From 31ead5af5869cb4b42658e01bcc6fa28166e3cf7 Mon Sep 17 00:00:00 2001
From: Steven Rowe <sarowe@apache.org>
Date: Wed, 16 May 2012 20:03:49 +0000
Subject: [PATCH] LUCENE-4063: FrenchLightStemmer no longer deletes repeated
 digits.

git-svn-id: https://svn.apache.org/repos/asf/lucene/dev/trunk@1339333 13f79535-47bb-0310-9956-ffa450edef68
---
 lucene/CHANGES.txt                               |  3 +++
 .../lucene/analysis/fr/FrenchLightStemmer.java   |  4 ++--
 .../analysis/fr/TestFrenchLightStemFilter.java   | 16 ++++++++++++++++
 3 files changed, 21 insertions(+), 2 deletions(-)

diff --git a/lucene/CHANGES.txt b/lucene/CHANGES.txt
index 6025da6c306..dce825156f8 100644
--- a/lucene/CHANGES.txt
+++ b/lucene/CHANGES.txt
@@ -513,6 +513,9 @@ API Changes
   immutable instances of NormalizeCharMap. (Dawid Weiss, Mike
   McCandless)
 
+* LUCENE-4063: FrenchLightStemmer no longer deletes repeated digits.
+  (Tanguy Moal via Steve Rowe)
+
 New features
 
 * LUCENE-2604: Added RegexpQuery support to QueryParser. Regular expressions
diff --git a/lucene/analysis/common/src/java/org/apache/lucene/analysis/fr/FrenchLightStemmer.java b/lucene/analysis/common/src/java/org/apache/lucene/analysis/fr/FrenchLightStemmer.java
index 43e2e0625d1..5be44bc032e 100644
--- a/lucene/analysis/common/src/java/org/apache/lucene/analysis/fr/FrenchLightStemmer.java
+++ b/lucene/analysis/common/src/java/org/apache/lucene/analysis/fr/FrenchLightStemmer.java
@@ -246,7 +246,7 @@ public class FrenchLightStemmer {
       
       char ch = s[0];
       for (int i = 1; i < len; i++) {
-        if (s[i] == ch)
+        if (s[i] == ch && Character.isLetter(ch))
           len = delete(s, i--, len);
         else
           ch = s[i];
@@ -260,7 +260,7 @@ public class FrenchLightStemmer {
         if (s[len-1] == 'r') len--;
         if (s[len-1] == 'e') len--;
         if (s[len-1] == 'e') len--;
-        if (s[len-1] == s[len-2]) len--;
+        if (s[len-1] == s[len-2] && Character.isLetter(s[len-1])) len--;
     }
     return len;
   }
diff --git a/lucene/analysis/common/src/test/org/apache/lucene/analysis/fr/TestFrenchLightStemFilter.java b/lucene/analysis/common/src/test/org/apache/lucene/analysis/fr/TestFrenchLightStemFilter.java
index 6012e8282a9..d0f7af15c24 100644
--- a/lucene/analysis/common/src/test/org/apache/lucene/analysis/fr/TestFrenchLightStemFilter.java
+++ b/lucene/analysis/common/src/test/org/apache/lucene/analysis/fr/TestFrenchLightStemFilter.java
@@ -153,6 +153,22 @@ public class TestFrenchLightStemFilter extends BaseTokenStreamTestCase {
     
     checkOneTerm(analyzer, "disposition", "dispos");
     checkOneTerm(analyzer, "dispose", "dispos");
+
+    // SOLR-3463 : abusive compression of repeated characters in numbers
+    // Trailing repeated char elision :
+    checkOneTerm(analyzer, "1234555", "1234555");
+    // Repeated char within numbers with more than 4 characters :
+    checkOneTerm(analyzer, "12333345", "12333345");
+    // Short numbers weren't affected already:
+    checkOneTerm(analyzer, "1234", "1234");
+    // Ensure behaviour is preserved for words!
+    // Trailing repeated char elision :
+    checkOneTerm(analyzer, "abcdeff", "abcdef");
+    // Repeated char within words with more than 4 characters :
+    checkOneTerm(analyzer, "abcccddeef", "abcdef");
+    checkOneTerm(analyzer, "créées", "cre");
+    // Combined letter and digit repetition
+    checkOneTerm(analyzer, "22hh00", "22h00"); // 10:00pm
   }
   
   /** Test against a vocabulary from the reference impl */