one BibEntry
      * @param two BibEntry
      * @return boolean
      */
     public static boolean isDuplicate(BibEntry one, BibEntry two, BibDatabaseMode bibDatabaseMode) {
-        // same entry type
-        if (!one.getType().equals(two.getType())) {
-            return false;
-        }
-
         // same identifier
         if (hasSameIdentifier(one, two)) {
             return true;
         }
 
+        // same entry type
+        if (!one.getType().equals(two.getType())) {
+            return false;
+        }
+
         EntryType type = EntryTypes.getTypeOrDefault(one.getType(), bibDatabaseMode);
         // The check if they have the same required fields:
         List<String> var = type.getRequiredFieldsFlat();
@@ -106,7 +100,7 @@ public static boolean isDuplicate(BibEntry one, BibEntry two, BibDatabaseMode bi
 
     private static boolean hasSameIdentifier(BibEntry one, BibEntry two) {
         for (String name : FieldName.getIdentifierFieldNames()) {
-            if (compareSingleField(name, one, two) == 1) {
+            if (one.getField(name).isPresent() && one.getField(name).equals(two.getField(name))) {
                 return true;
             }
         }

diff --git a/src/test/java/org/jabref/logic/bibtex/DuplicateCheckTest.java b/src/test/java/org/jabref/logic/bibtex/DuplicateCheckTest.java
@@ -2,8 +2,11 @@
 
 import org.jabref.model.database.BibDatabaseMode;
 import org.jabref.model.entry.BibEntry;
+import org.jabref.model.entry.BibtexEntryType;
 import org.jabref.model.entry.BibtexEntryTypes;
+import org.jabref.model.entry.FieldName;
 
+import org.junit.Before;
 import org.junit.Test;
 
 import static org.junit.Assert.assertEquals;
@@ -12,6 +15,21 @@
 
 public class DuplicateCheckTest {
 
+    private BibEntry simpleArticle;
+    private BibEntry unrelatedArticle;
+
+    @Before
+    public void setUp() {
+        simpleArticle = new BibEntry(BibtexEntryTypes.ARTICLE.getName())
+                .withField(FieldName.AUTHOR, "Single Author")
+                .withField(FieldName.TITLE, "A serious paper about something")
+                .withField(FieldName.YEAR, "2017");
+        unrelatedArticle = new BibEntry(BibtexEntryTypes.ARTICLE.getName())
+                .withField(FieldName.AUTHOR, "Completely Different")
+                .withField(FieldName.TITLE, "Holy Moly Uffdada und Trallalla")
+                .withField(FieldName.YEAR, "1992");
+    }
+
     @Test
     public void testDuplicateDetection() {
         BibEntry one = new BibEntry(BibtexEntryTypes.ARTICLE.getName());
@@ -79,4 +97,50 @@ public void testWordCorrelation() {
         assertEquals(0.78, (DuplicateCheck.correlateByWords(d2, d3)), 0.01);
     }
 
+    @Test
+    public void twoUnrelatedEntriesAreNoDuplicates() {
+        assertFalse(DuplicateCheck.isDuplicate(simpleArticle, unrelatedArticle, BibDatabaseMode.BIBTEX));
+    }
+
+    @Test
+    public void twoUnrelatedEntriesWithDifferentDoisAreNoDuplicates() {
+        simpleArticle.setField(FieldName.DOI, "10.1016/j.is.2004.02.002");
+        unrelatedArticle.setField(FieldName.DOI, "10.1016/j.is.2004.02.00X");
+
+        assertFalse(DuplicateCheck.isDuplicate(simpleArticle, unrelatedArticle, BibDatabaseMode.BIBTEX));
+    }
+
+    @Test
+    public void twoUnrelatedEntriesWithEqualDoisAreDuplicates() {
+        simpleArticle.setField(FieldName.DOI, "10.1016/j.is.2004.02.002");
+        unrelatedArticle.setField(FieldName.DOI, "10.1016/j.is.2004.02.002");
+
+        assertTrue(DuplicateCheck.isDuplicate(simpleArticle, unrelatedArticle, BibDatabaseMode.BIBTEX));
+    }
+
+    @Test
+    public void twoUnrelatedEntriesWithEqualPmidAreDuplicates() {
+        simpleArticle.setField(FieldName.PMID, "12345678");
+        unrelatedArticle.setField(FieldName.PMID, "12345678");
+
+        assertTrue(DuplicateCheck.isDuplicate(simpleArticle, unrelatedArticle, BibDatabaseMode.BIBTEX));
+    }
+
+    @Test
+    public void twoUnrelatedEntriesWithEqualEprintAreDuplicates() {
+        simpleArticle.setField(FieldName.EPRINT, "12345678");
+        unrelatedArticle.setField(FieldName.EPRINT, "12345678");
+
+        assertTrue(DuplicateCheck.isDuplicate(simpleArticle, unrelatedArticle, BibDatabaseMode.BIBTEX));
+    }
+
+    @Test
+    public void twoEntriesWithSameDoiButDifferentTypesAreDuplicates() {
+        simpleArticle.setField(FieldName.DOI, "10.1016/j.is.2004.02.002");
+        BibEntry duplicateWithDifferentType = (BibEntry) simpleArticle.clone();
+        duplicateWithDifferentType.setType(BibtexEntryTypes.INCOLLECTION);
+
+        assertTrue(DuplicateCheck.isDuplicate(simpleArticle, duplicateWithDifferentType, BibDatabaseMode.BIBTEX));
+    }
+
 }