aboutsummaryrefslogtreecommitdiffstats
path: root/java/src/com/android/inputmethod/latin/BinaryDictionary.java
diff options
context:
space:
mode:
Diffstat (limited to 'java/src/com/android/inputmethod/latin/BinaryDictionary.java')
-rw-r--r--java/src/com/android/inputmethod/latin/BinaryDictionary.java95
1 files changed, 62 insertions, 33 deletions
diff --git a/java/src/com/android/inputmethod/latin/BinaryDictionary.java b/java/src/com/android/inputmethod/latin/BinaryDictionary.java
index e66cfca49..00eb57c9f 100644
--- a/java/src/com/android/inputmethod/latin/BinaryDictionary.java
+++ b/java/src/com/android/inputmethod/latin/BinaryDictionary.java
@@ -22,12 +22,13 @@ 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;
import com.android.inputmethod.latin.utils.LanguageModelParam;
import com.android.inputmethod.latin.utils.StringUtils;
-import com.android.inputmethod.latin.utils.UnigramProperty;
+import com.android.inputmethod.latin.utils.WordProperty;
import java.io.File;
import java.util.ArrayList;
@@ -61,18 +62,19 @@ public final class BinaryDictionary extends Dictionary {
public static final int NOT_A_VALID_TIMESTAMP = -1;
- // Format to get unigram flags from native side via getUnigramPropertyNative().
- private static final int FORMAT_UNIGRAM_PROPERTY_OUTPUT_FLAG_COUNT = 4;
- private static final int FORMAT_UNIGRAM_PROPERTY_IS_NOT_A_WORD_INDEX = 0;
- private static final int FORMAT_UNIGRAM_PROPERTY_IS_BLACKLISTED_INDEX = 1;
- private static final int FORMAT_UNIGRAM_PROPERTY_HAS_BIGRAMS_INDEX = 2;
- private static final int FORMAT_UNIGRAM_PROPERTY_HAS_SHORTCUTS_INDEX = 3;
+ // Format to get unigram flags from native side via getWordPropertyNative().
+ private static final int FORMAT_WORD_PROPERTY_OUTPUT_FLAG_COUNT = 4;
+ private static final int FORMAT_WORD_PROPERTY_IS_NOT_A_WORD_INDEX = 0;
+ private static final int FORMAT_WORD_PROPERTY_IS_BLACKLISTED_INDEX = 1;
+ 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 getUnigramPropertyNative().
- private static final int FORMAT_UNIGRAM_PROPERTY_OUTPUT_HISTORICAL_INFO_COUNT = 3;
- private static final int FORMAT_UNIGRAM_PROPERTY_TIMESTAMP_INDEX = 0;
- private static final int FORMAT_UNIGRAM_PROPERTY_LEVEL_INDEX = 1;
- private static final int FORMAT_UNIGRAM_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;
@@ -143,10 +145,11 @@ public final class BinaryDictionary extends Dictionary {
private static native int getFormatVersionNative(long dict);
private static native int getProbabilityNative(long dict, int[] word);
private static native int getBigramProbabilityNative(long dict, int[] word0, int[] word1);
- private static native void getUnigramPropertyNative(long dict, int[] word,
- int[] outCodePoints, boolean[] outFlags, int[] outProbability,
- int[] outHistoricalInfo, ArrayList<int[]> outShortcutTargets,
- ArrayList<Integer> outShortcutProbabilities);
+ private static native void getWordPropertyNative(long dict, int[] word,
+ 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,
@@ -306,29 +309,56 @@ public final class BinaryDictionary extends Dictionary {
}
@UsedForTesting
- public UnigramProperty getUnigramProperty(final String word) {
+ public WordProperty getWordProperty(final String word) {
if (TextUtils.isEmpty(word)) {
return null;
}
final int[] codePoints = StringUtils.toCodePointArray(word);
final int[] outCodePoints = new int[MAX_WORD_LENGTH];
- final boolean[] outFlags = new boolean[FORMAT_UNIGRAM_PROPERTY_OUTPUT_FLAG_COUNT];
- final int[] outProbability = new int[1];
- final int[] outHistoricalInfo =
- new int[FORMAT_UNIGRAM_PROPERTY_OUTPUT_HISTORICAL_INFO_COUNT];
+ final boolean[] outFlags = new boolean[FORMAT_WORD_PROPERTY_OUTPUT_FLAG_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();
- getUnigramPropertyNative(mNativeDict, codePoints, outCodePoints, outFlags, outProbability,
- outHistoricalInfo, outShortcutTargets, outShortcutProbabilities);
- return new UnigramProperty(codePoints,
- outFlags[FORMAT_UNIGRAM_PROPERTY_IS_NOT_A_WORD_INDEX],
- outFlags[FORMAT_UNIGRAM_PROPERTY_IS_BLACKLISTED_INDEX],
- outFlags[FORMAT_UNIGRAM_PROPERTY_HAS_BIGRAMS_INDEX],
- outFlags[FORMAT_UNIGRAM_PROPERTY_HAS_SHORTCUTS_INDEX], outProbability[0],
- outHistoricalInfo[FORMAT_UNIGRAM_PROPERTY_TIMESTAMP_INDEX],
- outHistoricalInfo[FORMAT_UNIGRAM_PROPERTY_LEVEL_INDEX],
- outHistoricalInfo[FORMAT_UNIGRAM_PROPERTY_COUNT_INDEX],
- 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], outProbabilityInfo,
+ outBigramTargets, outBigramProbabilityInfo, outShortcutTargets,
+ 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.
@@ -379,7 +409,6 @@ public final class BinaryDictionary extends Dictionary {
return;
}
}
-
}
private void reopen() {
generated by cgit v1.2.3-83-g751a (git 2.47.1) at 2025-05-25 04:56:38 +0000