private static ArrayList<SuggestedWordInfo> getSuggestionsInfoListWithDebugInfo(
final String typedWord, final ArrayList<SuggestedWordInfo> suggestions) {
final SuggestedWordInfo typedWordInfo = suggestions.get(0);
typedWordInfo.setDebugString("+");
final int suggestionsSize = suggestions.size();
final ArrayList<SuggestedWordInfo> suggestionsList =
CollectionUtils.newArrayList(suggestionsSize);
suggestionsList.add(typedWordInfo);
// Note: i here is the index in mScores[], but the index in mSuggestions is one more
// than i because we added the typed word to mSuggestions without touching mScores.
for (int i = 0; i < suggestionsSize - 1; ++i) {
final SuggestedWordInfo cur = suggestions.get(i + 1);
final float normalizedScore =
BinaryDictionary.calcNormalizedScore(typedWord, cur.toString(), cur.mScore);
final String scoreInfoString;
if (normalizedScore > 0) {
scoreInfoString = String.format(Locale.ROOT, "%d (%4.2f)", cur.mScore, normalizedScore);
} else {
scoreInfoString = Integer.toString(cur.mScore);
}
cur.setDebugString(scoreInfoString);
suggestionsList.add(cur);
}
return suggestionsList;
}
public void testVer2DictGetWordProperty() {
final FormatOptions formatOptions = BinaryDictUtils.VERSION2_OPTIONS;
final ArrayList<String> words = sWords;
final HashMap<String, List<String>> shortcuts = sShortcuts;
final String dictName = "testGetWordProperty";
final String dictVersion = Long.toString(System.currentTimeMillis());
final FusionDictionary dict =
new FusionDictionary(
new PtNodeArray(),
BinaryDictUtils.makeDictionaryOptions(dictName, dictVersion, formatOptions));
addUnigrams(words.size(), dict, words, shortcuts);
addBigrams(dict, words, sEmptyBigrams);
final File file =
BinaryDictUtils.getDictFile(
dictName, dictVersion, formatOptions, getContext().getCacheDir());
file.delete();
timeWritingDictToFile(file, dict, formatOptions);
final BinaryDictionary binaryDictionary =
new BinaryDictionary(
file.getAbsolutePath(),
0 /* offset */,
file.length(),
true /* useFullEditDistance */,
Locale.ENGLISH,
dictName,
false /* isUpdatable */);
for (final String word : words) {
final WordProperty wordProperty =
binaryDictionary.getWordProperty(word, false /* isBeginningOfSentence */);
assertEquals(word, wordProperty.mWord);
assertEquals(UNIGRAM_FREQ, wordProperty.getProbability());
if (shortcuts.containsKey(word)) {
assertEquals(shortcuts.get(word).size(), wordProperty.mShortcutTargets.size());
final List<String> shortcutList = shortcuts.get(word);
assertTrue(wordProperty.mHasShortcuts);
for (final WeightedString shortcutTarget : wordProperty.mShortcutTargets) {
assertTrue(shortcutList.contains(shortcutTarget.mWord));
assertEquals(UNIGRAM_FREQ, shortcutTarget.getProbability());
shortcutList.remove(shortcutTarget.mWord);
}
assertTrue(shortcutList.isEmpty());
}
}
}
public void testVer2DictIteration() {
final FormatOptions formatOptions = BinaryDictUtils.VERSION2_OPTIONS;
final ArrayList<String> words = sWords;
final HashMap<String, List<String>> shortcuts = sShortcuts;
final SparseArray<List<Integer>> bigrams = sEmptyBigrams;
final String dictName = "testGetWordProperty";
final String dictVersion = Long.toString(System.currentTimeMillis());
final FusionDictionary dict =
new FusionDictionary(
new PtNodeArray(),
BinaryDictUtils.makeDictionaryOptions(dictName, dictVersion, formatOptions));
addUnigrams(words.size(), dict, words, shortcuts);
addBigrams(dict, words, bigrams);
final File file =
BinaryDictUtils.getDictFile(
dictName, dictVersion, formatOptions, getContext().getCacheDir());
timeWritingDictToFile(file, dict, formatOptions);
Log.d(TAG, file.getAbsolutePath());
final BinaryDictionary binaryDictionary =
new BinaryDictionary(
file.getAbsolutePath(),
0 /* offset */,
file.length(),
true /* useFullEditDistance */,
Locale.ENGLISH,
dictName,
false /* isUpdatable */);
final HashSet<String> wordSet = new HashSet<>(words);
final HashSet<Pair<String, String>> bigramSet = new HashSet<>();
for (int i = 0; i < words.size(); i++) {
final List<Integer> bigramList = bigrams.get(i);
if (bigramList != null) {
for (final Integer word1Index : bigramList) {
final String word1 = words.get(word1Index);
bigramSet.add(new Pair<>(words.get(i), word1));
}
}
}
int token = 0;
do {
final BinaryDictionary.GetNextWordPropertyResult result =
binaryDictionary.getNextWordProperty(token);
final WordProperty wordProperty = result.mWordProperty;
final String word0 = wordProperty.mWord;
assertEquals(UNIGRAM_FREQ, wordProperty.mProbabilityInfo.mProbability);
wordSet.remove(word0);
if (shortcuts.containsKey(word0)) {
assertEquals(shortcuts.get(word0).size(), wordProperty.mShortcutTargets.size());
final List<String> shortcutList = shortcuts.get(word0);
assertNotNull(wordProperty.mShortcutTargets);
for (final WeightedString shortcutTarget : wordProperty.mShortcutTargets) {
assertTrue(shortcutList.contains(shortcutTarget.mWord));
assertEquals(UNIGRAM_FREQ, shortcutTarget.getProbability());
shortcutList.remove(shortcutTarget.mWord);
}
assertTrue(shortcutList.isEmpty());
}
for (int j = 0; j < wordProperty.mBigrams.size(); j++) {
final String word1 = wordProperty.mBigrams.get(j).mWord;
final Pair<String, String> bigram = new Pair<>(word0, word1);
assertTrue(bigramSet.contains(bigram));
bigramSet.remove(bigram);
}
token = result.mNextToken;
} while (token != 0);
assertTrue(wordSet.isEmpty());
assertTrue(bigramSet.isEmpty());
}