Home | History | Annotate | Download | only in makedict 
      1 /*
      2  * Copyright (C) 2011 The Android Open Source Project
      3  *
      4  * Licensed under the Apache License, Version 2.0 (the "License");
      5  * you may not use this file except in compliance with the License.
      6  * You may obtain a copy of 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,
     12  * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
     13  * See the License for the specific language governing permissions and
     14  * limitations under the License.
     15  */
     16 
     17 package com.android.inputmethod.latin.makedict;
     18 
     19 import com.android.inputmethod.annotations.UsedForTesting;
     20 import com.android.inputmethod.latin.BinaryDictionary;
     21 import com.android.inputmethod.latin.Dictionary;
     22 import com.android.inputmethod.latin.NgramContext;
     23 import com.android.inputmethod.latin.NgramContext.WordInfo;
     24 import com.android.inputmethod.latin.common.StringUtils;
     25 import com.android.inputmethod.latin.utils.CombinedFormatUtils;
     26 
     27 import java.util.ArrayList;
     28 import java.util.Arrays;
     29 
     30 import javax.annotation.Nullable;
     31 
     32 /**
     33  * Utility class for a word with a probability.
     34  *
     35  * This is chiefly used to iterate a dictionary.
     36  */
     37 public final class WordProperty implements Comparable<WordProperty> {
     38     public final String mWord;
     39     public final ProbabilityInfo mProbabilityInfo;
     40     public final ArrayList<NgramProperty> mNgrams;
     41     // TODO: Support mIsBeginningOfSentence.
     42     public final boolean mIsBeginningOfSentence;
     43     public final boolean mIsNotAWord;
     44     public final boolean mIsPossiblyOffensive;
     45     public final boolean mHasNgrams;
     46 
     47     private int mHashCode = 0;
     48 
     49     // TODO: Support n-gram.
     50     @UsedForTesting
     51     public WordProperty(final String word, final ProbabilityInfo probabilityInfo,
     52             @Nullable final ArrayList<WeightedString> bigrams,
     53             final boolean isNotAWord, final boolean isPossiblyOffensive) {
     54         mWord = word;
     55         mProbabilityInfo = probabilityInfo;
     56         if (null == bigrams) {
     57             mNgrams = null;
     58         } else {
     59             mNgrams = new ArrayList<>();
     60             final NgramContext ngramContext = new NgramContext(new WordInfo(mWord));
     61             for (final WeightedString bigramTarget : bigrams) {
     62                 mNgrams.add(new NgramProperty(bigramTarget, ngramContext));
     63             }
     64         }
     65         mIsBeginningOfSentence = false;
     66         mIsNotAWord = isNotAWord;
     67         mIsPossiblyOffensive = isPossiblyOffensive;
     68         mHasNgrams = bigrams != null && !bigrams.isEmpty();
     69     }
     70 
     71     private static ProbabilityInfo createProbabilityInfoFromArray(final int[] probabilityInfo) {
     72       return new ProbabilityInfo(
     73               probabilityInfo[BinaryDictionary.FORMAT_WORD_PROPERTY_PROBABILITY_INDEX],
     74               probabilityInfo[BinaryDictionary.FORMAT_WORD_PROPERTY_TIMESTAMP_INDEX],
     75               probabilityInfo[BinaryDictionary.FORMAT_WORD_PROPERTY_LEVEL_INDEX],
     76               probabilityInfo[BinaryDictionary.FORMAT_WORD_PROPERTY_COUNT_INDEX]);
     77     }
     78 
     79     // Construct word property using information from native code.
     80     // This represents invalid word when the probability is BinaryDictionary.NOT_A_PROBABILITY.
     81     public WordProperty(final int[] codePoints, final boolean isNotAWord,
     82             final boolean isPossiblyOffensive, final boolean hasBigram,
     83             final boolean isBeginningOfSentence, final int[] probabilityInfo,
     84             final ArrayList<int[][]> ngramPrevWordsArray,
     85             final ArrayList<boolean[]> ngramPrevWordIsBeginningOfSentenceArray,
     86             final ArrayList<int[]> ngramTargets, final ArrayList<int[]> ngramProbabilityInfo) {
     87         mWord = StringUtils.getStringFromNullTerminatedCodePointArray(codePoints);
     88         mProbabilityInfo = createProbabilityInfoFromArray(probabilityInfo);
     89         final ArrayList<NgramProperty> ngrams = new ArrayList<>();
     90         mIsBeginningOfSentence = isBeginningOfSentence;
     91         mIsNotAWord = isNotAWord;
     92         mIsPossiblyOffensive = isPossiblyOffensive;
     93         mHasNgrams = hasBigram;
     94 
     95         final int relatedNgramCount = ngramTargets.size();
     96         for (int i = 0; i < relatedNgramCount; i++) {
     97             final String ngramTargetString =
     98                     StringUtils.getStringFromNullTerminatedCodePointArray(ngramTargets.get(i));
     99             final WeightedString ngramTarget = new WeightedString(ngramTargetString,
    100                     createProbabilityInfoFromArray(ngramProbabilityInfo.get(i)));
    101             final int[][] prevWords = ngramPrevWordsArray.get(i);
    102             final boolean[] isBeginningOfSentenceArray =
    103                     ngramPrevWordIsBeginningOfSentenceArray.get(i);
    104             final WordInfo[] wordInfoArray = new WordInfo[prevWords.length];
    105             for (int j = 0; j < prevWords.length; j++) {
    106                 wordInfoArray[j] = isBeginningOfSentenceArray[j]
    107                         ? WordInfo.BEGINNING_OF_SENTENCE_WORD_INFO
    108                         : new WordInfo(StringUtils.getStringFromNullTerminatedCodePointArray(
    109                                 prevWords[j]));
    110             }
    111             final NgramContext ngramContext = new NgramContext(wordInfoArray);
    112             ngrams.add(new NgramProperty(ngramTarget, ngramContext));
    113         }
    114         mNgrams = ngrams.isEmpty() ? null : ngrams;
    115     }
    116 
    117     // TODO: Remove
    118     @UsedForTesting
    119     public ArrayList<WeightedString> getBigrams() {
    120         if (null == mNgrams) {
    121             return null;
    122         }
    123         final ArrayList<WeightedString> bigrams = new ArrayList<>();
    124         for (final NgramProperty ngram : mNgrams) {
    125             if (ngram.mNgramContext.getPrevWordCount() == 1) {
    126                 bigrams.add(ngram.mTargetWord);
    127             }
    128         }
    129         return bigrams;
    130     }
    131 
    132     public int getProbability() {
    133         return mProbabilityInfo.mProbability;
    134     }
    135 
    136     private static int computeHashCode(WordProperty word) {
    137         return Arrays.hashCode(new Object[] {
    138                 word.mWord,
    139                 word.mProbabilityInfo,
    140                 word.mNgrams,
    141                 word.mIsNotAWord,
    142                 word.mIsPossiblyOffensive
    143         });
    144     }
    145 
    146     /**
    147      * Three-way comparison.
    148      *
    149      * A Word x is greater than a word y if x has a higher frequency. If they have the same
    150      * frequency, they are sorted in lexicographic order.
    151      */
    152     @Override
    153     public int compareTo(final WordProperty w) {
    154         if (getProbability() < w.getProbability()) return 1;
    155         if (getProbability() > w.getProbability()) return -1;
    156         return mWord.compareTo(w.mWord);
    157     }
    158 
    159     /**
    160      * Equality test.
    161      *
    162      * Words are equal if they have the same frequency, the same spellings, and the same
    163      * attributes.
    164      */
    165     @Override
    166     public boolean equals(Object o) {
    167         if (o == this) return true;
    168         if (!(o instanceof WordProperty)) return false;
    169         WordProperty w = (WordProperty)o;
    170         return mProbabilityInfo.equals(w.mProbabilityInfo)
    171                 && mWord.equals(w.mWord) && equals(mNgrams, w.mNgrams)
    172                 && mIsNotAWord == w.mIsNotAWord && mIsPossiblyOffensive == w.mIsPossiblyOffensive
    173                 && mHasNgrams == w.mHasNgrams;
    174     }
    175 
    176     // TDOO: Have a utility method like java.util.Objects.equals.
    177     private static <T> boolean equals(final ArrayList<T> a, final ArrayList<T> b) {
    178         if (null == a) {
    179             return null == b;
    180         }
    181         return a.equals(b);
    182     }
    183 
    184     @Override
    185     public int hashCode() {
    186         if (mHashCode == 0) {
    187             mHashCode = computeHashCode(this);
    188         }
    189         return mHashCode;
    190     }
    191 
    192     @UsedForTesting
    193     public boolean isValid() {
    194         return getProbability() != Dictionary.NOT_A_PROBABILITY;
    195     }
    196 
    197     @Override
    198     public String toString() {
    199         return CombinedFormatUtils.formatWordProperty(this);
    200     }
    201 }
    202