Add a jni method to iterate words in a dictionary.

Bug: 12810574
Change-Id: I87b13559765f0262dfd06c2ba40d43af6437ea7f

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

@@ -22,6 +22,7 @@ import android.util.SparseArray;
 import com.android.inputmethod.annotations.UsedForTesting;
 import com.android.inputmethod.keyboard.ProximityInfo;
 import com.android.inputmethod.latin.SuggestedWords.SuggestedWordInfo;
+import com.android.inputmethod.latin.makedict.Word;
 import com.android.inputmethod.latin.settings.NativeSuggestOptions;
 import com.android.inputmethod.latin.utils.CollectionUtils;
 import com.android.inputmethod.latin.utils.JniUtils;
@@ -148,6 +149,7 @@ public final class BinaryDictionary extends Dictionary {
             int[] outCodePoints, boolean[] outFlags, int[] outProbabilityInfo,
             ArrayList<int[]> outBigramTargets, ArrayList<int[]> outBigramProbabilityInfo,
             ArrayList<int[]> outShortcutTargets, ArrayList<Integer> outShortcutProbabilities);
+    private static native int getNextWordNative(long dict, int token, int[] outCodePoints);
     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,
@@ -332,6 +334,33 @@ public final class BinaryDictionary extends Dictionary {
                 outShortcutProbabilities);
     }
 
+    public static class GetNextWordPropertyResult {
+        public WordProperty mWordProperty;
+        public int mNextToken;
+
+        public GetNextWordPropertyResult(final WordProperty wordPreperty, final int nextToken) {
+            mWordProperty = wordPreperty;
+            mNextToken = nextToken;
+        }
+    }
+
+    /**
+     * Method to iterate all words in the dictionary for makedict.
+     * If token is 0, this method newly starts iterating the dictionary.
+     */
+    @UsedForTesting
+    public GetNextWordPropertyResult getNextWordProperty(final int token) {
+        final int[] codePoints = new int[MAX_WORD_LENGTH];
+        final int nextToken = getNextWordNative(mNativeDict, token, codePoints);
+        int len = 0;
+        // codePoints is null-terminated if its length is shorter than the array length.
+        while (len < MAX_WORD_LENGTH && codePoints[len] != 0) {
+            ++len;
+        }
+        final String word = new String(mOutputCodePoints, 0, len);
+        return new GetNextWordPropertyResult(getWordProperty(word), nextToken);
+    }
+
     // Add a unigram entry to binary dictionary with unigram attributes in native code.
     public void addUnigramWord(final String word, final int probability,
             final String shortcutTarget, final int shortcutProbability, final boolean isNotAWord,
@@ -380,7 +409,6 @@ public final class BinaryDictionary extends Dictionary {
                 return;
             }
         }
-
     }
 
     private void reopen() {

native/jni/com_android_inputmethod_latin_BinaryDictionary.cpp

@@ -260,6 +260,26 @@ static jint latinime_BinaryDictionary_getBigramProbability(JNIEnv *env, jclass c
             word1Length);
 }
 
+// Method to iterate all words in the dictionary for makedict.
+// If token is 0, this method newly starts iterating the dictionary. This method returns 0 when
+// the dictionary does not have a next word.
+static jint latinime_BinaryDictionary_getNextWord(JNIEnv *env, jclass clazz,
+        jlong dict, jint token, jintArray outCodePoints) {
+    Dictionary *dictionary = reinterpret_cast<Dictionary *>(dict);
+    if (!dictionary) return 0;
+    const jsize outCodePointsLength = env->GetArrayLength(outCodePoints);
+    if (outCodePointsLength != MAX_WORD_LENGTH) {
+        AKLOGE("Invalid outCodePointsLength: %d", outCodePointsLength);
+        ASSERT(false);
+        return 0;
+    }
+    int wordCodePoints[outCodePointsLength];
+    memset(wordCodePoints, 0, sizeof(wordCodePoints));
+    const int nextToken = dictionary->getNextWordAndNextToken(token, wordCodePoints);
+    env->SetIntArrayRegion(outCodePoints, 0, outCodePointsLength, wordCodePoints);
+    return nextToken;
+}
+
 static void latinime_BinaryDictionary_getWordProperty(JNIEnv *env, jclass clazz,
         jlong dict, jintArray word, jintArray outCodePoints, jbooleanArray outFlags,
         jintArray outProbabilityInfo, jobject outBigramTargets, jobject outBigramProbabilityInfo,
@@ -526,6 +546,11 @@ static const JNINativeMethod sMethods[] = {
                 "Ljava/util/ArrayList;Ljava/util/ArrayList;)V"),
         reinterpret_cast<void *>(latinime_BinaryDictionary_getWordProperty)
     },
+    {
+        const_cast<char *>("getNextWordNative"),
+        const_cast<char *>("(JI[I)I"),
+        reinterpret_cast<void *>(latinime_BinaryDictionary_getNextWord)
+    },
     {
         const_cast<char *>("calcNormalizedScoreNative"),
         const_cast<char *>("([I[II)F"),

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

@@ -150,6 +150,12 @@ const WordProperty Dictionary::getWordProperty(const int *const codePoints,
             codePoints, codePointCount);
 }
 
+int Dictionary::getNextWordAndNextToken(const int token, int *const outCodePoints) {
+    TimeKeeper::setCurrentTime();
+    return mDictionaryStructureWithBufferPolicy.get()->getNextWordAndNextToken(
+            token, outCodePoints);
+}
+
 void Dictionary::logDictionaryInfo(JNIEnv *const env) const {
     int dictionaryIdCodePointBuffer[HEADER_ATTRIBUTE_BUFFER_SIZE];
     int versionStringCodePointBuffer[HEADER_ATTRIBUTE_BUFFER_SIZE];

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

@@ -96,6 +96,11 @@ class Dictionary {
 
     const WordProperty getWordProperty(const int *const codePoints, const int codePointCount);
 
+    // Method to iterate all words in the dictionary.
+    // The returned token has to be used to get the next word. If token is 0, this method newly
+    // starts iterating the dictionary.
+    int getNextWordAndNextToken(const int token, int *const outCodePoints);
+
     const DictionaryStructureWithBufferPolicy *getDictionaryStructurePolicy() const {
         return mDictionaryStructureWithBufferPolicy.get();
     }

native/jni/src/suggest/core/policy/dictionary_structure_with_buffer_policy.h

@@ -95,6 +95,11 @@ class DictionaryStructureWithBufferPolicy {
     virtual const WordProperty getWordProperty(const int *const codePonts,
             const int codePointCount) const = 0;
 
+    // Method to iterate all words in the dictionary.
+    // The returned token has to be used to get the next word. If token is 0, this method newly
+    // starts iterating the dictionary.
+    virtual int getNextWordAndNextToken(const int token, int *const outCodePoints) = 0;
+
  protected:
     DictionaryStructureWithBufferPolicy() {}
 

native/jni/src/suggest/policyimpl/dictionary/structure/v2/patricia_trie_policy.h

@@ -129,6 +129,11 @@ class PatriciaTriePolicy : public DictionaryStructureWithBufferPolicy {
         return WordProperty();
     }
 
+    int getNextWordAndNextToken(const int token, int *const outCodePoints) {
+        // getNextWordAndNextToken is not supported.
+        return 0;
+    }
+
  private:
     DISALLOW_IMPLICIT_CONSTRUCTORS(PatriciaTriePolicy);
 

native/jni/src/suggest/policyimpl/dictionary/structure/v4/ver4_patricia_trie_policy.cpp

@@ -392,4 +392,10 @@ const WordProperty Ver4PatriciaTriePolicy::getWordProperty(const int *const code
             historicalInfo->getCount(), &bigrams, &shortcuts);
 }
 
+int Ver4PatriciaTriePolicy::getNextWordAndNextToken(const int token,
+        int *const outCodePoints) {
+    // TODO: Implement.
+    return 0;
+}
+
 } // namespace latinime

native/jni/src/suggest/policyimpl/dictionary/structure/v4/ver4_patricia_trie_policy.h

@@ -109,6 +109,8 @@ class Ver4PatriciaTriePolicy : public DictionaryStructureWithBufferPolicy {
     const WordProperty getWordProperty(const int *const codePoints,
             const int codePointCount) const;
 
+    int getNextWordAndNextToken(const int token, int *const outCodePoints);
+
  private:
     DISALLOW_IMPLICIT_CONSTRUCTORS(Ver4PatriciaTriePolicy);