/// <summary>
/// Applies the final rules to convert from a language-specific phonetic representation to a
/// language-independent representation.
/// </summary>
/// <param name="phonemeBuilder">The current phonemes.</param>
/// <param name="finalRules">The final rules to apply.</param>
/// <returns>The resulting phonemes.</returns>
private PhonemeBuilder ApplyFinalRules(PhonemeBuilder phonemeBuilder,
IDictionary <string, IList <Rule> > finalRules)
{
if (finalRules == null)
{
throw new ArgumentNullException("finalRules can not be null");
}
if (finalRules.Count == 0)
{
return(phonemeBuilder);
}
ISet <Phoneme> phonemes = new SortedSet <Phoneme>(Phoneme.COMPARER);
foreach (Phoneme phoneme in phonemeBuilder.Phonemes)
{
PhonemeBuilder subBuilder = PhonemeBuilder.Empty(phoneme.Languages);
string phonemeText = phoneme.GetPhonemeText();
for (int i = 0; i < phonemeText.Length;)
{
RulesApplication rulesApplication =
new RulesApplication(finalRules, phonemeText, subBuilder, i, maxPhonemes).Invoke();
bool found = rulesApplication.IsFound;
subBuilder = rulesApplication.PhonemeBuilder;
if (!found)
{
// not found, appending as-is
subBuilder.Append(phonemeText.Substring(i, 1));
}
i = rulesApplication.I;
}
phonemes.UnionWith(subBuilder.Phonemes);
}
return(new PhonemeBuilder(phonemes.ToList()));
}
/// <summary>
/// Encodes an input string into an output phonetic representation, given a set of possible origin languages.
/// </summary>
/// <param name="input">String to phoneticise; a string with dashes or spaces separating each word.</param>
/// <param name="languageSet"></param>
/// <returns>A phonetic representation of the input; a string containing '-'-separated phonetic representations of the input.</returns>
public virtual string Encode(string input, LanguageSet languageSet)
{
IDictionary <string, IList <Rule> > rules = Rule.GetInstanceMap(this.nameType, RuleType.RULES, languageSet);
// rules common across many (all) languages
IDictionary <string, IList <Rule> > finalRules1 = Rule.GetInstanceMap(this.nameType, this.ruleType, "common");
// rules that apply to a specific language that may be ambiguous or wrong if applied to other languages
IDictionary <string, IList <Rule> > finalRules2 = Rule.GetInstanceMap(this.nameType, this.ruleType, languageSet);
// tidy the input
// lower case is a locale-dependent operation
input = input.ToLowerInvariant().Replace('-', ' ').Trim();
if (this.nameType == NameType.GENERIC)
{
if (input.Length >= 2 && input.Substring(0, 2 - 0).Equals("d'"))
{ // check for d'
string remainder = input.Substring(2);
string combined = "d" + remainder;
return("(" + Encode(remainder) + ")-(" + Encode(combined) + ")");
}
foreach (string l in NAME_PREFIXES[this.nameType])
{
// handle generic prefixes
if (input.StartsWith(l + " ", StringComparison.Ordinal))
{
// check for any prefix in the words list
string remainder = input.Substring(l.Length + 1); // input without the prefix
string combined = l + remainder; // input with prefix without space
return("(" + Encode(remainder) + ")-(" + Encode(combined) + ")");
}
}
}
IList <string> words = WHITESPACE.Split(input).ToList();
IList <string> words2 = new List <string>();
// special-case handling of word prefixes based upon the name type
switch (this.nameType)
{
case NameType.SEPHARDIC:
foreach (string aWord in words)
{
string[] parts = aWord.Split(new char[] { '\'' }, StringSplitOptions.RemoveEmptyEntries);
string lastPart = parts[parts.Length - 1];
words2.Add(lastPart);
}
words2.RemoveAll(NAME_PREFIXES[this.nameType]);
break;
case NameType.ASHKENAZI:
words2.AddRange(words);
words2.RemoveAll(NAME_PREFIXES[this.nameType]);
break;
case NameType.GENERIC:
words2.AddRange(words);
break;
default:
throw new InvalidOperationException("Unreachable case: " + this.nameType);
}
if (this.concat)
{
// concat mode enabled
input = Join(words2, " ");
}
else if (words2.Count == 1)
{
// not a multi-word name
//input = words.iterator().next();
input = words.FirstOrDefault();
}
else
{
// encode each word in a multi-word name separately (normally used for approx matches)
StringBuilder result = new StringBuilder();
foreach (string word in words2)
{
result.Append("-").Append(Encode(word));
}
// return the result without the leading "-"
return(result.ToString(1, result.Length - 1));
}
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, maxPhonemes).Invoke();
i = rulesApplication.I;
phonemeBuilder = rulesApplication.PhonemeBuilder;
}
// Apply the general rules
phonemeBuilder = ApplyFinalRules(phonemeBuilder, finalRules1);
// Apply the language-specific rules
phonemeBuilder = ApplyFinalRules(phonemeBuilder, finalRules2);
return(phonemeBuilder.MakeString());
}