[ https://issues.apache.org/jira/browse/LANG-936?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13878391#comment-13878391 ]
Henri Yandell edited comment on LANG-936 at 1/22/14 8:20 AM: ------------------------------------------------------------- Thanks Yaniv + Eli :) svn ci -m "Applying Eli Lindsey's patch to Yaniv Kunda's report in LANG-936 that StringUtils.getLevensteinDistance(String, String, int) gave the wrong answer when the int threshold is near Integer.MAX_VALUE" . Sending src/changes/changes.xml Sending src/main/java/org/apache/commons/lang3/StringUtils.java Sending src/test/java/org/apache/commons/lang3/StringUtilsTest.java Transmitting file data ... Committed revision 1560275. was (Author: bayard): Thanks Yanix + Eli :) svn ci -m "Applying Eli Lindsey's patch to Yaniv Kunda's report in LANG-936 that StringUtils.getLevensteinDistance(String, String, int) gave the wrong answer when the int threshold is near Integer.MAX_VALUE" . Sending src/changes/changes.xml Sending src/main/java/org/apache/commons/lang3/StringUtils.java Sending src/test/java/org/apache/commons/lang3/StringUtilsTest.java Transmitting file data ... Committed revision 1560275. > StringUtils.getLevenshteinDistance with too big of a threshold returns wrong > result > ----------------------------------------------------------------------------------- > > Key: LANG-936 > URL: https://issues.apache.org/jira/browse/LANG-936 > Project: Commons Lang > Issue Type: Bug > Components: lang.* > Affects Versions: 3.1 > Reporter: Yaniv Kunda > Priority: Minor > Fix For: 3.3 > > > StringUtils.getLevenshteinDistance(CharSequence s, CharSequence t, int > threshold) specifies: > {quote} > {{Find the Levenshtein distance between two Strings if it's _+*less than or > equal to*+_ a given threshold.}} > {quote} > When passing a threshold > *Integer.MAX_VALUE - max(s.length(), t.length())* > the method always returns -1. > The simplest use case is passing *Integer.MAX_VALUE* (a common practice if > one would want to find the min/max LD of a string to several other strings in > an iterative fashion. > The code should be fixed to consider the threshold in relation to the > source/target lengths, or alternatively the javadoc should be fixed to > pronounce the current limit. -- This message was sent by Atlassian JIRA (v6.1.5#6160)