src/com/hughes/android/dictionary/engine/IndexBuilder.java

   1 // Copyright 2011 Google Inc. All Rights Reserved.
   2 //
   3 // Licensed under the Apache License, Version 2.0 (the "License");
   4 // you may not use this file except in compliance with the License.
   5 // You may obtain a copy of the License at
   6 //
   7 //     http://www.apache.org/licenses/LICENSE-2.0
   8 //
   9 // Unless required by applicable law or agreed to in writing, software
  10 // distributed under the License is distributed on an "AS IS" BASIS,
  11 // WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
  12 // See the License for the specific language governing permissions and
  13 // limitations under the License.
  14
  15 package com.hughes.android.dictionary.engine;
  16
  17 import java.util.ArrayList;
  18 import java.util.Collections;
  19 import java.util.Comparator;
  20 import java.util.EnumMap;
  21 import java.util.HashSet;
  22 import java.util.List;
  23 import java.util.Map;
  24 import java.util.Set;
  25 import java.util.SortedMap;
  26 import java.util.TreeMap;
  27
  28 import com.hughes.android.dictionary.engine.Index.IndexEntry;
  29 import com.hughes.android.dictionary.parser.DictFileParser;
  30
  31 public class IndexBuilder {
  32
  33     final DictionaryBuilder dictionaryBuilder;
  34     public final Index index;
  35     final Set<String> stoplist;
  36
  37     final SortedMap<String, TokenData> tokenToData;
  38
  39     IndexBuilder(final DictionaryBuilder dictionaryBuilder, final String shortName, final String longName, final Language language, final String normalizerRules, final Set<String> stoplist, final boolean swapPairEntries) {
  40         this.dictionaryBuilder = dictionaryBuilder;
  41         index = new Index(dictionaryBuilder.dictionary, shortName, longName, language, normalizerRules, swapPairEntries, stoplist);
  42         tokenToData = new TreeMap<String, TokenData>(index.getSortComparator());
  43         this.stoplist = stoplist;
  44     }
  45
  46     public void build() {
  47         final Set<IndexedEntry> tokenIndexedEntries = new HashSet<IndexedEntry>();
  48         final List<RowBase> rows = index.rows;
  49         index.mainTokenCount = 0;
  50         for (final TokenData tokenData : tokenToData.values()) {
  51             tokenIndexedEntries.clear();
  52             final int indexIndex = index.sortedIndexEntries.size();
  53             final int startRow = rows.size();
  54
  55             TokenRow tokenRow = null;
  56             if (!tokenData.htmlEntries.isEmpty()) {
  57                 tokenRow = new TokenRow(indexIndex, rows.size(), index, tokenData.hasMainEntry);
  58                 rows.add(tokenRow);
  59             }
  60
  61 //    System.out.println("Added TokenRow: " + rows.get(rows.size() - 1));
  62
  63             int numRows = 0;  // off by one--doesn't count the token row!
  64 //      System.out.println("TOKEN: " + tokenData.token);
  65             for (final Map.Entry<EntryTypeName, List<IndexedEntry>> typeToIndexedEntries : tokenData.typeToEntries.entrySet()) {
  66                 for (final IndexedEntry indexedEntry : typeToIndexedEntries.getValue()) {
  67                     if (!indexedEntry.isValid) {
  68                         continue;
  69                     }
  70
  71                     if (tokenRow == null) {
  72                         tokenRow = new TokenRow(indexIndex, rows.size(), index, tokenData.hasMainEntry);
  73                         rows.add(tokenRow);
  74                     }
  75
  76                     if (indexedEntry.entry.index() == -1) {
  77                         indexedEntry.entry.addToDictionary(dictionaryBuilder.dictionary);
  78                         assert indexedEntry.entry.index() >= 0;
  79                     }
  80                     if (tokenIndexedEntries.add(indexedEntry) && !tokenData.htmlEntries.contains(indexedEntry.entry)) {
  81                         rows.add(indexedEntry.entry.CreateRow(rows.size(), index));
  82                         ++indexedEntry.entry.entrySource.numEntries;
  83                         ++numRows;
  84
  85 //            System.out.print("  " + typeToEntry.getKey() + ": ");
  86                         //          rows.get(rows.size() - 1).print(System.out);
  87 //            System.out.println();
  88                     }
  89                 }
  90             }
  91
  92             if (tokenRow != null) {
  93                 if (tokenRow.hasMainEntry) {
  94                     index.mainTokenCount++;
  95                 }
  96
  97                 final Index.IndexEntry indexEntry = new Index.IndexEntry(index, tokenData.token, index
  98                         .normalizer().transliterate(tokenData.token), startRow, numRows);
  99                 indexEntry.htmlEntries.addAll(tokenData.htmlEntries);
 100                 index.sortedIndexEntries.add(indexEntry);
 101             }
 102         }
 103
 104         final List<IndexEntry> entriesSortedByNumRows = new ArrayList<IndexEntry>(index.sortedIndexEntries);
 105         Collections.sort(entriesSortedByNumRows, new Comparator<IndexEntry>() {
 106             @Override
 107             public int compare(IndexEntry object1, IndexEntry object2) {
 108                 return object2.numRows - object1.numRows;
 109             }
 110         });
 111         System.out.println("Most common tokens:");
 112         for (int i = 0; i < 50 && i < entriesSortedByNumRows.size(); ++i) {
 113             System.out.println("  " + entriesSortedByNumRows.get(i));
 114         }
 115     }
 116
 117     public static class TokenData {
 118         final String token;
 119
 120         final Map<EntryTypeName, List<IndexedEntry>> typeToEntries = new EnumMap<EntryTypeName, List<IndexedEntry>>(EntryTypeName.class);
 121         public boolean hasMainEntry = false;
 122
 123         public List<HtmlEntry> htmlEntries = new ArrayList<HtmlEntry>();
 124
 125         TokenData(final String token) {
 126             assert token.equals(token.trim());
 127             assert token.length() > 0;
 128             this.token = token;
 129         }
 130     }
 131
 132     public TokenData getOrCreateTokenData(final String token) {
 133         TokenData tokenData = tokenToData.get(token);
 134         if (tokenData == null) {
 135             tokenData = new TokenData(token);
 136             tokenToData.put(token, tokenData);
 137         }
 138         return tokenData;
 139     }
 140
 141     private List<IndexedEntry> getOrCreateEntries(final String token, final EntryTypeName entryTypeName) {
 142         final TokenData tokenData = getOrCreateTokenData(token);
 143         List<IndexedEntry> entries = tokenData.typeToEntries.get(entryTypeName);
 144         if (entryTypeName.mainWord) {
 145             tokenData.hasMainEntry = true;
 146         }
 147         if (entries == null) {
 148             entries = new ArrayList<IndexedEntry>();
 149             tokenData.typeToEntries.put(entryTypeName, entries);
 150         }
 151         return entries;
 152     }
 153
 154     public void addEntryWithTokens(final IndexedEntry indexedEntry, final Set<String> tokens,
 155                                    final EntryTypeName entryTypeName) {
 156         if (indexedEntry == null) {
 157             System.out.println("asdfasdf");
 158         }
 159         assert indexedEntry != null;
 160         for (final String token : tokens) {
 161             if (entryTypeName.overridesStopList || !stoplist.contains(token)) {
 162                 getOrCreateEntries(token, entryTypeName).add(indexedEntry);
 163             }
 164         }
 165     }
 166
 167     public void addEntryWithString(final IndexedEntry indexedEntry, final String untokenizedString,
 168                                    final EntryTypeName entryTypeName) {
 169         final Set<String> tokens = DictFileParser.tokenize(untokenizedString, DictFileParser.NON_CHAR);
 170         addEntryWithTokens(indexedEntry, tokens, tokens.size() == 1 ? entryTypeName.singleWordInstance : entryTypeName);
 171     }
 172
 173     public void addEntryWithStringNoSingle(final IndexedEntry indexedEntry, final String untokenizedString,
 174                                            final EntryTypeName entryTypeName) {
 175         final Set<String> tokens = DictFileParser.tokenize(untokenizedString, DictFileParser.NON_CHAR);
 176         addEntryWithTokens(indexedEntry, tokens, entryTypeName);
 177     }
 178 }