aboutsummaryrefslogtreecommitdiffstats
path: root/native/jni/src
diff options
context:
space:
mode:
Diffstat (limited to 'native/jni/src')
-rw-r--r--native/jni/src/defines.h3
-rw-r--r--native/jni/src/suggest/core/dictionary/bigram_dictionary.cpp21
-rw-r--r--native/jni/src/suggest/core/dictionary/bigram_dictionary.h6
-rw-r--r--native/jni/src/suggest/core/dictionary/dictionary.cpp13
-rw-r--r--native/jni/src/suggest/core/dictionary/dictionary.h8
-rw-r--r--native/jni/src/suggest/core/session/dic_traverse_session.cpp21
-rw-r--r--native/jni/src/suggest/core/session/dic_traverse_session.h15
-rw-r--r--native/jni/src/suggest/core/session/prev_words_info.h65
8 files changed, 115 insertions, 37 deletions
diff --git a/native/jni/src/defines.h b/native/jni/src/defines.h
index a80c97530..24d04e51f 100644
--- a/native/jni/src/defines.h
+++ b/native/jni/src/defines.h
@@ -336,6 +336,9 @@ static inline void prof_out(void) {
#define MAX_POINTER_COUNT 1
#define MAX_POINTER_COUNT_G 2
+// (MAX_PREV_WORD_COUNT_FOR_N_GRAM + 1)-gram is supported.
+#define MAX_PREV_WORD_COUNT_FOR_N_GRAM 1
+
#define DISALLOW_DEFAULT_CONSTRUCTOR(TypeName) \
TypeName() = delete
diff --git a/native/jni/src/suggest/core/dictionary/bigram_dictionary.cpp b/native/jni/src/suggest/core/dictionary/bigram_dictionary.cpp
index f793363a8..847fa1b02 100644
--- a/native/jni/src/suggest/core/dictionary/bigram_dictionary.cpp
+++ b/native/jni/src/suggest/core/dictionary/bigram_dictionary.cpp
@@ -26,6 +26,7 @@
#include "suggest/core/dictionary/dictionary.h"
#include "suggest/core/policy/dictionary_structure_with_buffer_policy.h"
#include "suggest/core/result/suggestion_results.h"
+#include "suggest/core/session/prev_words_info.h"
#include "utils/char_utils.h"
namespace latinime {
@@ -42,19 +43,18 @@ BigramDictionary::~BigramDictionary() {
}
/* Parameters :
- * prevWord: the word before, the one for which we need to look up bigrams.
- * prevWordLength: its length.
+ * prevWordsInfo: Information of previous words to get the predictions.
* outSuggestionResults: SuggestionResults to put the predictions.
*/
-void BigramDictionary::getPredictions(const int *prevWord, const int prevWordLength,
+void BigramDictionary::getPredictions(const PrevWordsInfo *const prevWordsInfo,
SuggestionResults *const outSuggestionResults) const {
- int pos = getBigramListPositionForWord(prevWord, prevWordLength,
- false /* forceLowerCaseSearch */);
+ int pos = getBigramListPositionForWord(prevWordsInfo->getPrevWordCodePoints(),
+ prevWordsInfo->getPrevWordCodePointCount(), false /* forceLowerCaseSearch */);
// getBigramListPositionForWord returns 0 if this word isn't in the dictionary or has no bigrams
if (NOT_A_DICT_POS == pos) {
// If no bigrams for this exact word, search again in lower case.
- pos = getBigramListPositionForWord(prevWord, prevWordLength,
- true /* forceLowerCaseSearch */);
+ pos = getBigramListPositionForWord(prevWordsInfo->getPrevWordCodePoints(),
+ prevWordsInfo->getPrevWordCodePointCount(), true /* forceLowerCaseSearch */);
}
// If still no bigrams, we really don't have them!
if (NOT_A_DICT_POS == pos) return;
@@ -96,9 +96,10 @@ int BigramDictionary::getBigramListPositionForWord(const int *prevWord, const in
return mDictionaryStructurePolicy->getBigramsPositionOfPtNode(pos);
}
-int BigramDictionary::getBigramProbability(const int *word0, int length0, const int *word1,
- int length1) const {
- int pos = getBigramListPositionForWord(word0, length0, false /* forceLowerCaseSearch */);
+int BigramDictionary::getBigramProbability(const PrevWordsInfo *const prevWordsInfo,
+ const int *word1, int length1) const {
+ int pos = getBigramListPositionForWord(prevWordsInfo->getPrevWordCodePoints(),
+ prevWordsInfo->getPrevWordCodePointCount(), false /* forceLowerCaseSearch */);
// getBigramListPositionForWord returns 0 if this word isn't in the dictionary or has no bigrams
if (NOT_A_DICT_POS == pos) return NOT_A_PROBABILITY;
int nextWordPos = mDictionaryStructurePolicy->getTerminalPtNodePositionOfWord(word1, length1,
diff --git a/native/jni/src/suggest/core/dictionary/bigram_dictionary.h b/native/jni/src/suggest/core/dictionary/bigram_dictionary.h
index 12aaf20d3..bd3aed1bd 100644
--- a/native/jni/src/suggest/core/dictionary/bigram_dictionary.h
+++ b/native/jni/src/suggest/core/dictionary/bigram_dictionary.h
@@ -22,15 +22,17 @@
namespace latinime {
class DictionaryStructureWithBufferPolicy;
+class PrevWordsInfo;
class SuggestionResults;
class BigramDictionary {
public:
BigramDictionary(const DictionaryStructureWithBufferPolicy *const dictionaryStructurePolicy);
- void getPredictions(const int *word, int length,
+ void getPredictions(const PrevWordsInfo *const prevWordsInfo,
SuggestionResults *const outSuggestionResults) const;
- int getBigramProbability(const int *word1, int length1, const int *word2, int length2) const;
+ int getBigramProbability(const PrevWordsInfo *const prevWordsInfo,
+ const int *word1, int length1) const;
~BigramDictionary();
private:
diff --git a/native/jni/src/suggest/core/dictionary/dictionary.cpp b/native/jni/src/suggest/core/dictionary/dictionary.cpp
index aa52e63c0..c860d82af 100644
--- a/native/jni/src/suggest/core/dictionary/dictionary.cpp
+++ b/native/jni/src/suggest/core/dictionary/dictionary.cpp
@@ -44,11 +44,11 @@ Dictionary::Dictionary(JNIEnv *env, DictionaryStructureWithBufferPolicy::Structu
void Dictionary::getSuggestions(ProximityInfo *proximityInfo, DicTraverseSession *traverseSession,
int *xcoordinates, int *ycoordinates, int *times, int *pointerIds, int *inputCodePoints,
- int inputSize, int *prevWordCodePoints, int prevWordLength,
+ int inputSize, const PrevWordsInfo *const prevWordsInfo,
const SuggestOptions *const suggestOptions, const float languageWeight,
SuggestionResults *const outSuggestionResults) const {
TimeKeeper::setCurrentTime();
- traverseSession->init(this, prevWordCodePoints, prevWordLength, suggestOptions);
+ traverseSession->init(this, prevWordsInfo, suggestOptions);
const auto &suggest = suggestOptions->isGesture() ? mGestureSuggest : mTypingSuggest;
suggest->getSuggestions(proximityInfo, traverseSession, xcoordinates,
ycoordinates, times, pointerIds, inputCodePoints, inputSize,
@@ -58,11 +58,10 @@ void Dictionary::getSuggestions(ProximityInfo *proximityInfo, DicTraverseSession
}
}
-void Dictionary::getPredictions(const int *word, int length,
+void Dictionary::getPredictions(const PrevWordsInfo *const prevWordsInfo,
SuggestionResults *const outSuggestionResults) const {
TimeKeeper::setCurrentTime();
- if (length <= 0) return;
- mBigramDictionary.getPredictions(word, length, outSuggestionResults);
+ mBigramDictionary.getPredictions(prevWordsInfo, outSuggestionResults);
}
int Dictionary::getProbability(const int *word, int length) const {
@@ -75,10 +74,10 @@ int Dictionary::getProbability(const int *word, int length) const {
return getDictionaryStructurePolicy()->getUnigramProbabilityOfPtNode(pos);
}
-int Dictionary::getBigramProbability(const int *word0, int length0, const int *word1,
+int Dictionary::getBigramProbability(const PrevWordsInfo *const prevWordsInfo, const int *word1,
int length1) const {
TimeKeeper::setCurrentTime();
- return mBigramDictionary.getBigramProbability(word0, length0, word1, length1);
+ return mBigramDictionary.getBigramProbability(prevWordsInfo, word1, length1);
}
void Dictionary::addUnigramWord(const int *const word, const int length,
diff --git a/native/jni/src/suggest/core/dictionary/dictionary.h b/native/jni/src/suggest/core/dictionary/dictionary.h
index f0a7e5b6a..b63c61fbb 100644
--- a/native/jni/src/suggest/core/dictionary/dictionary.h
+++ b/native/jni/src/suggest/core/dictionary/dictionary.h
@@ -31,6 +31,7 @@ namespace latinime {
class DictionaryStructureWithBufferPolicy;
class DicTraverseSession;
+class PrevWordsInfo;
class ProximityInfo;
class SuggestionResults;
class SuggestOptions;
@@ -62,16 +63,17 @@ class Dictionary {
void getSuggestions(ProximityInfo *proximityInfo, DicTraverseSession *traverseSession,
int *xcoordinates, int *ycoordinates, int *times, int *pointerIds, int *inputCodePoints,
- int inputSize, int *prevWordCodePoints, int prevWordLength,
+ int inputSize, const PrevWordsInfo *const prevWordsInfo,
const SuggestOptions *const suggestOptions, const float languageWeight,
SuggestionResults *const outSuggestionResults) const;
- void getPredictions(const int *word, int length,
+ void getPredictions(const PrevWordsInfo *const prevWordsInfo,
SuggestionResults *const outSuggestionResults) const;
int getProbability(const int *word, int length) const;
- int getBigramProbability(const int *word0, int length0, const int *word1, int length1) const;
+ int getBigramProbability(const PrevWordsInfo *const prevWordsInfo,
+ const int *word1, int length1) const;
void addUnigramWord(const int *const codePoints, const int codePointCount,
const UnigramProperty *const unigramProperty);
diff --git a/native/jni/src/suggest/core/session/dic_traverse_session.cpp b/native/jni/src/suggest/core/session/dic_traverse_session.cpp
index 77b634e07..b9e9db719 100644
--- a/native/jni/src/suggest/core/session/dic_traverse_session.cpp
+++ b/native/jni/src/suggest/core/session/dic_traverse_session.cpp
@@ -20,6 +20,7 @@
#include "suggest/core/dictionary/dictionary.h"
#include "suggest/core/policy/dictionary_header_structure_policy.h"
#include "suggest/core/policy/dictionary_structure_with_buffer_policy.h"
+#include "suggest/core/session/prev_words_info.h"
namespace latinime {
@@ -28,24 +29,26 @@ namespace latinime {
const int DicTraverseSession::DICTIONARY_SIZE_THRESHOLD_TO_USE_LARGE_CACHE_FOR_SUGGESTION =
256 * 1024;
-void DicTraverseSession::init(const Dictionary *const dictionary, const int *prevWord,
- int prevWordLength, const SuggestOptions *const suggestOptions) {
+void DicTraverseSession::init(const Dictionary *const dictionary,
+ const PrevWordsInfo *const prevWordsInfo, const SuggestOptions *const suggestOptions) {
mDictionary = dictionary;
mMultiWordCostMultiplier = getDictionaryStructurePolicy()->getHeaderStructurePolicy()
->getMultiWordCostMultiplier();
mSuggestOptions = suggestOptions;
- if (!prevWord) {
- mPrevWordPtNodePos = NOT_A_DICT_POS;
+ if (!prevWordsInfo->getPrevWordCodePoints()) {
+ mPrevWordsPtNodePos[0] = NOT_A_DICT_POS;
return;
}
// TODO: merge following similar calls to getTerminalPosition into one case-insensitive call.
- mPrevWordPtNodePos = getDictionaryStructurePolicy()->getTerminalPtNodePositionOfWord(
- prevWord, prevWordLength, false /* forceLowerCaseSearch */);
- if (mPrevWordPtNodePos == NOT_A_DICT_POS) {
+ mPrevWordsPtNodePos[0] = getDictionaryStructurePolicy()->getTerminalPtNodePositionOfWord(
+ prevWordsInfo->getPrevWordCodePoints(), prevWordsInfo->getPrevWordCodePointCount(),
+ false /* forceLowerCaseSearch */);
+ if (mPrevWordsPtNodePos[0] == NOT_A_DICT_POS) {
// Check bigrams for lower-cased previous word if original was not found. Useful for
// auto-capitalized words like "The [current_word]".
- mPrevWordPtNodePos = getDictionaryStructurePolicy()->getTerminalPtNodePositionOfWord(
- prevWord, prevWordLength, true /* forceLowerCaseSearch */);
+ mPrevWordsPtNodePos[0] = getDictionaryStructurePolicy()->getTerminalPtNodePositionOfWord(
+ prevWordsInfo->getPrevWordCodePoints(), prevWordsInfo->getPrevWordCodePointCount(),
+ true /* forceLowerCaseSearch */);
}
}
diff --git a/native/jni/src/suggest/core/session/dic_traverse_session.h b/native/jni/src/suggest/core/session/dic_traverse_session.h
index 50bffebac..90aff06c3 100644
--- a/native/jni/src/suggest/core/session/dic_traverse_session.h
+++ b/native/jni/src/suggest/core/session/dic_traverse_session.h
@@ -29,6 +29,7 @@ namespace latinime {
class Dictionary;
class DictionaryStructureWithBufferPolicy;
+class PrevWordsInfo;
class ProximityInfo;
class SuggestOptions;
@@ -49,18 +50,20 @@ class DicTraverseSession {
}
AK_FORCE_INLINE DicTraverseSession(JNIEnv *env, jstring localeStr, bool usesLargeCache)
- : mPrevWordPtNodePos(NOT_A_DICT_POS), mProximityInfo(nullptr),
- mDictionary(nullptr), mSuggestOptions(nullptr), mDicNodesCache(usesLargeCache),
- mMultiBigramMap(), mInputSize(0), mMaxPointerCount(1),
+ : mProximityInfo(nullptr), mDictionary(nullptr), mSuggestOptions(nullptr),
+ mDicNodesCache(usesLargeCache), mMultiBigramMap(), mInputSize(0), mMaxPointerCount(1),
mMultiWordCostMultiplier(1.0f) {
// NOTE: mProximityInfoStates is an array of instances.
// No need to initialize it explicitly here.
+ for (size_t i = 0; i < NELEMS(mPrevWordsPtNodePos); ++i) {
+ mPrevWordsPtNodePos[i] = NOT_A_DICT_POS;
+ }
}
// Non virtual inline destructor -- never inherit this class
AK_FORCE_INLINE ~DicTraverseSession() {}
- void init(const Dictionary *dictionary, const int *prevWord, int prevWordLength,
+ void init(const Dictionary *dictionary, const PrevWordsInfo *const prevWordsInfo,
const SuggestOptions *const suggestOptions);
// TODO: Remove and merge into init
void setupForGetSuggestions(const ProximityInfo *pInfo, const int *inputCodePoints,
@@ -76,7 +79,7 @@ class DicTraverseSession {
//--------------------
const ProximityInfo *getProximityInfo() const { return mProximityInfo; }
const SuggestOptions *getSuggestOptions() const { return mSuggestOptions; }
- int getPrevWordPtNodePos() const { return mPrevWordPtNodePos; }
+ int getPrevWordPtNodePos() const { return mPrevWordsPtNodePos[0]; }
DicNodesCache *getDicTraverseCache() { return &mDicNodesCache; }
MultiBigramMap *getMultiBigramMap() { return &mMultiBigramMap; }
const ProximityInfoState *getProximityInfoState(int id) const {
@@ -163,7 +166,7 @@ class DicTraverseSession {
const int *const inputYs, const int *const times, const int *const pointerIds,
const int inputSize, const float maxSpatialDistance, const int maxPointerCount);
- int mPrevWordPtNodePos;
+ int mPrevWordsPtNodePos[MAX_PREV_WORD_COUNT_FOR_N_GRAM];
const ProximityInfo *mProximityInfo;
const Dictionary *mDictionary;
const SuggestOptions *mSuggestOptions;
diff --git a/native/jni/src/suggest/core/session/prev_words_info.h b/native/jni/src/suggest/core/session/prev_words_info.h
new file mode 100644
index 000000000..bc685945e
--- /dev/null
+++ b/native/jni/src/suggest/core/session/prev_words_info.h
@@ -0,0 +1,65 @@
+/*
+ * Copyright (C) 2014 The Android Open Source Project
+ *
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+#ifndef LATINIME_PREV_WORDS_INFO_H
+#define LATINIME_PREV_WORDS_INFO_H
+
+#include "defines.h"
+
+namespace latinime {
+
+// TODO: Support n-gram.
+// TODO: Support beginning of sentence.
+// This class does not take ownership of any code point buffers.
+class PrevWordsInfo {
+ public:
+ // No prev word information.
+ PrevWordsInfo() {
+ clear();
+ }
+
+ PrevWordsInfo(const int *const prevWordCodePoints, const int prevWordCodePointCount,
+ const bool isBeginningOfSentence) {
+ clear();
+ mPrevWordCodePoints[0] = prevWordCodePoints;
+ mPrevWordCodePointCount[0] = prevWordCodePointCount;
+ mIsBeginningOfSentence[0] = isBeginningOfSentence;
+ }
+ const int *getPrevWordCodePoints() const {
+ return mPrevWordCodePoints[0];
+ }
+
+ int getPrevWordCodePointCount() const {
+ return mPrevWordCodePointCount[0];
+ }
+
+ private:
+ DISALLOW_COPY_AND_ASSIGN(PrevWordsInfo);
+
+ void clear() {
+ for (size_t i = 0; i < NELEMS(mPrevWordCodePoints); ++i) {
+ mPrevWordCodePoints[i] = nullptr;
+ mPrevWordCodePointCount[i] = 0;
+ mIsBeginningOfSentence[i] = false;
+ }
+ }
+
+ const int *mPrevWordCodePoints[MAX_PREV_WORD_COUNT_FOR_N_GRAM];
+ int mPrevWordCodePointCount[MAX_PREV_WORD_COUNT_FOR_N_GRAM];
+ bool mIsBeginningOfSentence[MAX_PREV_WORD_COUNT_FOR_N_GRAM];
+};
+} // namespace latinime
+#endif // LATINIME_PREV_WORDS_INFO_H
generated by cgit v1.2.3-83-g751a (git 2.47.1) at 2025-06-26 06:34:17 +0000