com.swabunga.spell.engine

Class SpellDictionaryDisk

public class SpellDictionaryDisk extends SpellDictionaryASpell

An implementation of SpellDictionary that doesn't cache any words in memory. Avoids the huge footprint of SpellDictionaryHashMap at the cost of relatively minor latency. A future version of this class that implements some caching strategies might be a good idea in the future, if there's any demand for it.

This class makes use of the "classic" Java IO library (java.io). However, it could probably benefit from the new IO APIs (java.nio) and it is anticipated that a future version of this class, probably called SpellDictionaryDiskNIO will appear at some point.

Since: 0.5

Version: 0.1

Author: Ben Galbraith (ben@galbraiths.org)

Field Summary
protected booleanready
The flag indicating if the initial preparation or loading of the on disk dictionary is complete.
Constructor Summary
SpellDictionaryDisk(File base, File phonetic, boolean block)
Construct a spell dictionary on disk.
Method Summary
voidaddWord(String word)
Adds another word to the dictionary.
protected voidbuildNewDictionaryDatabase()
Builds the file words database file and the contents file for the on disk dictionary.
ListgetWords(String code)
Returns a list of words that have the same phonetic code.
booleanisReady()
Indicates if the initial preparation or loading of the on disk dictionary is complete.
protected voidloadIndex()
Loads the index file from disk.

Field Detail

ready

protected boolean ready
The flag indicating if the initial preparation or loading of the on disk dictionary is complete.

Constructor Detail

SpellDictionaryDisk

public SpellDictionaryDisk(File base, File phonetic, boolean block)
Construct a spell dictionary on disk. The spell dictionary is created from words list(s) contained in file(s). A words list file is a file with one word per line. Words list files are located in a base/words dictionary where base is the path to words dictionary. The on disk spell dictionary is created in base/db dictionary and contains files: The contents file has a list of filename, size indicating the name and length of each files in the base/words dictionary. If one of theses files was changed, added or deleted before the call to the constructor, the process of producing new or updated words.db and words.idx files is started again.

The spellchecking process is then worked upon the words.db and words.idx files.

NOTE: Do *not* create two instances of this class pointing to the same base unless you are sure that a new dictionary does not have to be created. In the future, some sort of external locking mechanism may be created that handles this scenario gracefully.

Parameters: base the base directory in which SpellDictionaryDisk can expect to find its necessary files. phonetic the phonetic file used by the spellchecker. block if a new word db needs to be created, there can be a considerable delay before the constructor returns. If block is true, this method will block while the db is created and return when done. If block is false, this method will create a thread to create the new dictionary and return immediately.

Throws: java.io.FileNotFoundException indicates problems locating the files on the system java.io.IOException indicates problems reading the files

Method Detail

addWord

public void addWord(String word)
Adds another word to the dictionary. This method is not yet implemented for this class.

Parameters: word The word to add.

buildNewDictionaryDatabase

protected void buildNewDictionaryDatabase()
Builds the file words database file and the contents file for the on disk dictionary.

getWords

public List getWords(String code)
Returns a list of words that have the same phonetic code.

Parameters: code The phonetic code common to the list of words

Returns: A list of words having the same phonetic code

isReady

public boolean isReady()
Indicates if the initial preparation or loading of the on disk dictionary is complete.

Returns: the indication that the dictionary initial setup is done.

loadIndex

protected void loadIndex()
Loads the index file from disk. The index file accelerates words lookup into the dictionary db file.