[CODEC-125] Implement a Beider-Morse phonetic matching codec. Applied patch https://issues.apache.org/jira/secure/attachment/12489755/fightingMemoryChurn.patch

git-svn-id: https://svn.apache.org/repos/asf/commons/proper/codec/trunk@1155135 13f79535-47bb-0310-9956-ffa450edef68

Author: garydgregory

src/java/org/apache/commons/codec/language/bm/Lang.java

@@ -24,6 +24,7 @@
 import java.util.EnumMap;
 import java.util.HashSet;
 import java.util.List;
+import java.util.Locale;
 import java.util.Map;
 import java.util.Scanner;
 import java.util.Set;
@@ -71,7 +72,7 @@
  */
 public class Lang {
 
-    private static class LangRule {
+    private static final class LangRule {
         private final boolean acceptOnMatch;
         private final Set<String> languages;
         private final Pattern pattern;
@@ -199,7 +200,7 @@ private Lang(List<LangRule> rules, Languages languages) {
      */
     public String guessLanguage(String text) {
         Languages.LanguageSet ls = guessLanguages(text);
-        return ls.isSingleton() ? ls.getAny() : Languages.ANY; 
+        return ls.isSingleton() ? ls.getAny() : Languages.ANY;
     }
 
     /**
@@ -210,7 +211,7 @@ public String guessLanguage(String text) {
      * @return a Set of Strings of language names that are potential matches for the input word
      */
     public Languages.LanguageSet guessLanguages(String input) {
-        String text = input.toLowerCase(); // todo: locale?
+        String text = input.toLowerCase(Locale.ENGLISH);
         // System.out.println("Testing text: '" + text + "'");
 
         Set<String> langs = new HashSet<String>(this.languages.getLanguages());

src/java/org/apache/commons/codec/language/bm/Languages.java

@@ -58,9 +58,9 @@ public class Languages {
      * A set of languages.
      */
     public static abstract class LanguageSet {
-        
+
         public static LanguageSet from(Set<String> langs) {
-            return langs.isEmpty() ? NO_LANGUAGES : new SomeLanguages(langs);  
+            return langs.isEmpty() ? NO_LANGUAGES : new SomeLanguages(langs);
         }
 
         public abstract boolean contains(String language);
@@ -77,7 +77,7 @@ public static LanguageSet from(Set<String> langs) {
     /**
      * Some languages, explicitly enumerated.
      */
-    public static class SomeLanguages extends LanguageSet {
+    public static final class SomeLanguages extends LanguageSet {
         private final Set<String> languages;
 
         private SomeLanguages(Set<String> languages) {
@@ -116,9 +116,13 @@ public LanguageSet restrictTo(LanguageSet other) {
                 return this;
             } else {
                 SomeLanguages sl = (SomeLanguages) other;
-                Set<String> ls = new HashSet<String>(this.languages);
-                ls.retainAll(sl.languages);
-                return from(ls);
+                if (sl.languages.containsAll(languages)) {
+                    return this;
+                } else {
+                    Set<String> ls = new HashSet<String>(this.languages);
+                    ls.retainAll(sl.languages);
+                    return from(ls);
+                }
             }
         }
 

src/java/org/apache/commons/codec/language/bm/NameType.java

@@ -24,13 +24,13 @@
  * @since 2.0
  */
 public enum NameType {
-    
+
     /** Ashkenazi family names */
     ASHKENAZI("ash"),
-    
+
     /** Generic names and words */
     GENERIC("gen"),
-    
+
     /** Sephardic family names */
     SEPHARDIC("sep");
 

src/java/org/apache/commons/codec/language/bm/PhoneticEngine.java

@@ -51,7 +51,7 @@
  */
 public class PhoneticEngine {
 
-    static class PhonemeBuilder {
+    static final class PhonemeBuilder {
 
         public static PhonemeBuilder empty(Languages.LanguageSet languages) {
             return new PhonemeBuilder(Collections.singleton(new Rule.Phoneme("", languages)));
@@ -108,7 +108,7 @@ public String makeString() {
         }
     }
 
-    private static class RulesApplication {
+    private static final class RulesApplication {
         private final List<Rule> finalRules;
         private final CharSequence input;
 
@@ -176,6 +176,32 @@ public boolean isFound() {
                 "de la", "della", "des", "di", "do", "dos", "du", "van", "von"))));
     }
 
+    private static CharSequence cacheSubSequence(final CharSequence cached) {
+        // return cached;
+        final CharSequence[][] cache = new CharSequence[cached.length()][cached.length()];
+        return new CharSequence() {
+            public char charAt(int index) {
+                return cached.charAt(index);
+            }
+
+            public int length() {
+                return cached.length();
+            }
+
+            public CharSequence subSequence(int start, int end) {
+                if (start == end)
+                    return "";
+
+                CharSequence res = cache[start][end - 1];
+                if (res == null) {
+                    res = cached.subSequence(start, end);
+                    cache[start][end - 1] = res;
+                }
+                return res;
+            }
+        };
+    }
+
     private static String join(Iterable<String> strings, String sep) {
         StringBuilder sb = new StringBuilder();
         Iterator<String> si = strings.iterator();
@@ -229,7 +255,7 @@ private PhonemeBuilder applyFinalRules(PhonemeBuilder phonemeBuilder, List<Rule>
 
         for (Rule.Phoneme phoneme : phonemeBuilder.getPhonemes()) {
             PhonemeBuilder subBuilder = PhonemeBuilder.empty(phoneme.getLanguages());
-            CharSequence phonemeText = phoneme.getPhonemeText();
+            CharSequence phonemeText = cacheSubSequence(phoneme.getPhonemeText());
             // System.err.println("Expanding: " + phonemeText);
 
             for (int i = 0; i < phonemeText.length();) {
@@ -248,7 +274,7 @@ private PhonemeBuilder applyFinalRules(PhonemeBuilder phonemeBuilder, List<Rule>
             }
 
             // System.err.println("Expanded to: " + subBuilder.makeString());
-
+            // System.err.println("phenomes in collection of type: " + subBuilder.getPhonemes().getClass());
             phonemes.addAll(subBuilder.getPhonemes());
         }
 
@@ -345,8 +371,9 @@ public String encode(String input, final Languages.LanguageSet languageSet) {
         PhonemeBuilder phonemeBuilder = PhonemeBuilder.empty(languageSet);
 
         // loop over each char in the input - we will handle the increment manually
-        for (int i = 0; i < input.length();) {
-            RulesApplication rulesApplication = new RulesApplication(rules, input, phonemeBuilder, i).invoke();
+        CharSequence inputCache = cacheSubSequence(input);
+        for (int i = 0; i < inputCache.length();) {
+            RulesApplication rulesApplication = new RulesApplication(rules, inputCache, phonemeBuilder, i).invoke();
             i = rulesApplication.getI();
             phonemeBuilder = rulesApplication.getPhonemeBuilder();
             // System.err.println(input + " " + i + ": " + phonemeBuilder.makeString());