org.apache.commons.codec.language.bm

Class Rule

java.lang.Object

org.apache.commons.codec.language.bm.Rule

public class Rule
extends Object

A phoneme rule.

Rules have a pattern, left context, right context, output phoneme, set of languages for which they apply and a logical flag indicating if all languages must be in play. A rule matches if:

• the pattern matches at the current position
• the string up until the beginning of the pattern matches the left context
• the string from the end of the pattern matches the right context
• logical is ALL and all languages are in scope; or
• logical is any other value and at least one language is in scope

Rules are typically generated by parsing rules resources. In normal use, there will be no need for the user to explicitly construct their own.

Rules are immutable and thread-safe.

Rules resources

Rules are typically loaded from resource files. These are UTF-8 encoded text files. They are systematically named following the pattern:

org/apache/commons/codec/language/bm/${NameType#getName}_${RuleType#getName}_${language}.txt

The format of these resources is the following:

• Rules: whitespace separated, double-quoted strings. There should be 4 columns to each row, and these will be interpreted as:
  1. pattern
  2. left context
  3. right context
  4. phoneme
• End-of-line comments: Any occurrence of '//' will cause all text following on that line to be discarded as a comment.
• Multi-line comments: Any line starting with '/*' will start multi-line commenting mode. This will skip all content until a line ending in '*' and '/' is found.
• Blank lines: All blank lines will be skipped.

Since:

1.6

Version:

$Id: Rule.java 1544532 2013-11-22 13:38:35Z ggregory $

Nested Class Summary

Nested Classes

Modifier and Type	Class and Description
static class	Rule.Phoneme
static interface	Rule.PhonemeExpr
static class	Rule.PhonemeList
static interface	Rule.RPattern A minimal wrapper around the functionality of Pattern that we use, to allow for alternate implementations.

Field Summary

Fields

Modifier and Type	Field and Description
static String	ALL
static Rule.RPattern	ALL_STRINGS_RMATCHER

Constructor Summary

Constructors

Constructor and Description
Rule(String pattern, String lContext, String rContext, Rule.PhonemeExpr phoneme) Creates a new rule.

Method Summary

Methods

Modifier and Type	Method and Description
static List<Rule>	getInstance(NameType nameType, RuleType rt, Languages.LanguageSet langs) Gets rules for a combination of name type, rule type and languages.
static List<Rule>	getInstance(NameType nameType, RuleType rt, String lang) Gets rules for a combination of name type, rule type and a single language.
static Map<String,List<Rule>>	getInstanceMap(NameType nameType, RuleType rt, Languages.LanguageSet langs) Gets rules for a combination of name type, rule type and languages.
static Map<String,List<Rule>>	getInstanceMap(NameType nameType, RuleType rt, String lang) Gets rules for a combination of name type, rule type and a single language.
Rule.RPattern	getLContext() Gets the left context.
String	getPattern() Gets the pattern.
Rule.PhonemeExpr	getPhoneme() Gets the phoneme.
Rule.RPattern	getRContext() Gets the right context.
boolean	patternAndContextMatches(CharSequence input, int i) Decides if the pattern and context match the input starting at a position.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Field Detail

ALL_STRINGS_RMATCHER

public static final Rule.RPattern ALL_STRINGS_RMATCHER

ALL

public static final String ALL

See Also:

Constant Field Values

Constructor Detail

Rule

public Rule(String pattern,
    String lContext,
    String rContext,
    Rule.PhonemeExpr phoneme)

Creates a new rule.

Parameters:

pattern - the pattern

lContext - the left context

rContext - the right context

phoneme - the resulting phoneme

Method Detail

getInstance

public static List<Rule> getInstance(NameType nameType,
                     RuleType rt,
                     Languages.LanguageSet langs)

Gets rules for a combination of name type, rule type and languages.

Parameters:

nameType - the NameType to consider

rt - the RuleType to consider

langs - the set of languages to consider

Returns:

a list of Rules that apply

getInstance

public static List<Rule> getInstance(NameType nameType,
                     RuleType rt,
                     String lang)

Gets rules for a combination of name type, rule type and a single language.

Parameters:

nameType - the NameType to consider

rt - the RuleType to consider

lang - the language to consider

Returns:

a list of Rules that apply

getInstanceMap

public static Map<String,List<Rule>> getInstanceMap(NameType nameType,
                                    RuleType rt,
                                    Languages.LanguageSet langs)

Gets rules for a combination of name type, rule type and languages.

Parameters:

nameType - the NameType to consider

rt - the RuleType to consider

langs - the set of languages to consider

Returns:

a map containing all Rules that apply, grouped by the first character of the rule pattern

Since:

1.9

getInstanceMap

public static Map<String,List<Rule>> getInstanceMap(NameType nameType,
                                    RuleType rt,
                                    String lang)

Gets rules for a combination of name type, rule type and a single language.

Parameters:

nameType - the NameType to consider

rt - the RuleType to consider

lang - the language to consider

Returns:

a map containing all Rules that apply, grouped by the first character of the rule pattern

Since:

1.9

getLContext

public Rule.RPattern getLContext()

Gets the left context. This is a regular expression that must match to the left of the pattern.

Returns:

the left context Pattern

getPattern

public String getPattern()

Gets the pattern. This is a string-literal that must exactly match.

Returns:

the pattern

getPhoneme

public Rule.PhonemeExpr getPhoneme()

Gets the phoneme. If the rule matches, this is the phoneme associated with the pattern match.

Returns:

the phoneme

getRContext

public Rule.RPattern getRContext()

Gets the right context. This is a regular expression that must match to the right of the pattern.

Returns:

the right context Pattern

patternAndContextMatches

public boolean patternAndContextMatches(CharSequence input,
                               int i)

Decides if the pattern and context match the input starting at a position. It is a match if the lContext matches input up to i, pattern matches at i and rContext matches from the end of the match of pattern to the end of input.

Parameters:

input - the input String

i - the int position within the input

Returns:

true if the pattern and left/right context match, false otherwise