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

SOLR-468: Changed semantics of CapitalizationFilterFactory to keep the original token, not the value in the map 

git-svn-id: https://svn.apache.org/repos/asf/lucene/solr/trunk@630572 13f79535-47bb-0310-9956-ffa450edef68
This commit is contained in:
Grant Ingersoll 2008-02-24 02:34:15 +00:00
parent 8cf3175518
commit ac58a1a4dd
3 changed files with 194 additions and 126 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
CHANGES.txt
View File

@ -106,6 +106,8 @@ New Features
17. SOLR-248: Added CapitalizationFilterFactory that creates tokens with
normalized capitalization. This filter is useful for facet display,
but will not work with a prefix query. (ryan)
SOLR-468: Change to the semantics to keep the original token, not the token in the Map. Also switched to
use Lucene's new reusable token capabilities. (gsingers)
18. SOLR-307: Added NGramFilterFactory and EdgeNGramFilterFactory.
(Thomas Peuss via Otis Gospodnetic)

267
src/java/org/apache/solr/analysis/CapitalizationFilterFactory.java
View File

@ -17,199 +17,226 @@
package org.apache.solr.analysis;
import java.io.IOException;
import java.util.ArrayList;
import java.util.Collection;
import java.util.HashMap;
import java.util.Map;
import java.util.StringTokenizer;
import org.apache.lucene.analysis.CharArraySet;
import org.apache.lucene.analysis.Token;
import org.apache.lucene.analysis.TokenFilter;
import org.apache.lucene.analysis.TokenStream;
import org.apache.solr.core.SolrConfig;
import org.apache.solr.analysis.BaseTokenFilterFactory;
import java.io.IOException;
import java.util.ArrayList;
import java.util.Collection;
import java.util.Collections;
import java.util.Map;
import java.util.StringTokenizer;
/**
* A filter to apply normal capitalization rules to Tokens. It will make the first letter
* capital and the rest lower case.
*
* capital and the rest lower case.
* <p/>
* This filter is particularly useful to build nice looking facet parameters. This filter
* is not appropriate if you intend to use a prefix query.
*
* The factory takes parameters:
* "onlyFirstWord" - should each word be capitalized or all of the words?
* "keep" - a keep word list. Each word that should be kept separated by whitespace.
* <p/>
* The factory takes parameters:<br/>
* "onlyFirstWord" - should each word be capitalized or all of the words?<br/>
* "keep" - a keep word list. Each word that should be kept separated by whitespace.<br/>
* "keepIgnoreCase - true or false. If true, the keep list will be considered case-insensitive.
* "forceFirstLetter" - Force the first letter to be capitalized even if it is in the keep list<br/>
* "okPrefix" - do not change word capitalization if a word begins with something in this list.
* for example if "McK" is on the okPrefix list, the word "McKinley" should not be changed to
* "Mckinley"
* "minWordLength" - how long the word needs to be to get capitalization applied. If the
* minWordLength is 3, "and" > "And" but "or" stays "or"
* for example if "McK" is on the okPrefix list, the word "McKinley" should not be changed to
* "Mckinley"<br/>
* "minWordLength" - how long the word needs to be to get capitalization applied. If the
* minWordLength is 3, "and" > "And" but "or" stays "or"<br/>
* "maxWordCount" - if the token contains more then maxWordCount words, the capitalization is
* assumed to be correct.
*
* @since solr 1.3
* assumed to be correct.<br/>
*
* @version $Id$
* @since solr 1.3
*/
public class CapitalizationFilterFactory extends BaseTokenFilterFactory
{
public class CapitalizationFilterFactory extends BaseTokenFilterFactory {
public static final int DEFAULT_MAX_WORD_COUNT = Integer.MAX_VALUE;
public static final String KEEP = "keep";
public static final String KEEP_IGNORE_CASE = "keepIgnoreCase";
public static final String OK_PREFIX = "okPrefix";
public static final String MIN_WORD_LENGTH = "minWordLength";
public static final String MAX_WORD_COUNT = "maxWordCount";
public static final String MAX_TOKEN_LENGTH = "maxTokenLength";
public static final String ONLY_FIRST_WORD = "onlyFirstWord";
public static final String FORCE_FIRST_LETTER = "forceFirstLetter";
Map<String,String> keep = new HashMap<String, String>(); // not synchronized because it is only initialized once
Collection<String> okPrefix = new ArrayList<String>(); // for Example: McK
//Map<String,String> keep = new HashMap<String, String>(); // not synchronized because it is only initialized once
CharArraySet keep;
Collection<char[]> okPrefix = Collections.emptyList(); // for Example: McK
int minWordLength = 0; // don't modify capitalization for words shorter then this
int maxWordCount = Integer.MAX_VALUE;
int maxTokenLength = Integer.MAX_VALUE;
int maxWordCount = DEFAULT_MAX_WORD_COUNT;
int maxTokenLength = DEFAULT_MAX_WORD_COUNT;
boolean onlyFirstWord = true;
boolean forceFirstLetter = true; // make sure the first letter is capitol even if it is in the keep list
@Override
public void init(Map<String,String> args) {
super.init( args );
String k = args.get( KEEP );
if( k != null ) {
StringTokenizer st = new StringTokenizer( k );
while( st.hasMoreTokens() ) {
public void init(Map<String, String> args) {
super.init(args);
String k = args.get(KEEP);
if (k != null) {
StringTokenizer st = new StringTokenizer(k);
boolean ignoreCase = false;
String ignoreStr = args.get(KEEP_IGNORE_CASE);
if ("true".equalsIgnoreCase(ignoreStr)) {
ignoreCase = true;
}
keep = new CharArraySet(10, ignoreCase);
while (st.hasMoreTokens()) {
k = st.nextToken().trim();
keep.put( k.toUpperCase(), k );
keep.add(k.toCharArray());
}
}
k = args.get( OK_PREFIX );
if( k != null ) {
StringTokenizer st = new StringTokenizer( k );
while( st.hasMoreTokens() ) {
okPrefix.add( st.nextToken().trim() );
k = args.get(OK_PREFIX);
if (k != null) {
okPrefix = new ArrayList<char[]>();
StringTokenizer st = new StringTokenizer(k);
while (st.hasMoreTokens()) {
okPrefix.add(st.nextToken().trim().toCharArray());
}
}
k = args.get( MIN_WORD_LENGTH );
if( k != null ) {
minWordLength = Integer.valueOf( k );
k = args.get(MIN_WORD_LENGTH);
if (k != null) {
minWordLength = Integer.valueOf(k);
}
k = args.get( MAX_WORD_COUNT );
if( k != null ) {
maxWordCount = Integer.valueOf( k );
k = args.get(MAX_WORD_COUNT);
if (k != null) {
maxWordCount = Integer.valueOf(k);
}
k = args.get( MAX_TOKEN_LENGTH );
if( k != null ) {
maxTokenLength = Integer.valueOf( k );
k = args.get(MAX_TOKEN_LENGTH);
if (k != null) {
maxTokenLength = Integer.valueOf(k);
}
k = args.get( ONLY_FIRST_WORD );
if( k != null ) {
onlyFirstWord = Boolean.valueOf( k );
k = args.get(ONLY_FIRST_WORD);
if (k != null) {
onlyFirstWord = Boolean.valueOf(k);
}
k = args.get( FORCE_FIRST_LETTER );
if( k != null ) {
forceFirstLetter = Boolean.valueOf( k );
k = args.get(FORCE_FIRST_LETTER);
if (k != null) {
forceFirstLetter = Boolean.valueOf(k);
}
}
public String processWord( String w, int wordCount )
{
if( w.length() < 1 ) {
return w;
public void processWord(char[] buffer, int offset, int length, int wordCount) {
if (length < 1) {
return;
}
if( onlyFirstWord && wordCount > 0 ) {
return w.toLowerCase();
}
String k = keep.get( w.toUpperCase() );
if( k != null ) {
if( wordCount == 0 && forceFirstLetter && Character.isLowerCase( k.charAt(0) ) ) {
return Character.toUpperCase( k.charAt(0) ) + k.substring( 1 );
if (onlyFirstWord && wordCount > 0) {
for (int i = 0; i < length; i++) {
buffer[offset + i] = Character.toLowerCase(buffer[offset + i]);
}
return k;
return;
}
if( w.length() < minWordLength ) {
return w;
if (keep.contains(buffer, offset, length)) {
if (wordCount == 0 && forceFirstLetter) {
buffer[offset] = Character.toUpperCase(buffer[offset]);
}
return;
}
for( String prefix : okPrefix ) {
if( w.startsWith( prefix ) ) {
return w;
if (length < minWordLength) {
return;
}
for (char[] prefix : okPrefix) {
if (length >= prefix.length) { //don't bother checking if the buffer length is less than the prefix
boolean match = true;
for (int i = 0; i < prefix.length; i++) {
if (prefix[i] != buffer[offset + i]) {
match = false;
break;
}
}
if (match == true) {
return;
}
}
}
// We know it has at least one character
char[] chars = w.toCharArray();
/*char[] chars = w.toCharArray();
StringBuilder word = new StringBuilder( w.length() );
word.append( Character.toUpperCase( chars[0] ) );
for( int i=1; i<chars.length; i++ ) {
word.append( Character.toLowerCase( chars[i] ) );
word.append( Character.toUpperCase( chars[0] ) );*/
buffer[offset] = Character.toUpperCase(buffer[offset]);
for (int i = 1; i < length; i++) {
buffer[offset + i] = Character.toLowerCase(buffer[offset + i]);
}
return word.toString();
//return word.toString();
}
public CapitalizationFilter create(TokenStream input) {
return new CapitalizationFilter(input,this);
return new CapitalizationFilter(input, this);
}
}
/**
* This relies on the Factory so that the difficult stuff does not need to be
* re-initialized each time the filter runs.
*
* <p/>
* This is package protected since it is not useful without the Factory
*/
class CapitalizationFilter extends TokenFilter
{
class CapitalizationFilter extends TokenFilter {
protected final CapitalizationFilterFactory factory;
public CapitalizationFilter(TokenStream in, final CapitalizationFilterFactory factory ) {
public CapitalizationFilter(TokenStream in, final CapitalizationFilterFactory factory) {
super(in);
this.factory = factory;
}
@Override
public final Token next() throws IOException {
Token t = input.next();
if( t != null ) {
String s = t.termText();
if( s.length() < factory.maxTokenLength ) {
public Token next(Token token) throws IOException {
Token t = input.next(token);
if (t != null) {
char[] termBuffer = t.termBuffer();
int termBufferLength = t.termLength();
char[] backup = null;
if (factory.maxWordCount < CapitalizationFilterFactory.DEFAULT_MAX_WORD_COUNT) {
//make a backup in case we exceed the word count
System.arraycopy(termBuffer, 0, backup, 0, termBufferLength);
}
if (termBuffer.length < factory.maxTokenLength) {
int wordCount = 0;
StringBuilder word = new StringBuilder( s.length() );
StringBuilder text = new StringBuilder( s.length() );
for( char c : s.toCharArray() ) {
if( c <= ' ' || c == '.' ) {
if( word.length() > 0 ) {
text.append( factory.processWord( word.toString(), wordCount++ ) );
word.setLength( 0 );
int lastWordStart = 0;
for (int i = 0; i < termBufferLength; i++) {
char c = termBuffer[i];
if (c <= ' ' || c == '.') {
int len = i - lastWordStart;
if (len > 0) {
factory.processWord(termBuffer, lastWordStart, len, wordCount++);
lastWordStart = i + 1;
i++;
}
text.append( c );
}
else {
word.append( c );
}
}
// Add the last word
if( word.length() > 0 ) {
text.append( factory.processWord( word.toString(), wordCount++ ) );
// process the last word
if (lastWordStart < termBuffer.length) {
factory.processWord(termBuffer, lastWordStart, termBuffer.length - lastWordStart, wordCount++);
}
if( wordCount <= factory.maxWordCount ) {
t.setTermText( text.toString() );
if (wordCount > factory.maxWordCount) {
t.setTermBuffer(backup, 0, termBufferLength);
}
}
}
return t;
}
}

51
src/test/org/apache/solr/analysis/TestCapitalizationFilter.java
View File

@ -34,14 +34,33 @@ public class TestCapitalizationFilter extends BaseTokenTestCase {
CapitalizationFilterFactory factory = new CapitalizationFilterFactory();
factory.init( args );
char[] termBuffer;
termBuffer = "kiTTEN".toCharArray();
factory.processWord(termBuffer, 0, termBuffer.length, 0 );
assertEquals( "Kitten", new String(termBuffer, 0, termBuffer.length));
assertEquals( "Kitten", factory.processWord( "kiTTEN", 0 ) );
factory.forceFirstLetter = true;
assertEquals( "And", factory.processWord( "AnD", 0 ) ); // first is forced
termBuffer = "and".toCharArray();
factory.processWord(termBuffer, 0, termBuffer.length, 0 );
assertEquals( "And", new String(termBuffer, 0, termBuffer.length));//first is forced
termBuffer = "AnD".toCharArray();
factory.processWord(termBuffer, 0, termBuffer.length, 0 );
assertEquals( "And", new String(termBuffer, 0, termBuffer.length));//first is forced, but it's not a keep word, either
factory.forceFirstLetter = false;
assertEquals( "and", factory.processWord( "AnD", 0 ) ); // first is forced
termBuffer = "AnD".toCharArray();
factory.processWord(termBuffer, 0, termBuffer.length, 0 );
assertEquals( "And", new String(termBuffer, 0, termBuffer.length)); //first is not forced, but it's not a keep word, either
factory.forceFirstLetter = true;
assertEquals( "BIG", factory.processWord( "big", 0 ) );
termBuffer = "big".toCharArray();
factory.processWord(termBuffer, 0, termBuffer.length, 0 );
assertEquals( "Big", new String(termBuffer, 0, termBuffer.length));
termBuffer = "BIG".toCharArray();
factory.processWord(termBuffer, 0, termBuffer.length, 0 );
assertEquals( "BIG", new String(termBuffer, 0, termBuffer.length));
String out = tsToString( factory.create( new IterTokenStream( "Hello thEre my Name is Ryan" ) ) );
assertEquals( "Hello there my name is ryan", out );
@ -74,7 +93,27 @@ public class TestCapitalizationFilter extends BaseTokenTestCase {
assertEquals( "1st 2nd Third", out );
factory.forceFirstLetter = true;
out = tsToString( factory.create( new IterTokenStream( "the The" ) ) );
assertEquals( "The the", out );
out = tsToString( factory.create( new IterTokenStream( "the The the" ) ) );
assertEquals( "The The the", out );
}
public void testKeepIgnoreCase() throws Exception {
Map<String,String> args = new HashMap<String, String>();
args.put( CapitalizationFilterFactory.KEEP, "kitten" );
args.put( CapitalizationFilterFactory.KEEP_IGNORE_CASE, "true" );
args.put( CapitalizationFilterFactory.ONLY_FIRST_WORD, "true" );
CapitalizationFilterFactory factory = new CapitalizationFilterFactory();
factory.init( args );
char[] termBuffer;
termBuffer = "kiTTEN".toCharArray();
factory.forceFirstLetter = true;
factory.processWord(termBuffer, 0, termBuffer.length, 0 );
assertEquals( "KiTTEN", new String(termBuffer, 0, termBuffer.length));
factory.forceFirstLetter = false;
termBuffer = "kiTTEN".toCharArray();
factory.processWord(termBuffer, 0, termBuffer.length, 0 );
assertEquals( "kiTTEN", new String(termBuffer, 0, termBuffer.length));
}
}