Home | History | Annotate | Download | only in dicttool 
      1 /**
      2  * Copyright (C) 2012 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.dicttool;
     18 
     19 import com.android.inputmethod.latin.makedict.FormatSpec;
     20 import com.android.inputmethod.latin.makedict.FusionDictionary;
     21 import com.android.inputmethod.latin.makedict.FusionDictionary.PtNode;
     22 import com.android.inputmethod.latin.makedict.WeightedString;
     23 import com.android.inputmethod.latin.makedict.WordProperty;
     24 
     25 import java.util.Arrays;
     26 import java.util.ArrayList;
     27 
     28 public class Info extends Dicttool.Command {
     29     public static final String COMMAND = "info";
     30 
     31     public Info() {
     32     }
     33 
     34     @Override
     35     public String getHelp() {
     36         return COMMAND + " <filename>: prints various information about a dictionary file";
     37     }
     38 
     39     private static void showInfo(final FusionDictionary dict, final boolean plumbing) {
     40         System.out.println("Header attributes :");
     41         System.out.print(dict.mOptions.toString(2, plumbing));
     42         int wordCount = 0;
     43         int bigramCount = 0;
     44         int shortcutCount = 0;
     45         int whitelistCount = 0;
     46         for (final WordProperty wordProperty : dict) {
     47             ++wordCount;
     48             if (wordProperty.mHasNgrams) {
     49                 bigramCount += wordProperty.mNgrams.size();
     50             }
     51             if (null != wordProperty.mShortcutTargets) {
     52                 shortcutCount += wordProperty.mShortcutTargets.size();
     53                 for (WeightedString shortcutTarget : wordProperty.mShortcutTargets) {
     54                     if (FormatSpec.SHORTCUT_WHITELIST_FREQUENCY
     55                             == shortcutTarget.getProbability()) {
     56                         ++whitelistCount;
     57                     }
     58                 }
     59             }
     60         }
     61         System.out.println("Words in the dictionary : " + wordCount);
     62         System.out.println("Bigram count : " + bigramCount);
     63         System.out.println("Shortcuts : " + shortcutCount + " (out of which " + whitelistCount
     64                 + " whitelist entries)");
     65     }
     66 
     67     private static void showWordInfo(final FusionDictionary dict, final String word) {
     68         final PtNode ptNode = FusionDictionary.findWordInTree(dict.mRootNodeArray, word);
     69         if (null == ptNode) {
     70             System.out.println(word + " is not in the dictionary");
     71             return;
     72         }
     73         System.out.println("Word: " + word);
     74         System.out.println("  Freq: " + ptNode.getProbability());
     75         if (ptNode.getIsNotAWord()) {
     76             System.out.println("  Is not a word");
     77         }
     78         if (ptNode.getIsPossiblyOffensive()) {
     79             System.out.println("  Is possibly offensive");
     80         }
     81         final ArrayList<WeightedString> shortcutTargets = ptNode.getShortcutTargets();
     82         if (null == shortcutTargets || shortcutTargets.isEmpty()) {
     83             System.out.println("  No shortcuts");
     84         } else {
     85             for (final WeightedString shortcutTarget : shortcutTargets) {
     86                 System.out.println("  Shortcut target: " + shortcutTarget.mWord + " ("
     87                         + (FormatSpec.SHORTCUT_WHITELIST_FREQUENCY
     88                                 == shortcutTarget.getProbability() ?
     89                                         "whitelist" : shortcutTarget.getProbability()) + ")");
     90             }
     91         }
     92         final ArrayList<WeightedString> bigrams = ptNode.getBigrams();
     93         if (null == bigrams || bigrams.isEmpty()) {
     94             System.out.println("  No bigrams");
     95         } else {
     96             for (final WeightedString bigram : bigrams) {
     97                 System.out.println(
     98                         "  Bigram: " + bigram.mWord + " (" + bigram.getProbability() + ")");
     99             }
    100         }
    101     }
    102 
    103     @Override
    104     public void run() {
    105         if (mArgs.length < 1) {
    106             throw new RuntimeException("Not enough arguments for command " + COMMAND);
    107         }
    108         final boolean plumbing;
    109         if ("-p".equals(mArgs[0])) {
    110             plumbing = true;
    111             mArgs = Arrays.copyOfRange(mArgs, 1, mArgs.length);
    112             if (mArgs.length != 1) { // There should be only 1 argument left
    113                 throw new RuntimeException("Wrong number of arguments for command " + COMMAND);
    114             }
    115         } else {
    116             plumbing = false;
    117         }
    118         final String filename = mArgs[0];
    119         final boolean hasWordArguments = (1 == mArgs.length);
    120         final FusionDictionary dict = BinaryDictOffdeviceUtils.getDictionary(filename,
    121                 hasWordArguments /* report */);
    122         if (hasWordArguments) {
    123             showInfo(dict, plumbing);
    124         } else {
    125             for (int i = 1; i < mArgs.length; ++i) {
    126                 showWordInfo(dict, mArgs[i]);
    127             }
    128         }
    129     }
    130 }
    131