eu.etrap.medusa.uima.analysis_engine

Class StringToNumberMapping

java.lang.Object

eu.etrap.medusa.uima.analysis_engine.StringToNumberMapping

public class StringToNumberMapping
extends java.lang.Object

This class manages word numbers.

Author:

lp

See Also:

WordNumberAnnotator

Field Summary

Fields

Modifier and Type	Field and Description
protected bak.pcj.map.ObjectKeyIntMap	map maps word --> count
protected int	nextNumber next free number for a word

Constructor Summary

Constructors

Constructor and Description
StringToNumberMapping()

Method Summary

Modifier and Type	Method and Description
void	clearMap() clears all mappings
void	load(java.lang.String fileName, java.lang.String encoding, boolean frequencyFormat) clears and then populates the wordnumbers using the given file.
int	stringToNumber(java.lang.String s) maps the given String s to a wordnumber.

Methods inherited from class java.lang.Object

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

Field Detail

map

protected bak.pcj.map.ObjectKeyIntMap map

maps word --> count

nextNumber

protected int nextNumber

next free number for a word

Constructor Detail

StringToNumberMapping

public StringToNumberMapping()

Method Detail

clearMap

public void clearMap()

clears all mappings

stringToNumber

public int stringToNumber(java.lang.String s)

maps the given String s to a wordnumber. If no mapping is defined already a new one is generated.

Parameters:

s - word to generate wordnumber for

Returns:

wordnumber for s

load

public void load(java.lang.String fileName,
                 java.lang.String encoding,
                 boolean frequencyFormat)
          throws java.io.FileNotFoundException,
                 java.io.UnsupportedEncodingException,
                 java.util.InputMismatchException

clears and then populates the wordnumbers using the given file. Each line contains one word and the first line will get the lowest wordnumber. Two input formats are supported. For the first input format (frequencyFormat=false) each line contains only a word. For the second input format (frequencyFormat=true) each line contains several columns which are separated by TAB-characters and the word is given in the second column. This second input format can be used for output of the WncConsumer

Parameters:

fileName - name of the file to load

encoding - character encoding of the file

frequencyFormat - use second column of multicolumn input?

Throws:

java.io.FileNotFoundException - thrown if the given file can not be found

java.io.UnsupportedEncodingException - thrown if the given character encoding is not supported

java.util.InputMismatchException - throw in case the multicolumn format is not used correctly