Fixes JabRef#1181 and JabRef#1504: Improved "Normalize to BibTeX name…

… format"

Added the jr, sr,... special cases for semicolon partition.
Fixed to avoid the "and", "{", ";" cases.
Added Test for every case.

CHANGELOG.md

@@ -30,7 +30,7 @@ We refer to [GitHub issues](https://github.com/JabRef/jabref/issues) by using `#
 - Fixed field `location` containing only city is not exported correctly to MS-Office 2007 xml format
 
 ### Removed
-- The non-supported feature of being able to define file directories for any extension is removed. Still, it should work for older databases using the legacy `ps` and `pdf` fields, although we strongly encourage using the `file` field. 
+- The non-supported feature of being able to define file directories for any extension is removed. Still, it should work for older databases using the legacy `ps` and `pdf` fields, although we strongly encourage using the `file` field.
 
 
 
@@ -195,6 +195,7 @@ We refer to [GitHub issues](https://github.com/JabRef/jabref/issues) by using `#
 - Manage content selectors now saves edited existing lists again and only marks database as changed when the content selectors are changed
 - When inserting a duplicate the right entry will be selected
 - Preview panel height is now saved immediately, thus is shown correctly if the panel height is changed, closed and opened again
+- Fixed [#1181](https://github.com/JabRef/jabref/issues/1181) and [#1504](https://github.com/JabRef/jabref/issues/1504): Improved "Normalize to BibTeX name format": Support separated names with commas and colons. Considered name affixes such as "Jr".
 
 ### Removed
 - [#1610](https://github.com/JabRef/jabref/issues/1610) Removed the possibility to auto show or hide the groups interface

src/main/java/net/sf/jabref/logic/formatter/bibtexfields/NormalizeNamesFormatter.java

@@ -1,5 +1,11 @@
 package net.sf.jabref.logic.formatter.bibtexfields;
 
+import java.util.Arrays;
+import java.util.Collection;
+import java.util.HashSet;
+import java.util.Objects;
+import java.util.StringJoiner;
+
 import net.sf.jabref.logic.formatter.Formatter;
 import net.sf.jabref.logic.l10n.Localization;
 import net.sf.jabref.model.entry.AuthorList;
@@ -9,6 +15,9 @@
  */
 public class NormalizeNamesFormatter implements Formatter {
 
+    // Avoid partition where these values are contained
+    private final Collection<String> avoidTermsInLowerCase = Arrays.asList("jr", "sr", "jnr", "snr", "von", "zu", "van", "der");
+
     @Override
     public String getName() {
         return Localization.lang("Normalize names of persons");
@@ -21,6 +30,73 @@ public String getKey() {
 
     @Override
     public String format(String value) {
+        Objects.requireNonNull(value);
+        // Handle case names in order lastname, firstname and separated by ","
+        // E.g., Ali Babar, M., Dingsøyr, T., Lago, P., van der Vliet, H.
+        if (!value.contains(" and ") && !value.contains("{") && !value.contains(";")) {
+            String[] valueParts = value.split(",");
+            // Delete spaces for correct case identification
+            for(int i=0; i < valueParts.length; i++) {
+                valueParts[i] = valueParts[i].trim();
+            }
+            // Looking for space between pre- and lastname
+            boolean spaceInAllParts = false;
+            for (int i=0; i<valueParts.length; i++) {
+                if (valueParts[i].contains(" ") ) {
+                    spaceInAllParts = true;
+                } else {
+                    spaceInAllParts = false;
+                    break;
+                }
+            }
+
+            // We hit the comma name separator case
+            // Usually the getAsLastFirstNamesWithAnd method would separate them if pre- and lastname are separated with "and"
+            // If not, we check if spaces separate pre- and lastname
+            if (spaceInAllParts) {
+                value = value.replaceAll(",", " and");
+            } else {
+                // Looking for name affixes to avoid
+                // partCount need to reduce by the count off avoiding terms
+                int valuePartsCount = valueParts.length;
+                // Holding the index of every term, which need to avoid
+                Collection<Integer> avoidIndex = new HashSet<>();
+
+                for (int i = 0; i < valueParts.length; i++) {
+                    if (avoidTermsInLowerCase.contains(valueParts[i].toLowerCase())) {
+                        avoidIndex.add(i);
+                        valuePartsCount--;
+                    }
+                }
+
+                if ((valuePartsCount % 2) == 0) {
+                    // We hit the described special case with name affix like Jr
+                    StringBuilder stringBuilder = new StringBuilder();
+                    // avoidedTimes need to increase the count of avoided terms for correct module calculation
+                    int avoidedTimes = 0;
+                    for (int i = 0; i < valueParts.length; i++) {
+                        if (avoidIndex.contains(i)) {
+                            // We hit a name affix
+                            stringBuilder.append(valueParts[i]);
+                            stringBuilder.append(',');
+                            avoidedTimes++;
+                        } else {
+                            stringBuilder.append(valueParts[i]);
+                            if (((i + avoidedTimes) % 2) == 0) {
+                                // Hit separation between last name and firstname --> comma has to be kept
+                                stringBuilder.append(',');
+                            } else {
+                                // Hit separation between full names (e.g., Ali Babar, M. and Dingsøyr, T.) --> semicolon has to be used
+                                // Will be treated correctly by AuthorList.parse(value);
+                                stringBuilder.append(';');
+                            }
+                        }
+                    }
+                    value = stringBuilder.toString();
+                }
+            }
+        }
+
         AuthorList authorList = AuthorList.parse(value);
         return authorList.getAsLastFirstNamesWithAnd(false);
     }
@@ -35,4 +111,14 @@ public String getExampleInput() {
         return "Albert Einstein and Alan Turing";
     }
 
+    private static boolean contains(final String[] array, final String[] searchTerms) {
+        for (String currentTerm : array) {
+            for (String beCompared : searchTerms) {
+                if (beCompared.trim().toLowerCase().equals(currentTerm.trim().toLowerCase())) {
+                    return true;
+                }
+            }
+        }
+        return false;
+    }
 }

src/test/java/net/sf/jabref/logic/formatter/bibtexfields/NormalizeNamesFormatterTest.java

@@ -98,6 +98,18 @@ public void lastThenJuniorThenFirst() {
         expectCorrect("Name, della, first", "Name, della, first");
     }
 
+    @Test
+    public void testConcatenationOfAuthorsWithCommas() {
+        expectCorrect("Ali Babar, M., Dingsøyr, T., Lago, P., van der Vliet, H.",
+                "Ali Babar, M. and Dingsøyr, T. and Lago, P. and van der Vliet, H.");
+        expectCorrect("Ali Babar, M.", "Ali Babar, M.");
+    }
+
+    @Test
+    public void testOddCountOfCommas() {
+        expectCorrect("Ali Babar, M., Dingsøyr, T., Lago P.", "Ali Babar, M., Dingsøyr T. Lago P.");
+    }
+
     private void expectCorrect(String input, String expected) {
         Assert.assertEquals(expected, formatter.format(input));
     }
@@ -107,4 +119,35 @@ public void formatExample() {
         assertEquals("Einstein, Albert and Turing, Alan", formatter.format(formatter.getExampleInput()));
     }
 
-}
+    @Test
+    public void testNameAffixe() {
+        expectCorrect("Surname, jr, First, Surname2, First2", "Surname, jr, First and Surname2, First2");
+    }
+
+    @Test
+    public void testAvoidSpecialCharacter() {
+        expectCorrect("Surname, {, First; Surname2, First2", "Surname, {, First; Surname2, First2");
+    }
+
+    @Test
+    public void testAndInName() {
+        expectCorrect("Surname, and , First, Surname2, First2", "Surname and , First, Surname2 First2");
+    }
+
+    @Test
+    public void testMultipleNameAffixes() {
+        expectCorrect("Mair, Jr, Daniel, Brühl, Sr, Daniel", "Mair, Jr, Daniel and Brühl, Sr, Daniel");
+    }
+
+    @Test
+    public void testCommaSeperatedNames() {
+        expectCorrect("Cristina Bosoi, Mariana Oliveira, Rafael Ochoa Sanchez, Mélanie Tremblay, Gabrie TenHave, Nicoolas Deutz, Christopher F. Rose, Chantal Bemeur",
+                "Bosoi, Cristina and Oliveira, Mariana and Sanchez, Rafael Ochoa and Tremblay, Mélanie and TenHave, Gabrie and Deutz, Nicoolas and Rose, Christopher F. and Bemeur, Chantal");
+    }
+
+    @Test
+    public void testMultipleSpaces() {
+        expectCorrect("Cristina    Bosoi,    Mariana Oliveira, Rafael Ochoa Sanchez   ,   Mélanie Tremblay  , Gabrie TenHave, Nicoolas Deutz, Christopher F. Rose, Chantal Bemeur",
+                "Bosoi, Cristina and Oliveira, Mariana and Sanchez, Rafael Ochoa and Tremblay, Mélanie and TenHave, Gabrie and Deutz, Nicoolas and Rose, Christopher F. and Bemeur, Chantal");
+    }
+}

src/test/java/net/sf/jabref/model/entry/AuthorListTest.java

@@ -524,6 +524,15 @@ public void testGetAuthorsLastFirstAnds() {
 
     }
 
+    @Test
+    public void testGetAuthorsLastFirstAndsCaching() {
+        // getAsLastFirstNamesWithAnd caches its results, therefore we call the method twice using the same arguments
+        Assert.assertEquals("Smith, John", AuthorList.parse("John Smith").getAsLastFirstNamesWithAnd(false));
+        Assert.assertEquals("Smith, John", AuthorList.parse("John Smith").getAsLastFirstNamesWithAnd(false));
+        Assert.assertEquals("Smith, J.", AuthorList.parse("John Smith").getAsLastFirstNamesWithAnd(true));
+        Assert.assertEquals("Smith, J.", AuthorList.parse("John Smith").getAsLastFirstNamesWithAnd(true));
+    }
+
     @Test
     public void testGetAuthorsFirstFirst() {
 
@@ -636,4 +645,5 @@ public void parseNameWithBraces() throws Exception {
         Author expected = new Author("H{e}lene", "H.", null, "Fiaux", null);
         Assert.assertEquals(new AuthorList(expected), AuthorList.parse("H{e}lene Fiaux"));
     }
+
 }