Donate to e Foundation | Murena handsets with /e/OS | Own a part of Murena! Learn more

Commit 94d9a230 authored by Keisuke Kuroyanagi's avatar Keisuke Kuroyanagi
Browse files

Extend jni interface to get bigrams in WordProperty. 

Bug: 12810574
Change-Id: Ia4b88d02ea8790a5c47d32376cc0b84c3e071ddd
parent 17e7768a

java/src/com/android/inputmethod/latin/BinaryDictionary.java

+19 −18
Original line number Diff line number Diff line
@@ -68,11 +68,12 @@ public final class BinaryDictionary extends Dictionary { 
    private static final int FORMAT_WORD_PROPERTY_HAS_BIGRAMS_INDEX = 2;

    private static final int FORMAT_WORD_PROPERTY_HAS_SHORTCUTS_INDEX = 3;



    // Format to get unigram historical info from native side via getWordPropertyNative().

    private static final int FORMAT_WORD_PROPERTY_OUTPUT_HISTORICAL_INFO_COUNT = 3;

    private static final int FORMAT_WORD_PROPERTY_TIMESTAMP_INDEX = 0;

    private static final int FORMAT_WORD_PROPERTY_LEVEL_INDEX = 1;

    private static final int FORMAT_WORD_PROPERTY_COUNT_INDEX = 2;

    // Format to get probability and historical info from native side via getWordPropertyNative().

    public static final int FORMAT_WORD_PROPERTY_OUTPUT_PROBABILITY_INFO_COUNT = 4;

    public static final int FORMAT_WORD_PROPERTY_PROBABILITY_INDEX = 0;

    public static final int FORMAT_WORD_PROPERTY_TIMESTAMP_INDEX = 1;

    public static final int FORMAT_WORD_PROPERTY_LEVEL_INDEX = 2;

    public static final int FORMAT_WORD_PROPERTY_COUNT_INDEX = 3;



    private long mNativeDict;

    private final Locale mLocale;
@@ -144,9 +145,9 @@ public final class BinaryDictionary extends Dictionary { 
    private static native int getProbabilityNative(long dict, int[] word);

    private static native int getBigramProbabilityNative(long dict, int[] word0, int[] word1);

    private static native void getWordPropertyNative(long dict, int[] word,

            int[] outCodePoints, boolean[] outFlags, int[] outProbability,

            int[] outHistoricalInfo, ArrayList<int[]> outShortcutTargets,

            ArrayList<Integer> outShortcutProbabilities);

            int[] outCodePoints, boolean[] outFlags, int[] outProbabilityInfo,

            ArrayList<int[]> outBigramTargets, ArrayList<int[]> outBigramProbabilityInfo,

            ArrayList<int[]> outShortcutTargets, ArrayList<Integer> outShortcutProbabilities);

    private static native int getSuggestionsNative(long dict, long proximityInfo,

            long traverseSession, int[] xCoordinates, int[] yCoordinates, int[] times,

            int[] pointerIds, int[] inputCodePoints, int inputSize, int commitPoint,
@@ -313,22 +314,22 @@ public final class BinaryDictionary extends Dictionary { 
        final int[] codePoints = StringUtils.toCodePointArray(word);

        final int[] outCodePoints = new int[MAX_WORD_LENGTH];

        final boolean[] outFlags = new boolean[FORMAT_WORD_PROPERTY_OUTPUT_FLAG_COUNT];

        final int[] outProbability = new int[1];

        final int[] outHistoricalInfo =

                new int[FORMAT_WORD_PROPERTY_OUTPUT_HISTORICAL_INFO_COUNT];

        final int[] outProbabilityInfo =

                new int[FORMAT_WORD_PROPERTY_OUTPUT_PROBABILITY_INFO_COUNT];

        final ArrayList<int[]> outBigramTargets = CollectionUtils.newArrayList();

        final ArrayList<int[]> outBigramProbabilityInfo = CollectionUtils.newArrayList();

        final ArrayList<int[]> outShortcutTargets = CollectionUtils.newArrayList();

        final ArrayList<Integer> outShortcutProbabilities = CollectionUtils.newArrayList();

        getWordPropertyNative(mNativeDict, codePoints, outCodePoints, outFlags, outProbability,

                outHistoricalInfo, outShortcutTargets, outShortcutProbabilities);

        getWordPropertyNative(mNativeDict, codePoints, outCodePoints, outFlags, outProbabilityInfo,

                outBigramTargets, outBigramProbabilityInfo, outShortcutTargets,

                outShortcutProbabilities);

        return new WordProperty(codePoints,

                outFlags[FORMAT_WORD_PROPERTY_IS_NOT_A_WORD_INDEX],

                outFlags[FORMAT_WORD_PROPERTY_IS_BLACKLISTED_INDEX],

                outFlags[FORMAT_WORD_PROPERTY_HAS_BIGRAMS_INDEX],

                outFlags[FORMAT_WORD_PROPERTY_HAS_SHORTCUTS_INDEX], outProbability[0],

                outHistoricalInfo[FORMAT_WORD_PROPERTY_TIMESTAMP_INDEX],

                outHistoricalInfo[FORMAT_WORD_PROPERTY_LEVEL_INDEX],

                outHistoricalInfo[FORMAT_WORD_PROPERTY_COUNT_INDEX],

                outShortcutTargets, outShortcutProbabilities);

                outFlags[FORMAT_WORD_PROPERTY_HAS_SHORTCUTS_INDEX], outProbabilityInfo,

                outBigramTargets, outBigramProbabilityInfo, outShortcutTargets,

                outShortcutProbabilities);

    }



    // Add a unigram entry to binary dictionary with unigram attributes in native code.

java/src/com/android/inputmethod/latin/utils/WordProperty.java

+40 −14
Original line number Diff line number Diff line
@@ -32,14 +32,29 @@ public class WordProperty { 
    public final boolean mIsBlacklisted;

    public final boolean mHasBigrams;

    public final boolean mHasShortcuts;

    public final ProbabilityInfo mProbabilityInfo;

    public final ArrayList<WeightedString> mBigramTargets = CollectionUtils.newArrayList();

    public final ArrayList<ProbabilityInfo> mBigramProbabilityInfo = CollectionUtils.newArrayList();

    public final ArrayList<WeightedString> mShortcutTargets = CollectionUtils.newArrayList();



    // TODO: Use this kind of Probability class for dictionary read/write code under the makedict

    // package.

    public static final class ProbabilityInfo {

        public final int mProbability;

    // mTimestamp, mLevel and mCount are historical info. These values are depend on the

        // wTimestamp, mLevel and mCount are historical info. These values are depend on the

        // implementation in native code; thus, we must not use them and have any assumptions about

        // them except for tests.

        public final int mTimestamp;

        public final int mLevel;

        public final int mCount;

    public final ArrayList<WeightedString> mShortcutTargets = CollectionUtils.newArrayList();



        public ProbabilityInfo(final int[] probabilityInfo) {

            mProbability = probabilityInfo[BinaryDictionary.FORMAT_WORD_PROPERTY_PROBABILITY_INDEX];

            mTimestamp = probabilityInfo[BinaryDictionary.FORMAT_WORD_PROPERTY_TIMESTAMP_INDEX];

            mLevel = probabilityInfo[BinaryDictionary.FORMAT_WORD_PROPERTY_LEVEL_INDEX];

            mCount = probabilityInfo[BinaryDictionary.FORMAT_WORD_PROPERTY_COUNT_INDEX];

        }

    }



    private static int getCodePointCount(final int[] codePoints) {

        for (int i = 0; i < codePoints.length; i++) {
@@ -53,18 +68,29 @@ public class WordProperty { 
    // This represents invalid word when the probability is BinaryDictionary.NOT_A_PROBABILITY.

    public WordProperty(final int[] codePoints, final boolean isNotAWord,

            final boolean isBlacklisted, final boolean hasBigram,

            final boolean hasShortcuts, final int probability, final int timestamp,

            final int level, final int count, final ArrayList<int[]> shortcutTargets,

            final boolean hasShortcuts, final int[] probabilityInfo,

            final ArrayList<int[]> bigramTargets, final ArrayList<int[]> bigramProbabilityInfo,

            final ArrayList<int[]> shortcutTargets,

            final ArrayList<Integer> shortcutProbabilities) {

        mCodePoints = new String(codePoints, 0 /* offset */, getCodePointCount(codePoints));

        mIsNotAWord = isNotAWord;

        mIsBlacklisted = isBlacklisted;

        mHasBigrams = hasBigram;

        mHasShortcuts = hasShortcuts;

        mProbability = probability;

        mTimestamp = timestamp;

        mLevel = level;

        mCount = count;

        mProbabilityInfo = new ProbabilityInfo(probabilityInfo);



        final int bigramTargetCount = bigramTargets.size();

        for (int i = 0; i < bigramTargetCount; i++) {

            final int[] bigramTargetCodePointArray = bigramTargets.get(i);

            final String bigramTargetString = new String(bigramTargetCodePointArray,

                    0 /* offset */, getCodePointCount(bigramTargetCodePointArray));

            final ProbabilityInfo bigramProbability =

                    new ProbabilityInfo(bigramProbabilityInfo.get(i));

            mBigramTargets.add(

                    new WeightedString(bigramTargetString, bigramProbability.mProbability));

            mBigramProbabilityInfo.add(bigramProbability);

        }



        final int shortcutTargetCount = shortcutTargets.size();

        for (int i = 0; i < shortcutTargetCount; i++) {

            final int[] shortcutTargetCodePointArray = shortcutTargets.get(i);
@@ -77,6 +103,6 @@ public class WordProperty { 


    @UsedForTesting

    public boolean isValid() {

        return mProbability != BinaryDictionary.NOT_A_PROBABILITY;

        return mProbabilityInfo.mProbability != BinaryDictionary.NOT_A_PROBABILITY;

    }

}
 
 No newline at end of file

native/jni/com_android_inputmethod_latin_BinaryDictionary.cpp

+7 −5
Original line number Diff line number Diff line
@@ -262,16 +262,17 @@ static jint latinime_BinaryDictionary_getBigramProbability(JNIEnv *env, jclass c 


static void latinime_BinaryDictionary_getWordProperty(JNIEnv *env, jclass clazz,

        jlong dict, jintArray word, jintArray outCodePoints, jbooleanArray outFlags,

        jintArray outProbability, jintArray outHistoricalInfo, jobject outShortcutTargets,

        jobject outShortcutProbabilities) {

        jintArray outProbabilityInfo, jobject outBigramTargets, jobject outBigramProbabilityInfo,

        jobject outShortcutTargets, jobject outShortcutProbabilities) {

    Dictionary *dictionary = reinterpret_cast<Dictionary *>(dict);

    if (!dictionary) return;

    const jsize wordLength = env->GetArrayLength(word);

    int wordCodePoints[wordLength];

    env->GetIntArrayRegion(word, 0, wordLength, wordCodePoints);

    const WordProperty wordProperty = dictionary->getWordProperty(wordCodePoints, wordLength);

    wordProperty.outputProperties(env, outCodePoints, outFlags, outProbability,

            outHistoricalInfo, outShortcutTargets, outShortcutProbabilities);

    wordProperty.outputProperties(env, outCodePoints, outFlags, outProbabilityInfo,

            outBigramTargets, outBigramProbabilityInfo, outShortcutTargets,

            outShortcutProbabilities);

}



static jfloat latinime_BinaryDictionary_calcNormalizedScore(JNIEnv *env, jclass clazz,
@@ -521,7 +522,8 @@ static const JNINativeMethod sMethods[] = { 
    },

    {

        const_cast<char *>("getWordPropertyNative"),

        const_cast<char *>("(J[I[I[Z[I[ILjava/util/ArrayList;Ljava/util/ArrayList;)V"),

        const_cast<char *>("(J[I[I[Z[ILjava/util/ArrayList;Ljava/util/ArrayList;"

                "Ljava/util/ArrayList;Ljava/util/ArrayList;)V"),

        reinterpret_cast<void *>(latinime_BinaryDictionary_getWordProperty)

    },

    {

native/jni/src/suggest/core/dictionary/word_property.cpp

+9 −6
Original line number Diff line number Diff line
@@ -19,20 +19,23 @@ 
namespace latinime {



void WordProperty::outputProperties(JNIEnv *const env, jintArray outCodePoints,

        jbooleanArray outFlags, jintArray outProbability, jintArray outHistoricalInfo,

        jobject outShortcutTargets, jobject outShortcutProbabilities) const {

        jbooleanArray outFlags, jintArray outProbabilityInfo, jobject outBigramTargets,

        jobject outBigramProbabilities, jobject outShortcutTargets,

        jobject outShortcutProbabilities) const {

    env->SetIntArrayRegion(outCodePoints, 0 /* start */, mCodePoints.size(), &mCodePoints[0]);

    jboolean flags[] = {mIsNotAWord, mIsBlacklisted, mHasBigrams, mHasShortcuts};

    env->SetBooleanArrayRegion(outFlags, 0 /* start */, NELEMS(flags), flags);

    env->SetIntArrayRegion(outProbability, 0 /* start */, 1 /* len */, &mProbability);

    int historicalInfo[] = {mTimestamp, mLevel, mCount};

    env->SetIntArrayRegion(outHistoricalInfo, 0 /* start */, NELEMS(historicalInfo),

            historicalInfo);

    int probabilityInfo[] = {mProbability, mTimestamp, mLevel, mCount};

    env->SetIntArrayRegion(outProbabilityInfo, 0 /* start */, NELEMS(probabilityInfo),

            probabilityInfo);



    jclass integerClass = env->FindClass("java/lang/Integer");

    jmethodID intToIntegerConstructorId = env->GetMethodID(integerClass, "<init>", "(I)V");

    jclass arrayListClass = env->FindClass("java/util/ArrayList");

    jmethodID addMethodId = env->GetMethodID(arrayListClass, "add", "(Ljava/lang/Object;)Z");



    // TODO: Output bigrams.

    // Output shortcuts.

    const int shortcutTargetCount = mShortcuts.size();

    for (int i = 0; i < shortcutTargetCount; ++i) {

        const std::vector<int> *const targetCodePoints = mShortcuts[i].getTargetCodePoints();

native/jni/src/suggest/core/dictionary/word_property.h

+2 −2
Original line number Diff line number Diff line
@@ -78,8 +78,8 @@ class WordProperty { 
              mShortcuts(*shortcuts) {}



    void outputProperties(JNIEnv *const env, jintArray outCodePoints, jbooleanArray outFlags,

            jintArray outProbability, jintArray outHistoricalInfo, jobject outShortcutTargets,

            jobject outShortcutProbabilities) const;

            jintArray outProbabilityInfo, jobject outBigramTargets, jobject outBigramProbabilities,

            jobject outShortcutTargets, jobject outShortcutProbabilities) const;



 private:

    DISALLOW_ASSIGNMENT_OPERATOR(WordProperty);