nyancrimew/LatinIME
1
0
Fork 0
Code Issues Pull requests Projects Releases Wiki Activity

Remove PrevWordsInfo.getBigramsIteratorForPrediction().

Browse source 
This changes the behavior of predictions for some rare cases.
For example, let's consider the case of the prev word is
"ABC", "ABC" and "abc" is in the dict, "ABC" doesn't have
any bigrams, and "abc" has bigrams.
Without this change, "abc"'s bigrams are used for
prediction. But, with this change, "abc"'s bigrams are not
used for prediction. This strategy is the same as the the
current way of bigram probability computation during
traversal (for suggestion).

Bug: 14425059
Change-Id: I857ff39aa70aea265efa651451a63b759bb47d48
This commit is contained in:
Keisuke Kuroyanagi 2014-07-30 17:26:26 +09:00
parent 2d57b3339a
commit bd1f59bda5
9 changed files with 46 additions and 76 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

16
native/jni/src/suggest/core/dictionary/dictionary.cpp
View file

@ -92,7 +92,11 @@ void Dictionary::getPredictions(const PrevWordsInfo *const prevWordsInfo,
TimeKeeper::setCurrentTime();
NgramListenerForPrediction listener(prevWordsInfo, outSuggestionResults,
mDictionaryStructureWithBufferPolicy.get());
mDictionaryStructureWithBufferPolicy->iterateNgramEntries(prevWordsInfo, &listener);
int prevWordsPtNodePos[MAX_PREV_WORD_COUNT_FOR_N_GRAM];
prevWordsInfo->getPrevWordsTerminalPtNodePos(
mDictionaryStructureWithBufferPolicy.get(), prevWordsPtNodePos,
true /* tryLowerCaseSearch */);
mDictionaryStructureWithBufferPolicy->iterateNgramEntries(prevWordsPtNodePos, &listener);
}
int Dictionary::getProbability(const int *word, int length) const {
@ -111,7 +115,15 @@ int Dictionary::getNgramProbability(const PrevWordsInfo *const prevWordsInfo, co
int nextWordPos = mDictionaryStructureWithBufferPolicy->getTerminalPtNodePositionOfWord(word,
length, false /* forceLowerCaseSearch */);
if (NOT_A_DICT_POS == nextWordPos) return NOT_A_PROBABILITY;
return getDictionaryStructurePolicy()->getProbabilityOfPtNode(prevWordsInfo, nextWordPos);
if (!prevWordsInfo) {
return getDictionaryStructurePolicy()->getProbabilityOfPtNode(
nullptr /* prevWordsPtNodePos */, nextWordPos);
}
int prevWordsPtNodePos[MAX_PREV_WORD_COUNT_FOR_N_GRAM];
prevWordsInfo->getPrevWordsTerminalPtNodePos(
mDictionaryStructureWithBufferPolicy.get(), prevWordsPtNodePos,
true /* tryLowerCaseSearch */);
return getDictionaryStructurePolicy()->getProbabilityOfPtNode(prevWordsPtNodePos, nextWordPos);
}
bool Dictionary::addUnigramEntry(const int *const word, const int length,

4
native/jni/src/suggest/core/policy/dictionary_structure_with_buffer_policy.h
View file

@ -59,10 +59,10 @@ class DictionaryStructureWithBufferPolicy {
virtual int getProbability(const int unigramProbability,
const int bigramProbability) const = 0;
virtual int getProbabilityOfPtNode(const PrevWordsInfo *const prevWordsInfo,
virtual int getProbabilityOfPtNode(const int *const prevWordsPtNodePos,
const int nodePos) const = 0;
virtual void iterateNgramEntries(const PrevWordsInfo *const prevWordsInfo,
virtual void iterateNgramEntries(const int *const prevWordsPtNodePos,
NgramListener *const listener) const = 0;
virtual int getShortcutPositionOfPtNode(const int nodePos) const = 0;

47
native/jni/src/suggest/core/session/prev_words_info.h
View file

@ -90,13 +90,6 @@ class PrevWordsInfo {
}
}
BinaryDictionaryBigramsIterator getBigramsIteratorForPrediction(
const DictionaryStructureWithBufferPolicy *const dictStructurePolicy) const {
return getBigramsIteratorForWordWithTryingLowerCaseSearch(
dictStructurePolicy, mPrevWordCodePoints[0], mPrevWordCodePointCount[0],
mIsBeginningOfSentence[0]);
}
// n is 1-indexed.
const int *getNthPrevWordCodePoints(const int n) const {
if (n <= 0 || n > MAX_PREV_WORD_COUNT_FOR_N_GRAM) {
@ -154,46 +147,6 @@ class PrevWordsInfo {
codePoints, codePointCount, true /* forceLowerCaseSearch */);
}
static BinaryDictionaryBigramsIterator getBigramsIteratorForWordWithTryingLowerCaseSearch(
const DictionaryStructureWithBufferPolicy *const dictStructurePolicy,
const int *const wordCodePoints, const int wordCodePointCount,
const bool isBeginningOfSentence) {
if (!dictStructurePolicy || !wordCodePoints || wordCodePointCount > MAX_WORD_LENGTH) {
return BinaryDictionaryBigramsIterator();
}
int codePoints[MAX_WORD_LENGTH];
int codePointCount = wordCodePointCount;
memmove(codePoints, wordCodePoints, sizeof(int) * codePointCount);
if (isBeginningOfSentence) {
codePointCount = CharUtils::attachBeginningOfSentenceMarker(codePoints,
codePointCount, MAX_WORD_LENGTH);
if (codePointCount <= 0) {
return BinaryDictionaryBigramsIterator();
}
}
BinaryDictionaryBigramsIterator bigramsIt = getBigramsIteratorForWord(dictStructurePolicy,
codePoints, codePointCount, false /* forceLowerCaseSearch */);
// getBigramsIteratorForWord returns an empty iterator if this word isn't in the dictionary
// or has no bigrams.
if (bigramsIt.hasNext()) {
return bigramsIt;
}
// If no bigrams for this exact word, search again in lower case.
return getBigramsIteratorForWord(dictStructurePolicy, codePoints,
codePointCount, true /* forceLowerCaseSearch */);
}
static BinaryDictionaryBigramsIterator getBigramsIteratorForWord(
const DictionaryStructureWithBufferPolicy *const dictStructurePolicy,
const int *wordCodePoints, const int wordCodePointCount,
const bool forceLowerCaseSearch) {
if (!wordCodePoints || wordCodePointCount <= 0) return BinaryDictionaryBigramsIterator();
const int terminalPtNodePos = dictStructurePolicy->getTerminalPtNodePositionOfWord(
wordCodePoints, wordCodePointCount, forceLowerCaseSearch);
if (NOT_A_DICT_POS == terminalPtNodePos) return BinaryDictionaryBigramsIterator();
return dictStructurePolicy->getBigramsIteratorOfPtNode(terminalPtNodePos);
}
void clear() {
for (size_t i = 0; i < NELEMS(mPrevWordCodePoints); ++i) {
mPrevWordCodePointCount[i] = 0;

14
native/jni/src/suggest/policyimpl/dictionary/structure/backward/v402/ver4_patricia_trie_policy.cpp
View file

@ -132,7 +132,7 @@ int Ver4PatriciaTriePolicy::getProbability(const int unigramProbability,
}
}
int Ver4PatriciaTriePolicy::getProbabilityOfPtNode(const PrevWordsInfo *const prevWordsInfo,
int Ver4PatriciaTriePolicy::getProbabilityOfPtNode(const int *const prevWordsPtNodePos,
const int ptNodePos) const {
if (ptNodePos == NOT_A_DICT_POS) {
return NOT_A_PROBABILITY;
@ -141,9 +141,9 @@ int Ver4PatriciaTriePolicy::getProbabilityOfPtNode(const PrevWordsInfo *const pr
if (ptNodeParams.isDeleted() || ptNodeParams.isBlacklisted() || ptNodeParams.isNotAWord()) {
return NOT_A_PROBABILITY;
}
if (prevWordsInfo) {
if (prevWordsPtNodePos) {
BinaryDictionaryBigramsIterator bigramsIt =
prevWordsInfo->getBigramsIteratorForPrediction(this /* dictStructurePolicy */);
getBigramsIteratorOfPtNode(prevWordsPtNodePos[0]);
while (bigramsIt.hasNext()) {
bigramsIt.next();
if (bigramsIt.getBigramPos() == ptNodePos
@ -156,10 +156,12 @@ int Ver4PatriciaTriePolicy::getProbabilityOfPtNode(const PrevWordsInfo *const pr
return getProbability(ptNodeParams.getProbability(), NOT_A_PROBABILITY);
}
void Ver4PatriciaTriePolicy::iterateNgramEntries(const PrevWordsInfo *const prevWordsInfo,
void Ver4PatriciaTriePolicy::iterateNgramEntries(const int *const prevWordsPtNodePos,
NgramListener *const listener) const {
BinaryDictionaryBigramsIterator bigramsIt = prevWordsInfo->getBigramsIteratorForPrediction(
this /* dictStructurePolicy */);
if (!prevWordsPtNodePos) {
return;
}
BinaryDictionaryBigramsIterator bigramsIt = getBigramsIteratorOfPtNode(prevWordsPtNodePos[0]);
while (bigramsIt.hasNext()) {
bigramsIt.next();
listener->onVisitEntry(bigramsIt.getProbability(), bigramsIt.getBigramPos());

5
native/jni/src/suggest/policyimpl/dictionary/structure/backward/v402/ver4_patricia_trie_policy.h
View file

@ -90,10 +90,9 @@ class Ver4PatriciaTriePolicy : public DictionaryStructureWithBufferPolicy {
int getProbability(const int unigramProbability, const int bigramProbability) const;
int getProbabilityOfPtNode(const PrevWordsInfo *const prevWordsInfo,
const int ptNodePos) const;
int getProbabilityOfPtNode(const int *const prevWordsPtNodePos, const int ptNodePos) const;
void iterateNgramEntries(const PrevWordsInfo *const prevWordsInfo,
void iterateNgramEntries(const int *const prevWordsPtNodePos,
NgramListener *const listener) const;
int getShortcutPositionOfPtNode(const int ptNodePos) const;

14
native/jni/src/suggest/policyimpl/dictionary/structure/v2/patricia_trie_policy.cpp
View file

@ -297,7 +297,7 @@ int PatriciaTriePolicy::getProbability(const int unigramProbability,
}
}
int PatriciaTriePolicy::getProbabilityOfPtNode(const PrevWordsInfo *const prevWordsInfo,
int PatriciaTriePolicy::getProbabilityOfPtNode(const int *const prevWordsPtNodePos,
const int ptNodePos) const {
if (ptNodePos == NOT_A_DICT_POS) {
return NOT_A_PROBABILITY;
@ -310,9 +310,9 @@ int PatriciaTriePolicy::getProbabilityOfPtNode(const PrevWordsInfo *const prevWo
// for shortcuts).
return NOT_A_PROBABILITY;
}
if (prevWordsInfo) {
if (prevWordsPtNodePos) {
BinaryDictionaryBigramsIterator bigramsIt =
prevWordsInfo->getBigramsIteratorForPrediction(this /* dictStructurePolicy */);
getBigramsIteratorOfPtNode(prevWordsPtNodePos[0]);
while (bigramsIt.hasNext()) {
bigramsIt.next();
if (bigramsIt.getBigramPos() == ptNodePos
@ -325,10 +325,12 @@ int PatriciaTriePolicy::getProbabilityOfPtNode(const PrevWordsInfo *const prevWo
return getProbability(ptNodeParams.getProbability(), NOT_A_PROBABILITY);
}
void PatriciaTriePolicy::iterateNgramEntries(const PrevWordsInfo *const prevWordsInfo,
void PatriciaTriePolicy::iterateNgramEntries(const int *const prevWordsPtNodePos,
NgramListener *const listener) const {
BinaryDictionaryBigramsIterator bigramsIt = prevWordsInfo->getBigramsIteratorForPrediction(
this /* dictStructurePolicy */);
if (!prevWordsPtNodePos) {
return;
}
BinaryDictionaryBigramsIterator bigramsIt = getBigramsIteratorOfPtNode(prevWordsPtNodePos[0]);
while (bigramsIt.hasNext()) {
bigramsIt.next();
listener->onVisitEntry(bigramsIt.getProbability(), bigramsIt.getBigramPos());

4
native/jni/src/suggest/policyimpl/dictionary/structure/v2/patricia_trie_policy.h
View file

@ -63,9 +63,9 @@ class PatriciaTriePolicy : public DictionaryStructureWithBufferPolicy {
int getProbability(const int unigramProbability, const int bigramProbability) const;
int getProbabilityOfPtNode(const PrevWordsInfo *const prevWordsInfo, const int ptNodePos) const;
int getProbabilityOfPtNode(const int *const prevWordsPtNodePos, const int ptNodePos) const;
void iterateNgramEntries(const PrevWordsInfo *const prevWordsInfo,
void iterateNgramEntries(const int *const prevWordsPtNodePos,
NgramListener *const listener) const;
int getShortcutPositionOfPtNode(const int ptNodePos) const;

14
native/jni/src/suggest/policyimpl/dictionary/structure/v4/ver4_patricia_trie_policy.cpp
View file

@ -122,7 +122,7 @@ int Ver4PatriciaTriePolicy::getProbability(const int unigramProbability,
}
}
int Ver4PatriciaTriePolicy::getProbabilityOfPtNode(const PrevWordsInfo *const prevWordsInfo,
int Ver4PatriciaTriePolicy::getProbabilityOfPtNode(const int *const prevWordsPtNodePos,
const int ptNodePos) const {
if (ptNodePos == NOT_A_DICT_POS) {
return NOT_A_PROBABILITY;
@ -131,9 +131,9 @@ int Ver4PatriciaTriePolicy::getProbabilityOfPtNode(const PrevWordsInfo *const pr
if (ptNodeParams.isDeleted() || ptNodeParams.isBlacklisted() || ptNodeParams.isNotAWord()) {
return NOT_A_PROBABILITY;
}
if (prevWordsInfo) {
if (prevWordsPtNodePos) {
BinaryDictionaryBigramsIterator bigramsIt =
prevWordsInfo->getBigramsIteratorForPrediction(this /* dictStructurePolicy */);
getBigramsIteratorOfPtNode(prevWordsPtNodePos[0]);
while (bigramsIt.hasNext()) {
bigramsIt.next();
if (bigramsIt.getBigramPos() == ptNodePos
@ -146,10 +146,12 @@ int Ver4PatriciaTriePolicy::getProbabilityOfPtNode(const PrevWordsInfo *const pr
return getProbability(ptNodeParams.getProbability(), NOT_A_PROBABILITY);
}
void Ver4PatriciaTriePolicy::iterateNgramEntries(const PrevWordsInfo *const prevWordsInfo,
void Ver4PatriciaTriePolicy::iterateNgramEntries(const int *const prevWordsPtNodePos,
NgramListener *const listener) const {
BinaryDictionaryBigramsIterator bigramsIt = prevWordsInfo->getBigramsIteratorForPrediction(
this /* dictStructurePolicy */);
if (!prevWordsPtNodePos) {
return;
}
BinaryDictionaryBigramsIterator bigramsIt = getBigramsIteratorOfPtNode(prevWordsPtNodePos[0]);
while (bigramsIt.hasNext()) {
bigramsIt.next();
listener->onVisitEntry(bigramsIt.getProbability(), bigramsIt.getBigramPos());

4
native/jni/src/suggest/policyimpl/dictionary/structure/v4/ver4_patricia_trie_policy.h
View file

@ -72,9 +72,9 @@ class Ver4PatriciaTriePolicy : public DictionaryStructureWithBufferPolicy {
int getProbability(const int unigramProbability, const int bigramProbability) const;
int getProbabilityOfPtNode(const PrevWordsInfo *const prevWordsInfo, const int ptNodePos) const;
int getProbabilityOfPtNode(const int *const prevWordsPtNodePos, const int ptNodePos) const;
void iterateNgramEntries(const PrevWordsInfo *const prevWordsInfo,
void iterateNgramEntries(const int *const prevWordsPtNodePos,
NgramListener *const listener) const;
int getShortcutPositionOfPtNode(const int ptNodePos) const;