svn commit: r1608005 - in /lucene/dev/trunk: lucene/core/src/java/org/apache/lucene/analysis/AnalyzerWrapper.java solr/CHANGES.txt

[us...@apache.org]

Author: uschindler
Date: Sat Jul  5 12:19:51 2014
New Revision: 1608005

URL: http://svn.apache.org/r1608005
Log:
LUCENE-5803: Add more Javadocs to AnalyzerWrapper to encourage people to use DelegatingAnalyzerWrapper (if possible). Add changes entry for Solr.

Modified:
    lucene/dev/trunk/lucene/core/src/java/org/apache/lucene/analysis/AnalyzerWrapper.java
    lucene/dev/trunk/solr/CHANGES.txt

Modified: lucene/dev/trunk/lucene/core/src/java/org/apache/lucene/analysis/AnalyzerWrapper.java
URL: http://svn.apache.org/viewvc/lucene/dev/trunk/lucene/core/src/java/org/apache/lucene/analysis/AnalyzerWrapper.java?rev=1608005&r1=1608004&r2=1608005&view=diff
==============================================================================
--- lucene/dev/trunk/lucene/core/src/java/org/apache/lucene/analysis/AnalyzerWrapper.java (original)
+++ lucene/dev/trunk/lucene/core/src/java/org/apache/lucene/analysis/AnalyzerWrapper.java Sat Jul  5 12:19:51 2014
@@ -22,13 +22,24 @@ import java.io.Reader;
 /**
  * Extension to {@link Analyzer} suitable for Analyzers which wrap
  * other Analyzers.
- * <p/>
- * {@link #getWrappedAnalyzer(String)} allows the Analyzer
+ * 
+ * <p>{@link #getWrappedAnalyzer(String)} allows the Analyzer
  * to wrap multiple Analyzers which are selected on a per field basis.
- * <p/>
- * {@link #wrapComponents(String, Analyzer.TokenStreamComponents)} allows the
+ * 
+ * <p>{@link #wrapComponents(String, Analyzer.TokenStreamComponents)} allows the
  * TokenStreamComponents of the wrapped Analyzer to then be wrapped
  * (such as adding a new {@link TokenFilter} to form new TokenStreamComponents.
+ *
+ * <p>{@link #wrapReader(String, Reader)} allows the Reader of the wrapped
+ * Analyzer to then be wrapped (such as adding a new {@link CharFilter}.
+ *
+ * <p><b>Important:</b> If you do not want to wrap the TokenStream
+ * using {@link #wrapComponents(String, Analyzer.TokenStreamComponents)}
+ * or the Reader using {@link #wrapReader(String, Reader)} and just delegate
+ * to other analyzers (like by field name), use {@link DelegatingAnalyzerWrapper}
+ * as superclass!
+ *
+ * @see DelegatingAnalyzerWrapper
  */
 public abstract class AnalyzerWrapper extends Analyzer {
 

Modified: lucene/dev/trunk/solr/CHANGES.txt
URL: http://svn.apache.org/viewvc/lucene/dev/trunk/solr/CHANGES.txt?rev=1608005&r1=1608004&r2=1608005&view=diff
==============================================================================
--- lucene/dev/trunk/solr/CHANGES.txt (original)
+++ lucene/dev/trunk/solr/CHANGES.txt Sat Jul  5 12:19:51 2014
@@ -158,6 +158,14 @@ Bug Fixes
 * SOLR-6223: SearchComponents may throw NPE when using shards.tolerant and there is a failure
   in the 'GET_FIELDS/GET_HIGHLIGHTS/GET_DEBUG' phase. (Tomás Fernández Löbbe via shalin)
 
+Optimizations
+---------------------
+
+* LUCENE-5803: Solr's schema now uses DelegatingAnalyzerWrapper. This uses less heap
+  for cached TokenStreamComponents because it caches per FieldType not per Field, so
+  indexes with many fields of same type just use one TokenStream per thread.
+  (Shay Banon, Uwe Schindler, Robert Muir)
+
 Other Changes
 ---------------------