Home | History | Annotate | Download | only in makedict 
      1 /*
      2  * Copyright (C) 2013 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.utils.FileUtils;
     22 
     23 import java.io.File;
     24 import java.io.FileNotFoundException;
     25 import java.io.IOException;
     26 import java.util.ArrayList;
     27 
     28 /**
     29  * An implementation of binary dictionary decoder for version 4 binary dictionary.
     30  */
     31 @UsedForTesting
     32 public class Ver4DictDecoder extends AbstractDictDecoder {
     33     final File mDictDirectory;
     34 
     35     @UsedForTesting
     36     /* package */ Ver4DictDecoder(final File dictDirectory, final int factoryFlag) {
     37         this(dictDirectory, null /* factory */);
     38     }
     39 
     40     @UsedForTesting
     41     /* package */ Ver4DictDecoder(final File dictDirectory, final DictionaryBufferFactory factory) {
     42         mDictDirectory = dictDirectory;
     43 
     44     }
     45 
     46     @Override
     47     public DictionaryHeader readHeader() throws IOException, UnsupportedFormatException {
     48         // dictType is not being used in dicttool. Passing an empty string.
     49         final BinaryDictionary binaryDictionary= new BinaryDictionary(
     50               mDictDirectory.getAbsolutePath(), 0 /* offset */, 0 /* length */,
     51               true /* useFullEditDistance */, null /* locale */,
     52               "" /* dictType */, true /* isUpdatable */);
     53         final DictionaryHeader header = binaryDictionary.getHeader();
     54         binaryDictionary.close();
     55         if (header == null) {
     56             throw new IOException("Cannot read the dictionary header.");
     57         }
     58         return header;
     59     }
     60 
     61     @Override
     62     public FusionDictionary readDictionaryBinary(final boolean deleteDictIfBroken)
     63             throws FileNotFoundException, IOException, UnsupportedFormatException {
     64         // dictType is not being used in dicttool. Passing an empty string.
     65         final BinaryDictionary binaryDictionary = new BinaryDictionary(
     66               mDictDirectory.getAbsolutePath(), 0 /* offset */, 0 /* length */,
     67               true /* useFullEditDistance */, null /* locale */,
     68               "" /* dictType */, true /* isUpdatable */);
     69         final DictionaryHeader header = readHeader();
     70         final FusionDictionary fusionDict =
     71                 new FusionDictionary(new FusionDictionary.PtNodeArray(), header.mDictionaryOptions);
     72         int token = 0;
     73         final ArrayList<WordProperty> wordProperties = new ArrayList<>();
     74         do {
     75             final BinaryDictionary.GetNextWordPropertyResult result =
     76                     binaryDictionary.getNextWordProperty(token);
     77             final WordProperty wordProperty = result.mWordProperty;
     78             if (wordProperty == null) {
     79                 binaryDictionary.close();
     80                 if (deleteDictIfBroken) {
     81                     FileUtils.deleteRecursively(mDictDirectory);
     82                 }
     83                 return null;
     84             }
     85             wordProperties.add(wordProperty);
     86             token = result.mNextToken;
     87         } while (token != 0);
     88 
     89         // Insert unigrams into the fusion dictionary.
     90         for (final WordProperty wordProperty : wordProperties) {
     91             if (wordProperty.mIsBlacklistEntry) {
     92                 fusionDict.addBlacklistEntry(wordProperty.mWord, wordProperty.mShortcutTargets,
     93                         wordProperty.mIsNotAWord);
     94             } else {
     95                 fusionDict.add(wordProperty.mWord, wordProperty.mProbabilityInfo,
     96                         wordProperty.mShortcutTargets, wordProperty.mIsNotAWord);
     97             }
     98         }
     99         // Insert bigrams into the fusion dictionary.
    100         for (final WordProperty wordProperty : wordProperties) {
    101             if (wordProperty.mBigrams == null) {
    102                 continue;
    103             }
    104             final String word0 = wordProperty.mWord;
    105             for (final WeightedString bigram : wordProperty.mBigrams) {
    106                 fusionDict.setBigram(word0, bigram.mWord, bigram.mProbabilityInfo);
    107             }
    108         }
    109         binaryDictionary.close();
    110         return fusionDict;
    111     }
    112 }
    113