de.susebox.jtopas.spi

Interface PatternHandler

All Known Subinterfaces:

DataMapper

All Known Implementing Classes:

PatternMatcher, StandardTokenizerProperties

public interface PatternHandler

This interface must be implemented by classes that should be used as a pattern handler for a Tokenizer. Pattern are usually regular expressions that are applied on token images to check if that image matches the pattern.

Author:

Heiko Blau

See Also:

Tokenizer, TokenizerProperties, DataMapper

Nested Class Summary

Nested Classes

Modifier and Type	Interface and Description
static interface	PatternHandler.Result An inner interface for the pattern match result.

Method Summary

Modifier and Type	Method and Description
boolean	hasPattern() This method can be used by a Tokenizer implementation for a fast detection if pattern matching must be performed at all.
PatternHandler.Result	matches(DataProvider dataProvider) This method checks if the start of a character range given through the DataProvider matches a pattern.

Method Detail

hasPattern

boolean hasPattern()

This method can be used by a Tokenizer implementation for a fast detection if pattern matching must be performed at all. If the method returns false time-consuming preparations can be skipped.

Returns:

true if there actually are pattern that can be tested for a match, false otherwise.

matches

PatternHandler.Result matches(DataProvider dataProvider)
                       throws TokenizerException,
                              java.lang.NullPointerException

This method checks if the start of a character range given through the DataProvider matches a pattern. An implementation should use a TokenizerException to report problems.
The method returns null if the beginning of the character range doesn't match a pattern known to the PatternHandler. Otherwise it returns an object with the implemented interface PatternHandler.Result.
The pattern check is repeated if the method returns a match that is exactly as long as the given data range and more data is available. Since it is probably a rare case, that where are not enough data to find a complete or no match, the overhead of a repeated check on part of the data is neglected.
If a pattern handler has more than one pattern that could be applied to the given data, it should return the longest possible match.

Parameters:

dataProvider - the source to get the data from

lengthOfMatch - if a match is found, the method places the length of it into the first element of this array

Returns:

a PatternHandler.Result object or null if no match was found

Throws:

TokenizerException - generic exception

java.lang.NullPointerException - if no DataProvider is given