diff options
Diffstat (limited to 'java/src/com')
| -rw-r--r-- | java/src/com/android/inputmethod/latin/makedict/WordProperty.java | 19 | ||||
| -rw-r--r-- | java/src/com/android/inputmethod/latin/utils/CombinedFormatUtils.java | 18 |
2 files changed, 27 insertions, 10 deletions
diff --git a/java/src/com/android/inputmethod/latin/makedict/WordProperty.java b/java/src/com/android/inputmethod/latin/makedict/WordProperty.java index b129c3e40..e7808e46e 100644 --- a/java/src/com/android/inputmethod/latin/makedict/WordProperty.java +++ b/java/src/com/android/inputmethod/latin/makedict/WordProperty.java @@ -87,7 +87,7 @@ public final class WordProperty implements Comparable<WordProperty> { final boolean isPossiblyOffensive, final boolean hasBigram, final boolean hasShortcuts, final boolean isBeginningOfSentence, final int[] probabilityInfo, final ArrayList<int[][]> ngramPrevWordsArray, - final ArrayList<boolean[]> outNgramPrevWordIsBeginningOfSentenceArray, + final ArrayList<boolean[]> ngramPrevWordIsBeginningOfSentenceArray, final ArrayList<int[]> ngramTargets, final ArrayList<int[]> ngramProbabilityInfo, final ArrayList<int[]> shortcutTargets, final ArrayList<Integer> shortcutProbabilities) { @@ -102,16 +102,22 @@ public final class WordProperty implements Comparable<WordProperty> { mHasNgrams = hasBigram; final int relatedNgramCount = ngramTargets.size(); - final WordInfo currentWordInfo = - mIsBeginningOfSentence ? WordInfo.BEGINNING_OF_SENTENCE_WORD_INFO - : new WordInfo(mWord); - final NgramContext ngramContext = new NgramContext(currentWordInfo); for (int i = 0; i < relatedNgramCount; i++) { final String ngramTargetString = StringUtils.getStringFromNullTerminatedCodePointArray(ngramTargets.get(i)); final WeightedString ngramTarget = new WeightedString(ngramTargetString, createProbabilityInfoFromArray(ngramProbabilityInfo.get(i))); - // TODO: Support n-gram. + final int[][] prevWords = ngramPrevWordsArray.get(i); + final boolean[] isBeginningOfSentenceArray = + ngramPrevWordIsBeginningOfSentenceArray.get(i); + final WordInfo[] wordInfoArray = new WordInfo[prevWords.length]; + for (int j = 0; j < prevWords.length; j++) { + wordInfoArray[j] = isBeginningOfSentenceArray[j] + ? WordInfo.BEGINNING_OF_SENTENCE_WORD_INFO + : new WordInfo(StringUtils.getStringFromNullTerminatedCodePointArray( + prevWords[j])); + } + final NgramContext ngramContext = new NgramContext(wordInfoArray); ngrams.add(new NgramProperty(ngramTarget, ngramContext)); } mNgrams = ngrams.isEmpty() ? null : ngrams; @@ -126,6 +132,7 @@ public final class WordProperty implements Comparable<WordProperty> { } // TODO: Remove + @UsedForTesting public ArrayList<WeightedString> getBigrams() { if (null == mNgrams) { return null; diff --git a/java/src/com/android/inputmethod/latin/utils/CombinedFormatUtils.java b/java/src/com/android/inputmethod/latin/utils/CombinedFormatUtils.java index 248246232..4e0f5f583 100644 --- a/java/src/com/android/inputmethod/latin/utils/CombinedFormatUtils.java +++ b/java/src/com/android/inputmethod/latin/utils/CombinedFormatUtils.java @@ -17,6 +17,7 @@ package com.android.inputmethod.latin.utils; import com.android.inputmethod.latin.makedict.DictionaryHeader; +import com.android.inputmethod.latin.makedict.NgramProperty; import com.android.inputmethod.latin.makedict.ProbabilityInfo; import com.android.inputmethod.latin.makedict.WeightedString; import com.android.inputmethod.latin.makedict.WordProperty; @@ -26,6 +27,8 @@ import java.util.HashMap; public class CombinedFormatUtils { public static final String DICTIONARY_TAG = "dictionary"; public static final String BIGRAM_TAG = "bigram"; + public static final String NGRAM_TAG = "ngram"; + public static final String NGRAM_PREV_WORD_TAG = "prev_word"; public static final String SHORTCUT_TAG = "shortcut"; public static final String PROBABILITY_TAG = "f"; public static final String HISTORICAL_INFO_TAG = "historicalInfo"; @@ -76,12 +79,19 @@ public class CombinedFormatUtils { } } if (wordProperty.mHasNgrams) { - // TODO: Support ngram. - for (final WeightedString bigram : wordProperty.getBigrams()) { - builder.append(" " + BIGRAM_TAG + "=" + bigram.mWord); + for (final NgramProperty ngramProperty : wordProperty.mNgrams) { + builder.append(" " + NGRAM_TAG + "=" + ngramProperty.mTargetWord.mWord); builder.append(","); - builder.append(formatProbabilityInfo(bigram.mProbabilityInfo)); + builder.append(formatProbabilityInfo(ngramProperty.mTargetWord.mProbabilityInfo)); builder.append("\n"); + for (int i = 0; i < ngramProperty.mNgramContext.getPrevWordCount(); i++) { + builder.append(" " + NGRAM_PREV_WORD_TAG + "[" + i + "]=" + + ngramProperty.mNgramContext.getNthPrevWord(i + 1)); + if (ngramProperty.mNgramContext.isNthPrevWordBeginningOfSontence(i + 1)) { + builder.append("," + BEGINNING_OF_SENTENCE_TAG + "=true"); + } + builder.append("\n"); + } } } return builder.toString(); |