Package org.apache.lucene.analysis.util

Class CharTokenizer

  • All Implemented Interfaces:
    Closeable, AutoCloseable
    Direct Known Subclasses:
    LetterTokenizer, UnicodeWhitespaceTokenizer, WhitespaceTokenizer
    public abstract class CharTokenizer
extends Tokenizer
    An abstract base class for simple, character-oriented tokenizers.

    The base class also provides factories to create instances of CharTokenizer using Java 8 lambdas or method references. It is possible to create an instance which behaves exactly like LetterTokenizer: 

     Tokenizer tok = CharTokenizer.fromTokenCharPredicate(Character::isLetter);

    • Constructor Detail

      • CharTokenizer

        public CharTokenizer()
        Creates a new CharTokenizer instance

      • CharTokenizer

        public CharTokenizer​(AttributeFactory factory,
                     int maxTokenLen)
        Creates a new CharTokenizer instance
        Parameters:
        factory - the attribute factory to use for this Tokenizer
        maxTokenLen - maximum token length the tokenizer will emit. Must be greater than 0 and less than MAX_TOKEN_LENGTH_LIMIT (1024*1024)
        Throws:
        IllegalArgumentException - if maxTokenLen is invalid.

    • Method Detail

      • fromTokenCharPredicate

        public static CharTokenizer fromTokenCharPredicate​(IntPredicate tokenCharPredicate)
        Creates a new instance of CharTokenizer using a custom predicate, supplied as method reference or lambda expression. The predicate should return true for all valid token characters.

        This factory is intended to be used with lambdas or method references. E.g., an elegant way to create an instance which behaves exactly as LetterTokenizer is: 

         Tokenizer tok = CharTokenizer.fromTokenCharPredicate(Character::isLetter);

      • fromTokenCharPredicate

        public static CharTokenizer fromTokenCharPredicate​(AttributeFactory factory,
                                                   IntPredicate tokenCharPredicate)
        Creates a new instance of CharTokenizer with the supplied attribute factory using a custom predicate, supplied as method reference or lambda expression. The predicate should return true for all valid token characters.

        This factory is intended to be used with lambdas or method references. E.g., an elegant way to create an instance which behaves exactly as LetterTokenizer is: 

         Tokenizer tok = CharTokenizer.fromTokenCharPredicate(factory, Character::isLetter);

      • fromSeparatorCharPredicate

        public static CharTokenizer fromSeparatorCharPredicate​(IntPredicate separatorCharPredicate)
        Creates a new instance of CharTokenizer using a custom predicate, supplied as method reference or lambda expression. The predicate should return true for all valid token separator characters. This method is provided for convenience to easily use predicates that are negated (they match the separator characters, not the token characters).

        This factory is intended to be used with lambdas or method references. E.g., an elegant way to create an instance which behaves exactly as WhitespaceTokenizer is: 

         Tokenizer tok = CharTokenizer.fromSeparatorCharPredicate(Character::isWhitespace);

      • fromSeparatorCharPredicate

        public static CharTokenizer fromSeparatorCharPredicate​(AttributeFactory factory,
                                                       IntPredicate separatorCharPredicate)
        Creates a new instance of CharTokenizer with the supplied attribute factory using a custom predicate, supplied as method reference or lambda expression. The predicate should return true for all valid token separator characters.

        This factory is intended to be used with lambdas or method references. E.g., an elegant way to create an instance which behaves exactly as WhitespaceTokenizer is: 

         Tokenizer tok = CharTokenizer.fromSeparatorCharPredicate(factory, Character::isWhitespace);

      • isTokenChar

        protected abstract boolean isTokenChar​(int c)
        Returns true iff a codepoint should be included in a token. This tokenizer generates as tokens adjacent sequences of codepoints which satisfy this predicate. Codepoints for which this is false are used to define token boundaries and are not included in tokens.