Apache
/
commons-lang
mirror of https://github.com/apache/commons-lang.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

LANG-1269: Wrong name or result of StringUtils#getJaroWinklerDistance (closes #198) 

deprecate StringUtils#getJaroWinklerDistance and add StringUtils#getJaroWinklerSimilarity instead
This commit is contained in:
pascalschumacher 2016-10-22 11:55:32 +02:00
parent 1e5c2b8779
commit a40b2a907a
3 changed files with 80 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

1
src/changes/changes.xml
View File

@ -46,6 +46,7 @@ The <action> type attribute can be add,update,fix,remove.
<body>
<release version="3.6" date="2016-MM-DD" description="TBD">
<action issue="LANG-1269" type="fix" dev="paschuma">Wrong name or result of StringUtils#getJaroWinklerDistance</action>
<action issue="LANG-1188" type="fix" dev="paschuma">StringUtils#join(T...): warning: [unchecked] Possible heap pollution from parameterized vararg type T</action>
<action issue="LANG-1144" type="fix" dev="ggregory" due-to="Waldemar Maier, Gary Gregory">Multiple calls of org.apache.commons.lang3.concurrent.LazyInitializer.initialize() are possible.</action>
<action issue="LANG-1276" type="fix" dev="pschumacher" due-to="Andy Klimczak">StrBuilder#replaceAll ArrayIndexOutOfBoundsException</action>

51
src/main/java/org/apache/commons/lang3/StringUtils.java
View File

@ -8009,7 +8009,9 @@ public class StringUtils {
* @return result distance
* @throws IllegalArgumentException if either String input {@code null}
* @since 3.3
* @deprecated as of 3.6, due to a misleading name, use {@link #getJaroWinklerSimilarity()} instead
*/
@Deprecated
public static double getJaroWinklerDistance(final CharSequence first, final CharSequence second) {
final double DEFAULT_SCALING_FACTOR = 0.1;
@ -8027,6 +8029,55 @@ public class StringUtils {
return Math.round(jw * 100.0D) / 100.0D;
}
/**
* <p>Find the Jaro Winkler Similarity which indicates the similarity score between two Strings.</p>
*
* <p>The Jaro measure is the weighted sum of percentage of matched characters from each file and transposed characters.
* Winkler increased this measure for matching initial characters.</p>
*
* <p>This implementation is based on the Jaro Winkler similarity algorithm
* from <a href="http://en.wikipedia.org/wiki/Jaro%E2%80%93Winkler_distance">http://en.wikipedia.org/wiki/Jaro%E2%80%93Winkler_distance</a>.</p>
*
* <pre>
* StringUtils.getJaroWinklerSimilarity(null, null) = IllegalArgumentException
* StringUtils.getJaroWinklerSimilarity("","") = 0.0
* StringUtils.getJaroWinklerSimilarity("","a") = 0.0
* StringUtils.getJaroWinklerSimilarity("aaapppp", "") = 0.0
* StringUtils.getJaroWinklerSimilarity("frog", "fog") = 0.93
* StringUtils.getJaroWinklerSimilarity("fly", "ant") = 0.0
* StringUtils.getJaroWinklerSimilarity("elephant", "hippo") = 0.44
* StringUtils.getJaroWinklerSimilarity("hippo", "elephant") = 0.44
* StringUtils.getJaroWinklerSimilarity("hippo", "zzzzzzzz") = 0.0
* StringUtils.getJaroWinklerSimilarity("hello", "hallo") = 0.88
* StringUtils.getJaroWinklerSimilarity("ABC Corporation", "ABC Corp") = 0.93
* StringUtils.getJaroWinklerSimilarity("D N H Enterprises Inc", "D &amp; H Enterprises, Inc.") = 0.95
* StringUtils.getJaroWinklerSimilarity("My Gym Children's Fitness Center", "My Gym. Childrens Fitness") = 0.92
* StringUtils.getJaroWinklerSimilarity("PENNSYLVANIA", "PENNCISYLVNIA") = 0.88
* </pre>
*
* @param first the first String, must not be null
* @param second the second String, must not be null
* @return result similarity
* @throws IllegalArgumentException if either String input {@code null}
* @since 3.6
*/
public static double getJaroWinklerSimilarity(final CharSequence first, final CharSequence second) {
final double DEFAULT_SCALING_FACTOR = 0.1;
if (first == null || second == null) {
throw new IllegalArgumentException("Strings must not be null");
}
int[] mtp = matches(first, second);
double m = mtp[0];
if (m == 0) {
return 0D;
}
double j = ((m / first.length() + m / second.length() + (m - mtp[1]) / m)) / 3;
double jw = j < 0.7D ? j : j + Math.min(DEFAULT_SCALING_FACTOR, 1D / mtp[3]) * mtp[2] * (1D - j);
return Math.round(jw * 100.0D) / 100.0D;
}
private static int[] matches(final CharSequence first, final CharSequence second) {
CharSequence max, min;
if (first.length() > second.length()) {

28
src/test/java/org/apache/commons/lang3/StringUtilsTest.java
View File

@ -2384,6 +2384,34 @@ public class StringUtilsTest {
StringUtils.getJaroWinklerDistance(null, "clear");
}
@Test
public void testGetJaroWinklerSimilarity_StringString() {
assertEquals(0.93d, StringUtils.getJaroWinklerSimilarity("frog", "fog"), 0.0d);
assertEquals(0.0d, StringUtils.getJaroWinklerSimilarity("fly", "ant"), 0.0d);
assertEquals(0.44d, StringUtils.getJaroWinklerSimilarity("elephant", "hippo"), 0.0d);
assertEquals(0.84d, StringUtils.getJaroWinklerSimilarity("dwayne", "duane"), 0.0d);
assertEquals(0.93d, StringUtils.getJaroWinklerSimilarity("ABC Corporation", "ABC Corp"), 0.0d);
assertEquals(0.95d, StringUtils.getJaroWinklerSimilarity("D N H Enterprises Inc", "D & H Enterprises, Inc."), 0.0d);
assertEquals(0.92d, StringUtils.getJaroWinklerSimilarity("My Gym Children's Fitness Center", "My Gym. Childrens Fitness"), 0.0d);
assertEquals(0.88d, StringUtils.getJaroWinklerSimilarity("PENNSYLVANIA", "PENNCISYLVNIA"), 0.0d);
assertEquals(0.63d, StringUtils.getJaroWinklerSimilarity("Haus Ingeborg", "Ingeborg Esser"), 0.0d);
}
@Test(expected = IllegalArgumentException.class)
public void testGetJaroWinklerSimilarity_NullNull() throws Exception {
StringUtils.getJaroWinklerSimilarity(null, null);
}
@Test(expected = IllegalArgumentException.class)
public void testGetJaroWinklerSimilarity_StringNull() throws Exception {
StringUtils.getJaroWinklerSimilarity(" ", null);
}
@Test(expected = IllegalArgumentException.class)
public void testGetJaroWinklerSimilarity_NullString() throws Exception {
StringUtils.getJaroWinklerSimilarity(null, "clear");
}
@Test
public void testGetFuzzyDistance() throws Exception {
assertEquals(0, StringUtils.getFuzzyDistance("", "", Locale.ENGLISH));