/**
* Retreives the Soundex code for a given String object.
*
* @param str String to encode using the Soundex algorithm
* @return A soundex code for the String supplied
* @throws IllegalArgumentException if a character is not mapped
*/
@DSGenerator(
tool_name = "Doppelganger",
tool_version = "2.0",
generated_on = "2014-09-03 14:59:48.718 -0400",
hash_original_method = "0EBBB65C00DFAFB7B6E0400F5755712C",
hash_generated_method = "94B188BA634D9793A5790601BA8B7F45")
public String soundex(String str) {
if (str == null) {
return null;
}
str = SoundexUtils.clean(str);
if (str.length() == 0) {
return str;
}
char out[] = {'0', '0', '0', '0'};
char last, mapped;
int incount = 1, count = 1;
out[0] = str.charAt(0);
last = getMappingCode(str, 0);
while ((incount < str.length()) && (count < out.length)) {
mapped = getMappingCode(str, incount++);
if (mapped != 0) {
if ((mapped != '0') && (mapped != last)) {
out[count++] = mapped;
}
last = mapped;
}
}
return new String(out);
}
/**
* Retreives the Soundex code for a given String object.
*
* @param str String to encode using the Soundex algorithm
* @return A soundex code for the String supplied
* @throws IllegalArgumentException if a character is not mapped
*/
public String soundex(String str) {
if (str == null) {
return null;
}
str = SoundexUtils.clean(str);
if (str.length() == 0) {
return str;
}
char out[] = {'0', '0', '0', '0'};
char last, mapped;
int incount = 1, count = 1;
out[0] = str.charAt(0);
// getMappingCode() throws IllegalArgumentException
last = getMappingCode(str, 0);
while ((incount < str.length()) && (count < out.length)) {
mapped = getMappingCode(str, incount++);
if (mapped != 0) {
if ((mapped != '0') && (mapped != last)) {
out[count++] = mapped;
}
last = mapped;
}
}
return new String(out);
}
/**
* Encodes the Strings and returns the number of characters in the two encoded Strings that are
* the same. This return value ranges from 0 through 4: 0 indicates little or no similarity, and 4
* indicates strong similarity or identical values.
*
* @param s1 A String that will be encoded and compared.
* @param s2 A String that will be encoded and compared.
* @return The number of characters in the two encoded Strings that are the same from 0 to 4.
* @see <a
* href="http://msdn.microsoft.com/library/default.asp?url=/library/en-us/tsqlref/ts_de-dz_8co5.asp">
* MS T-SQL DIFFERENCE </a>
* @throws EncoderException if an error occurs encoding one of the strings
* @since 1.3
*/
@DSGenerator(
tool_name = "Doppelganger",
tool_version = "2.0",
generated_on = "2014-09-03 14:59:48.711 -0400",
hash_original_method = "9372A1D570FF0E071E2A73F0471280C8",
hash_generated_method = "73E7D5E07B73B540EF5D01C579CF6548")
public int difference(String s1, String s2) throws EncoderException {
return SoundexUtils.difference(this, s1, s2);
}
public String nysiis(String str) {
if (str == null) {
return null;
}
str = SoundexUtils.clean(str);
if (str.length() == 0) {
return str;
}
str = PAT_MAC.matcher(str).replaceFirst("MCC");
str = PAT_KN.matcher(str).replaceFirst("NN");
str = PAT_K.matcher(str).replaceFirst("C");
str = PAT_PH_PF.matcher(str).replaceFirst("FF");
str = PAT_SCH.matcher(str).replaceFirst("SSS");
str = PAT_EE_IE.matcher(str).replaceFirst("Y");
str = PAT_DT_ETC.matcher(str).replaceFirst("D");
StringBuilder key = new StringBuilder(str.length());
key.append(str.charAt(0));
char[] chars = str.toCharArray();
int len = chars.length;
for (int i = 1; i < len; i++) {
char next = i < len - 1 ? chars[(i + 1)] : ' ';
char aNext = i < len - 2 ? chars[(i + 2)] : ' ';
char[] transcoded = transcodeRemaining(chars[(i - 1)], chars[i], next, aNext);
System.arraycopy(transcoded, 0, chars, i, transcoded.length);
if (chars[i] != chars[(i - 1)]) {
key.append(chars[i]);
}
}
if (key.length() > 1) {
char lastChar = key.charAt(key.length() - 1);
if (lastChar == 'S') {
key.deleteCharAt(key.length() - 1);
lastChar = key.charAt(key.length() - 1);
}
if (key.length() > 2) {
char last2Char = key.charAt(key.length() - 2);
if ((last2Char == 'A') && (lastChar == 'Y')) {
key.deleteCharAt(key.length() - 2);
}
}
if (lastChar == 'A') {
key.deleteCharAt(key.length() - 1);
}
}
String string = key.toString();
return isStrict() ? string.substring(0, Math.min(6, string.length())) : string;
}
/**
* Retrieves the NYSIIS code for a given String object.
*
* @param str String to encode using the NYSIIS algorithm
* @return A NYSIIS code for the String supplied
*/
public String nysiis(String str) {
if (str == null) {
return null;
}
// Use the same clean rules as Soundex
str = SoundexUtils.clean(str);
if (str.length() == 0) {
return str;
}
// Translate first characters of name:
// MAC -> MCC, KN -> NN, K -> C, PH | PF -> FF, SCH -> SSS
str = PAT_MAC.matcher(str).replaceFirst("MCC");
str = PAT_KN.matcher(str).replaceFirst("NN");
str = PAT_K.matcher(str).replaceFirst("C");
str = PAT_PH_PF.matcher(str).replaceFirst("FF");
str = PAT_SCH.matcher(str).replaceFirst("SSS");
// Translate last characters of name:
// EE -> Y, IE -> Y, DT | RT | RD | NT | ND -> D
str = PAT_EE_IE.matcher(str).replaceFirst("Y");
str = PAT_DT_ETC.matcher(str).replaceFirst("D");
// First character of key = first character of name.
StringBuilder key = new StringBuilder(str.length());
key.append(str.charAt(0));
// Transcode remaining characters, incrementing by one character each
// time
final char[] chars = str.toCharArray();
final int len = chars.length;
for (int i = 1; i < len; i++) {
final char next = i < len - 1 ? chars[i + 1] : SPACE;
final char aNext = i < len - 2 ? chars[i + 2] : SPACE;
final char[] transcoded = transcodeRemaining(chars[i - 1], chars[i], next, aNext);
System.arraycopy(transcoded, 0, chars, i, transcoded.length);
// only append the current char to the key if it is different from
// the last one
if (chars[i] != chars[i - 1]) {
key.append(chars[i]);
}
}
if (key.length() > 1) {
char lastChar = key.charAt(key.length() - 1);
// If last character is S, remove it.
if (lastChar == 'S') {
key.deleteCharAt(key.length() - 1);
lastChar = key.charAt(key.length() - 1);
}
if (key.length() > 2) {
final char last2Char = key.charAt(key.length() - 2);
// If last characters are AY, replace with Y.
if (last2Char == 'A' && lastChar == 'Y') {
key.deleteCharAt(key.length() - 2);
}
}
// If last character is A, remove it.
if (lastChar == 'A') {
key.deleteCharAt(key.length() - 1);
}
}
final String string = key.toString();
return this.isStrict() ? string.substring(0, Math.min(TRUE_LENGTH, string.length())) : string;
}
/**
* Encodes the Strings and returns the number of characters in the two encoded Strings that are
* the same. This return value ranges from 0 through 4: 0 indicates little or no similarity, and 4
* indicates strong similarity or identical values.
*
* @param s1 A String that will be encoded and compared.
* @param s2 A String that will be encoded and compared.
* @return The number of characters in the two encoded Strings that are the same from 0 to 4.
* @see SoundexUtils#difference(StringEncoder,String,String)
* @see <a
* href="http://msdn.microsoft.com/library/default.asp?url=/library/en-us/tsqlref/ts_de-dz_8co5.asp">
* MS T-SQL DIFFERENCE </a>
* @throws EncoderException if an error occurs encoding one of the strings
* @since 1.3
*/
public int difference(String s1, String s2) throws EncoderException {
return SoundexUtils.difference(this, s1, s2);
}
