Home | History | Annotate | Download | only in latin
      1 /*
      2  * Copyright (C) 2008 The Android Open Source Project
      3  *
      4  * Licensed under the Apache License, Version 2.0 (the "License"); you may not
      5  * use this file except in compliance with the License. You may obtain a copy of
      6  * the License at
      7  *
      8  * http://www.apache.org/licenses/LICENSE-2.0
      9  *
     10  * Unless required by applicable law or agreed to in writing, software
     11  * distributed under the License is distributed on an "AS IS" BASIS, WITHOUT
     12  * WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the
     13  * License for the specific language governing permissions and limitations under
     14  * the License.
     15  */
     16 
     17 package com.android.inputmethod.latin;
     18 
     19 /**
     20  * Abstract base class for a dictionary that can do a fuzzy search for words based on a set of key
     21  * strokes.
     22  */
     23 abstract public class Dictionary {
     24     /**
     25      * Whether or not to replicate the typed word in the suggested list, even if it's valid.
     26      */
     27     protected static final boolean INCLUDE_TYPED_WORD_IF_VALID = false;
     28 
     29     /**
     30      * The weight to give to a word if it's length is the same as the number of typed characters.
     31      */
     32     protected static final int FULL_WORD_FREQ_MULTIPLIER = 2;
     33 
     34     public static enum DataType {
     35         UNIGRAM, BIGRAM
     36     }
     37 
     38     /**
     39      * Interface to be implemented by classes requesting words to be fetched from the dictionary.
     40      * @see #getWords(WordComposer, WordCallback)
     41      */
     42     public interface WordCallback {
     43         /**
     44          * Adds a word to a list of suggestions. The word is expected to be ordered based on
     45          * the provided frequency.
     46          * @param word the character array containing the word
     47          * @param wordOffset starting offset of the word in the character array
     48          * @param wordLength length of valid characters in the character array
     49          * @param frequency the frequency of occurence. This is normalized between 1 and 255, but
     50          * can exceed those limits
     51          * @param dicTypeId of the dictionary where word was from
     52          * @param dataType tells type of this data
     53          * @return true if the word was added, false if no more words are required
     54          */
     55         boolean addWord(char[] word, int wordOffset, int wordLength, int frequency, int dicTypeId,
     56                 DataType dataType);
     57     }
     58 
     59     /**
     60      * Searches for words in the dictionary that match the characters in the composer. Matched
     61      * words are added through the callback object.
     62      * @param composer the key sequence to match
     63      * @param callback the callback object to send matched words to as possible candidates
     64      * @param nextLettersFrequencies array of frequencies of next letters that could follow the
     65      *        word so far. For instance, "bracke" can be followed by "t", so array['t'] will have
     66      *        a non-zero value on returning from this method.
     67      *        Pass in null if you don't want the dictionary to look up next letters.
     68      * @see WordCallback#addWord(char[], int, int)
     69      */
     70     abstract public void getWords(final WordComposer composer, final WordCallback callback,
     71             int[] nextLettersFrequencies);
     72 
     73     /**
     74      * Searches for pairs in the bigram dictionary that matches the previous word and all the
     75      * possible words following are added through the callback object.
     76      * @param composer the key sequence to match
     77      * @param callback the callback object to send possible word following previous word
     78      * @param nextLettersFrequencies array of frequencies of next letters that could follow the
     79      *        word so far. For instance, "bracke" can be followed by "t", so array['t'] will have
     80      *        a non-zero value on returning from this method.
     81      *        Pass in null if you don't want the dictionary to look up next letters.
     82      */
     83     public void getBigrams(final WordComposer composer, final CharSequence previousWord,
     84             final WordCallback callback, int[] nextLettersFrequencies) {
     85         // empty base implementation
     86     }
     87 
     88     /**
     89      * Checks if the given word occurs in the dictionary
     90      * @param word the word to search for. The search should be case-insensitive.
     91      * @return true if the word exists, false otherwise
     92      */
     93     abstract public boolean isValidWord(CharSequence word);
     94 
     95     /**
     96      * Compares the contents of the character array with the typed word and returns true if they
     97      * are the same.
     98      * @param word the array of characters that make up the word
     99      * @param length the number of valid characters in the character array
    100      * @param typedWord the word to compare with
    101      * @return true if they are the same, false otherwise.
    102      */
    103     protected boolean same(final char[] word, final int length, final CharSequence typedWord) {
    104         if (typedWord.length() != length) {
    105             return false;
    106         }
    107         for (int i = 0; i < length; i++) {
    108             if (word[i] != typedWord.charAt(i)) {
    109                 return false;
    110             }
    111         }
    112         return true;
    113     }
    114 
    115     /**
    116      * Override to clean up any resources.
    117      */
    118     public void close() {
    119     }
    120 }
    121