Add DOI matching with not full coincidence.

JabRef · Aug 8, 2020 · 6c8b7d7 · 6c8b7d7
1 parent 04a8d3e
commit 6c8b7d7
Show file tree

Hide file tree

Showing 5 changed files with 57 additions and 1 deletion.
diff --git a/src/main/java/org/jabref/logic/bibtex/DuplicateCheck.java b/src/main/java/org/jabref/logic/bibtex/DuplicateCheck.java
@@ -303,6 +303,15 @@ public boolean isDuplicate(final BibEntry one, final BibEntry two, final BibData
  return true;
  }
 
+ // check DOI
+ if (one.getDOI().isPresent() && two.getDOI().isPresent() && one.getDOI().get().isCompareNotExact(two.getDOI().get())) {
+ return true;
+ }
+ // check ISBN
+ if (one.getISBN().isPresent() && two.getISBN().isPresent() && one.getISBN().get().equals(two.getISBN().get())) {
+ return true;
+ }
+
  if (haveDifferentEntryType(one, two) ||
  haveDifferentEditions(one, two) ||
  haveDifferentChaptersOrPagesOfTheSameBook(one, two)) {

diff --git a/src/main/java/org/jabref/model/entry/BibEntry.java b/src/main/java/org/jabref/model/entry/BibEntry.java
@@ -30,6 +30,7 @@
 import org.jabref.model.entry.field.OrFields;
 import org.jabref.model.entry.field.StandardField;
 import org.jabref.model.entry.identifier.DOI;
+import org.jabref.model.entry.identifier.ISBN;
 import org.jabref.model.entry.types.EntryType;
 import org.jabref.model.entry.types.IEEETranEntryType;
 import org.jabref.model.entry.types.StandardEntryType;
@@ -459,6 +460,10 @@ public Optional<DOI> getDOI() {
  return getField(StandardField.DOI).flatMap(DOI::parse);
  }
 
+ public Optional<ISBN> getISBN() {
+ return getField(StandardField.ISBN).flatMap(ISBN::parse);
+ }
+
  /**
  * Return the LaTeX-free contents of the given field or its alias an an Optional
  * <p>

diff --git a/src/main/java/org/jabref/model/entry/identifier/DOI.java b/src/main/java/org/jabref/model/entry/identifier/DOI.java
@@ -234,6 +234,12 @@ public String getNormalized() {
  return doi;
  }
 
+ public boolean isCompareNotExact(DOI o2) {
+ String s1 = this.doi.replaceAll("[^\\w]", "");
+ String s2 = o2.doi.replaceAll("[^\\w]", "");
+ return s1.equalsIgnoreCase(s2);
+ }
+
  /**
  * DOIs are case-insensitive. Thus, 10.1109/cloud.2017.89 equals 10.1109/CLOUD.2017.89
  */
@@ -253,5 +259,4 @@ public boolean equals(Object o) {
  public int hashCode() {
  return Objects.hash(doi.toLowerCase(Locale.ENGLISH));
  }
-
 }
diff --git a/src/main/java/org/jabref/model/entry/identifier/ISBN.java b/src/main/java/org/jabref/model/entry/identifier/ISBN.java
@@ -2,6 +2,7 @@
 
 import java.net.URI;
 import java.net.URISyntaxException;
+import java.util.Locale;
 import java.util.Objects;
 import java.util.Optional;
 import java.util.regex.Matcher;
@@ -106,4 +107,21 @@ public Optional<URI> getExternalURI() {
  return Optional.empty();
  }
  }
+
+ @Override
+ public boolean equals(Object o) {
+ if (this == o) {
+ return true;
+ }
+ if ((o == null) || (getClass() != o.getClass())) {
+ return false;
+ }
+ ISBN other = (ISBN) o;
+ return isbnString.equalsIgnoreCase(other.isbnString);
+ }
+
+ @Override
+ public int hashCode() {
+ return Objects.hash(isbnString.toLowerCase(Locale.ENGLISH));
+ }
 }
diff --git a/src/test/java/org/jabref/logic/bibtex/DuplicateCheckTest.java b/src/test/java/org/jabref/logic/bibtex/DuplicateCheckTest.java
@@ -161,6 +161,25 @@ public void twoEntriesWithSameDoiButDifferentTypesAreDuplicates() {
  assertTrue(duplicateChecker.isDuplicate(simpleArticle, duplicateWithDifferentType, BibDatabaseMode.BIBTEX));
  }
 
+ @Test
+ public void twoEntriesWithSameDoiButDifferentTypesAreDuplicates2() {
+ simpleArticle.setField(StandardField.DOI, "10.1016/j.is.2004.02.002");
+ unrelatedArticle.setField(StandardField.DOI, "10.1016/j.is.2004.02.0\\02");
+ BibEntry duplicateWithDifferentType = (BibEntry) unrelatedArticle;
+ duplicateWithDifferentType.setType(StandardEntryType.InCollection);
+
+ assertTrue(duplicateChecker.isDuplicate(simpleArticle, duplicateWithDifferentType, BibDatabaseMode.BIBTEX));
+ }
+ @Test
+ public void twoEntriesWithSameISBNButDifferentTypesAreDuplicates() {
+ simpleArticle.setField(StandardField.ISBN, "0-123456-47-9");
+ unrelatedArticle.setField(StandardField.ISBN, "0-123456-47-9");
+ BibEntry duplicateWithDifferentType = (BibEntry) unrelatedArticle;
+ duplicateWithDifferentType.setType(StandardEntryType.InCollection);
+
+ assertTrue(duplicateChecker.isDuplicate(simpleArticle, duplicateWithDifferentType, BibDatabaseMode.BIBTEX));
+ }
+
  @Test
  public void twoInbooksWithDifferentChaptersAreNotDuplicates() {
  twoEntriesWithDifferentSpecificFieldsAreNotDuplicates(simpleInbook, StandardField.CHAPTER,