diff --git a/native/jni/src/suggest/policyimpl/dictionary/structure/v4/ver4_patricia_trie_policy.cpp b/native/jni/src/suggest/policyimpl/dictionary/structure/v4/ver4_patricia_trie_policy.cpp
index 1336a6229..d537711b0 100644
--- a/native/jni/src/suggest/policyimpl/dictionary/structure/v4/ver4_patricia_trie_policy.cpp
+++ b/native/jni/src/suggest/policyimpl/dictionary/structure/v4/ver4_patricia_trie_policy.cpp
@@ -310,30 +310,32 @@ bool Ver4PatriciaTriePolicy::addNgramEntry(const PrevWordsInfo *const prevWordsI
     if (prevWordIds.empty()) {
         return false;
     }
-    // TODO: Support N-gram.
-    if (prevWordIds[0] == NOT_A_WORD_ID) {
-        if (prevWordsInfo->isNthPrevWordBeginningOfSentence(1 /* n */)) {
-            const std::vector<UnigramProperty::ShortcutProperty> shortcuts;
-            const UnigramProperty beginningOfSentenceUnigramProperty(
-                    true /* representsBeginningOfSentence */, true /* isNotAWord */,
-                    false /* isBlacklisted */, MAX_PROBABILITY /* probability */,
-                    NOT_A_TIMESTAMP /* timestamp */, 0 /* level */, 0 /* count */, &shortcuts);
-            if (!addUnigramEntry(prevWordsInfo->getNthPrevWordCodePoints(1 /* n */),
-                    &beginningOfSentenceUnigramProperty)) {
-                AKLOGE("Cannot add unigram entry for the beginning-of-sentence.");
-                return false;
-            }
-            // Refresh word ids.
-            prevWordsInfo->getPrevWordIds(this, &prevWordIdArray, false /* tryLowerCaseSearch */);
-        } else {
+    for (size_t i = 0; i < prevWordIds.size(); ++i) {
+        if (prevWordIds[i] != NOT_A_WORD_ID) {
+            continue;
+        }
+        if (!prevWordsInfo->isNthPrevWordBeginningOfSentence(i + 1 /* n */)) {
             return false;
         }
+        const std::vector<UnigramProperty::ShortcutProperty> shortcuts;
+        const UnigramProperty beginningOfSentenceUnigramProperty(
+                true /* representsBeginningOfSentence */, true /* isNotAWord */,
+                false /* isBlacklisted */, MAX_PROBABILITY /* probability */,
+                NOT_A_TIMESTAMP /* timestamp */, 0 /* level */, 0 /* count */, &shortcuts);
+        if (!addUnigramEntry(prevWordsInfo->getNthPrevWordCodePoints(1 /* n */),
+                &beginningOfSentenceUnigramProperty)) {
+            AKLOGE("Cannot add unigram entry for the beginning-of-sentence.");
+            return false;
+        }
+        // Refresh word ids.
+        prevWordsInfo->getPrevWordIds(this, &prevWordIdArray, false /* tryLowerCaseSearch */);
     }
     const int wordId = getWordId(CodePointArrayView(*bigramProperty->getTargetCodePoints()),
             false /* forceLowerCaseSearch */);
     if (wordId == NOT_A_WORD_ID) {
         return false;
     }
+    // TODO: Support N-gram.
     bool addedNewEntry = false;
     WordIdArray<MAX_PREV_WORD_COUNT_FOR_N_GRAM> prevWordsPtNodePos;
     for (size_t i = 0; i < prevWordsPtNodePos.size(); ++i) {
@@ -375,8 +377,7 @@ bool Ver4PatriciaTriePolicy::removeNgramEntry(const PrevWordsInfo *const prevWor
     WordIdArray<MAX_PREV_WORD_COUNT_FOR_N_GRAM> prevWordIdArray;
     const WordIdArrayView prevWordIds = prevWordsInfo->getPrevWordIds(this, &prevWordIdArray,
             false /* tryLowerCaseSerch */);
-    // TODO: Support N-gram.
-    if (prevWordIds.empty() || prevWordIds[0] == NOT_A_WORD_ID) {
+    if (prevWordIds.empty() || prevWordIds.contains(NOT_A_WORD_ID)) {
         return false;
     }
     const int wordId = getWordId(wordCodePoints, false /* forceLowerCaseSearch */);
diff --git a/native/jni/src/utils/int_array_view.h b/native/jni/src/utils/int_array_view.h
index caa13d976..cc5f328ba 100644
--- a/native/jni/src/utils/int_array_view.h
+++ b/native/jni/src/utils/int_array_view.h
@@ -17,6 +17,7 @@
 #ifndef LATINIME_INT_ARRAY_VIEW_H
 #define LATINIME_INT_ARRAY_VIEW_H
 
+#include <algorithm>
 #include <array>
 #include <cstdint>
 #include <cstring>
@@ -92,12 +93,16 @@ class IntArrayView {
         return mPtr + mSize;
     }
 
+    AK_FORCE_INLINE bool contains(const int value) const {
+        return std::find(begin(), end(), value) != end();
+    }
+
     // Returns the view whose size is smaller than or equal to the given count.
-    const IntArrayView limit(const size_t maxSize) const {
+    AK_FORCE_INLINE const IntArrayView limit(const size_t maxSize) const {
         return IntArrayView(mPtr, std::min(maxSize, mSize));
     }
 
-    const IntArrayView skip(const size_t n) const {
+    AK_FORCE_INLINE const IntArrayView skip(const size_t n) const {
         if (mSize <= n) {
             return IntArrayView();
         }
diff --git a/native/jni/tests/utils/int_array_view_test.cpp b/native/jni/tests/utils/int_array_view_test.cpp
index 3bc294cdd..934e27e1c 100644
--- a/native/jni/tests/utils/int_array_view_test.cpp
+++ b/native/jni/tests/utils/int_array_view_test.cpp
@@ -58,6 +58,19 @@ TEST(IntArrayViewTest, TestConstructFromObject) {
     EXPECT_EQ(object, intArrayView[0]);
 }
 
+TEST(IntArrayViewTest, TestContains) {
+    EXPECT_FALSE(IntArrayView().contains(0));
+    EXPECT_FALSE(IntArrayView().contains(1));
+
+    const std::vector<int> intVector = {3, 2, 1, 0, -1, -2};
+    IntArrayView intArrayView(intVector);
+    EXPECT_TRUE(intArrayView.contains(0));
+    EXPECT_TRUE(intArrayView.contains(3));
+    EXPECT_TRUE(intArrayView.contains(-2));
+    EXPECT_FALSE(intArrayView.contains(-3));
+    EXPECT_FALSE(intArrayView.limit(0).contains(3));
+}
+
 TEST(IntArrayViewTest, TestLimit) {
     const std::vector<int> intVector = {3, 2, 1, 0, -1, -2};
     IntArrayView intArrayView(intVector);