diff options
author | 2014-10-23 14:32:45 +0900 | |
---|---|---|
committer | 2014-10-23 14:32:45 +0900 | |
commit | 16cc3992d7468ef781753df7b4227330e0834501 (patch) | |
tree | a36882f970c5128107f61c119c7513d948133e7f /tests/src/com/android/inputmethod/latin/BinaryDictionaryDecayingTests.java | |
parent | d3a4c5132422b189c8dbb94dbbe84a9b9761b0a8 (diff) | |
download | latinime-16cc3992d7468ef781753df7b4227330e0834501.tar.gz latinime-16cc3992d7468ef781753df7b4227330e0834501.tar.xz latinime-16cc3992d7468ef781753df7b4227330e0834501.zip |
Use trigrams for personalization dict.
5Bug: 14425059
Change-Id: I73cf6904e569d60996a3b079f16ea6df0cb90f02
Diffstat (limited to 'tests/src/com/android/inputmethod/latin/BinaryDictionaryDecayingTests.java')
-rw-r--r-- | tests/src/com/android/inputmethod/latin/BinaryDictionaryDecayingTests.java | 64 |
1 files changed, 64 insertions, 0 deletions
diff --git a/tests/src/com/android/inputmethod/latin/BinaryDictionaryDecayingTests.java b/tests/src/com/android/inputmethod/latin/BinaryDictionaryDecayingTests.java index 15f7568c8..8f3373cc2 100644 --- a/tests/src/com/android/inputmethod/latin/BinaryDictionaryDecayingTests.java +++ b/tests/src/com/android/inputmethod/latin/BinaryDictionaryDecayingTests.java @@ -32,6 +32,7 @@ import com.android.inputmethod.latin.makedict.UnsupportedFormatException; import com.android.inputmethod.latin.utils.BinaryDictionaryUtils; import com.android.inputmethod.latin.utils.FileUtils; import com.android.inputmethod.latin.utils.LocaleUtils; +import com.android.inputmethod.latin.utils.WordInputEventForPersonalization; import java.io.File; import java.io.IOException; @@ -39,6 +40,7 @@ import java.util.ArrayList; import java.util.HashMap; import java.util.HashSet; import java.util.Locale; +import java.util.Map; import java.util.Random; import java.util.concurrent.TimeUnit; @@ -748,4 +750,66 @@ public class BinaryDictionaryDecayingTests extends AndroidTestCase { binaryDictionary.close(); } + + public void testUpdateEntriesForInputEvents() { + for (final int formatVersion : DICT_FORMAT_VERSIONS) { + testUpdateEntriesForInputEvents(formatVersion); + } + } + + private void testUpdateEntriesForInputEvents(final int formatVersion) { + setCurrentTimeForTestMode(mCurrentTime); + final int codePointSetSize = 20; + final int EVENT_COUNT = 1000; + final double CONTINUE_RATE = 0.9; + final long seed = System.currentTimeMillis(); + final Random random = new Random(seed); + final File dictFile = createEmptyDictionaryAndGetFile(formatVersion); + + final int[] codePointSet = CodePointUtils.generateCodePointSet(codePointSetSize, random); + final ArrayList<String> unigrams = new ArrayList<>(); + final ArrayList<Pair<String, String>> bigrams = new ArrayList<>(); + final ArrayList<Pair<Pair<String, String>, String>> trigrams = new ArrayList<>(); + + final WordInputEventForPersonalization[] inputEvents = + new WordInputEventForPersonalization[EVENT_COUNT]; + NgramContext ngramContext = NgramContext.EMPTY_PREV_WORDS_INFO; + int prevWordCount = 0; + for (int i = 0; i < inputEvents.length; i++) { + final String word = CodePointUtils.generateWord(random, codePointSet); + inputEvents[i] = new WordInputEventForPersonalization(word, ngramContext, + true /* isValid */, mCurrentTime); + unigrams.add(word); + if (prevWordCount >= 2) { + final Pair<String, String> prevWordsPair = bigrams.get(bigrams.size() - 1); + trigrams.add(new Pair<>(prevWordsPair, word)); + } + if (prevWordCount >= 1) { + bigrams.add(new Pair<>(ngramContext.getNthPrevWord(1 /* n */).toString(), word)); + } + if (random.nextDouble() > CONTINUE_RATE) { + ngramContext = NgramContext.EMPTY_PREV_WORDS_INFO; + prevWordCount = 0; + } else { + ngramContext = ngramContext.getNextNgramContext(new WordInfo(word)); + prevWordCount++; + } + } + final BinaryDictionary binaryDictionary = getBinaryDictionary(dictFile); + binaryDictionary.updateEntriesForInputEvents(inputEvents); + + for (final String word : unigrams) { + assertTrue(binaryDictionary.isInDictionary(word)); + } + for (final Pair<String, String> bigram : bigrams) { + assertTrue(isValidBigram(binaryDictionary, bigram.first, bigram.second)); + } + if (!supportsNgram(formatVersion)) { + return; + } + for (final Pair<Pair<String, String>, String> trigram : trigrams) { + assertTrue(isValidTrigram(binaryDictionary, trigram.first.first, trigram.first.second, + trigram.second)); + } + } } |