LUCENE-3720: fix BeiderMorseFilter OOM issues

git-svn-id: https://svn.apache.org/repos/asf/lucene/dev/trunk@1386662 13f79535-47bb-0310-9956-ffa450edef68

dev-tools/eclipse/dot.classpath

@@ -103,7 +103,7 @@
   <classpathentry kind="lib" path="lucene/test-framework/lib/junit-4.10.jar"/>
   <classpathentry kind="lib" path="lucene/sandbox/lib/jakarta-regexp-1.4.jar"/>
   <classpathentry kind="lib" path="lucene/analysis/icu/lib/icu4j-49.1.jar"/>
-  <classpathentry kind="lib" path="lucene/analysis/phonetic/lib/commons-codec-1.6.jar"/>
+  <classpathentry kind="lib" path="lucene/analysis/phonetic/lib/commons-codec-1.7.jar"/>
   <classpathentry kind="lib" path="lucene/analysis/morfologik/lib/morfologik-fsa-1.5.3.jar"/>
   <classpathentry kind="lib" path="lucene/analysis/morfologik/lib/morfologik-polish-1.5.3.jar"/>
   <classpathentry kind="lib" path="lucene/analysis/morfologik/lib/morfologik-stemming-1.5.3.jar"/>

dev-tools/maven/pom.xml.template

@@ -152,7 +152,7 @@
       <dependency>
         <groupId>commons-codec</groupId>
         <artifactId>commons-codec</artifactId>
-        <version>1.6</version>
+        <version>1.7</version>
       </dependency>
       <dependency>
         <groupId>commons-cli</groupId>

lucene/CHANGES.txt

@@ -164,6 +164,9 @@ Bug Fixes
   RAM, instead of checking the estimated size of the segment
   to decide whether to cache it. (Mike McCandless)
 
+* LUCENE-3720: fix memory-consumption issues with BeiderMorseFilter.
+  (Thomas Neidhart via Robert Muir)
+
 Optimizations
 
 * LUCENE-4322: Decrease lucene-core JAR size. The core JAR size had increased a

lucene/analysis/README.txt

@@ -31,7 +31,7 @@ lucene-analyzers-morfologik-XX.jar
 lucene-analyzers-phonetic-XX.jar
   An add-on analysis library that provides phonetic encoders via Apache
   Commons-Codec. Note: this module depends on the commons-codec jar 
-  file (version >= 1.4)
+  file
   
 lucene-analyzers-smartcn-XX.jar
   An add-on analysis library that provides word segmentation for Simplified

lucene/analysis/phonetic/build.xml

@@ -27,7 +27,7 @@
 
   <path id="classpath">
     <pathelement path="${analyzers-common.jar}"/>
-    <pathelement path="lib/commons-codec-1.6.jar"/>
+    <pathelement path="lib/commons-codec-1.7.jar"/>
     <path refid="base.classpath"/>
   </path>
 

lucene/analysis/phonetic/ivy.xml

@@ -19,7 +19,7 @@
 <ivy-module version="2.0">
     <info organisation="org.apache.lucene" module="analyzers-phonetic"/>
     <dependencies>
-      <dependency org="commons-codec" name="commons-codec" rev="1.6" transitive="false"/>
+      <dependency org="commons-codec" name="commons-codec" rev="1.7" transitive="false"/>
       <exclude org="*" ext="*" matcher="regexp" type="${ivy.exclude.types}"/> 
     </dependencies>
 </ivy-module>

lucene/analysis/phonetic/src/java/org/apache/lucene/analysis/phonetic/BeiderMorseFilter.java

@@ -32,12 +32,6 @@ import org.apache.lucene.analysis.tokenattributes.PositionIncrementAttribute;
 
 /**
  * TokenFilter for Beider-Morse phonetic encoding.
- * <p>
- * <b><font color="red">
- * WARNING: some inputs can cause extremely high RAM usage! 
- * https://issues.apache.org/jira/browse/CODEC-132
- * </font></b>
- * </p>
  * @see BeiderMorseEncoder
  * @lucene.experimental
  */

lucene/analysis/phonetic/src/test/org/apache/lucene/analysis/phonetic/TestBeiderMorseFilter.java

@@ -89,7 +89,6 @@ public class TestBeiderMorseFilter extends BaseTokenStreamTestCase {
         new int[] { 1 });
   }
 
-  @Ignore("broken: causes OOM on some strings (https://issues.apache.org/jira/browse/CODEC-132)")
   public void testRandom() throws Exception {
     checkRandomData(random(), analyzer, 1000 * RANDOM_MULTIPLIER); 
   }

lucene/licenses/commons-codec-1.6.jar.sha1

@@ -1 +0,0 @@
-b7f0fc8f61ecadeb3695f0b9464755eee44374d4

lucene/licenses/commons-codec-1.7.jar.sha1

@@ -0,0 +1 @@
+9cd61d269c88f9fb0eb36cea1efcd596ab74772f

solr/core/ivy.xml

@@ -20,7 +20,7 @@
     <info organisation="org.apache.solr" module="core"/>
 
     <dependencies>
-      <dependency org="commons-codec" name="commons-codec" rev="1.6" transitive="false"/>
+      <dependency org="commons-codec" name="commons-codec" rev="1.7" transitive="false"/>
       <dependency org="commons-fileupload" name="commons-fileupload" rev="1.2.1" transitive="false"/>
       <dependency org="commons-cli" name="commons-cli" rev="1.2" transitive="false"/>
       <dependency org="commons-lang" name="commons-lang" rev="2.6" transitive="false"/>

solr/licenses/commons-codec-1.6.jar.sha1

@@ -1 +0,0 @@
-b7f0fc8f61ecadeb3695f0b9464755eee44374d4

solr/licenses/commons-codec-1.7.jar.sha1

@@ -0,0 +1 @@
+9cd61d269c88f9fb0eb36cea1efcd596ab74772f