Donate to e Foundation | Murena handsets with /e/OS | Own a part of Murena! Learn more

Commit 9f8da0f8 authored by Keisuke Kuroyanagi's avatar Keisuke Kuroyanagi
Browse files

Use MultiBigramMap in structure policy. 

Bug: 14425059
Change-Id: I4d78da4839ef177e0223e6e5bcf0ebd7315c3099
parent 13864787

native/jni/src/suggest/core/dicnode/dic_node_utils.cpp

+2 −16
Original line number Diff line number Diff line
@@ -18,7 +18,6 @@ 


#include "suggest/core/dicnode/dic_node.h"

#include "suggest/core/dicnode/dic_node_vector.h"

#include "suggest/core/dictionary/multi_bigram_map.h"

#include "suggest/core/policy/dictionary_structure_with_buffer_policy.h"



namespace latinime {
@@ -73,25 +72,12 @@ namespace latinime { 
    if (dicNode->hasMultipleWords() && !dicNode->isValidMultipleWordSuggestion()) {

        return static_cast<float>(MAX_VALUE_FOR_WEIGHTING);

    }

    const int probability = getBigramNodeProbability(dictionaryStructurePolicy, dicNode,

            multiBigramMap);

    const int probability = dictionaryStructurePolicy->getProbabilityOfWordInContext(

            dicNode->getPrevWordIds(), dicNode->getWordId(), multiBigramMap);

    // TODO: This equation to calculate the improbability looks unreasonable.  Investigate this.

    const float cost = static_cast<float>(MAX_PROBABILITY - probability)

            / static_cast<float>(MAX_PROBABILITY);

    return cost;

}



/* static */ int DicNodeUtils::getBigramNodeProbability(

        const DictionaryStructureWithBufferPolicy *const dictionaryStructurePolicy,

        const DicNode *const dicNode, MultiBigramMap *const multiBigramMap) {

    const int unigramProbability = dicNode->getUnigramProbability();

    if (multiBigramMap) {

        const int *const prevWordIds = dicNode->getPrevWordIds();

        return multiBigramMap->getBigramProbability(dictionaryStructurePolicy,

                prevWordIds, dicNode->getWordId(), unigramProbability);

    }

    return dictionaryStructurePolicy->getProbability(unigramProbability,

            NOT_A_PROBABILITY);

}



} // namespace latinime

native/jni/src/suggest/core/dicnode/dic_node_utils.h

+0 −4
Original line number Diff line number Diff line
@@ -46,10 +46,6 @@ class DicNodeUtils { 
    DISALLOW_IMPLICIT_CONSTRUCTORS(DicNodeUtils);

    // Max number of bigrams to look up

    static const int MAX_BIGRAMS_CONSIDERED_PER_CONTEXT = 500;



    static int getBigramNodeProbability(

            const DictionaryStructureWithBufferPolicy *const dictionaryStructurePolicy,

            const DicNode *const dicNode, MultiBigramMap *const multiBigramMap);

};

} // namespace latinime

#endif // LATINIME_DIC_NODE_UTILS_H

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

+5 −0
Original line number Diff line number Diff line
@@ -29,6 +29,7 @@ namespace latinime { 
class DicNode;

class DicNodeVector;

class DictionaryHeaderStructurePolicy;

class MultiBigramMap;

class NgramListener;

class PrevWordsInfo;

class UnigramProperty;
@@ -56,6 +57,10 @@ class DictionaryStructureWithBufferPolicy { 
    virtual int getWordId(const CodePointArrayView wordCodePoints,

            const bool forceLowerCaseSearch) const = 0;



    virtual int getProbabilityOfWordInContext(const int *const prevWordIds, const int wordId,

            MultiBigramMap *const multiBigramMap) const = 0;



    // TODO: Remove

    virtual int getProbability(const int unigramProbability, const int bigramProbability) const = 0;



    virtual int getProbabilityOfWord(const int *const prevWordIds, const int wordId) const = 0;

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

+21 −0
Original line number Diff line number Diff line
@@ -28,6 +28,7 @@ 


#include "suggest/core/dicnode/dic_node.h"

#include "suggest/core/dicnode/dic_node_vector.h"

#include "suggest/core/dictionary/multi_bigram_map.h"

#include "suggest/core/dictionary/ngram_listener.h"

#include "suggest/core/dictionary/property/bigram_property.h"

#include "suggest/core/dictionary/property/unigram_property.h"
@@ -117,6 +118,26 @@ int Ver4PatriciaTriePolicy::getWordId(const CodePointArrayView wordCodePoints, 
    return getWordIdFromTerminalPtNodePos(ptNodePos);

}



int Ver4PatriciaTriePolicy::getProbabilityOfWordInContext(const int *const prevWordIds,

        const int wordId, MultiBigramMap *const multiBigramMap) const {

    if (wordId == NOT_A_WORD_ID) {

        return NOT_A_PROBABILITY;

    }

    const int ptNodePos = getTerminalPtNodePosFromWordId(wordId);

    const PtNodeParams ptNodeParams(mNodeReader.fetchPtNodeParamsInBufferFromPtNodePos(ptNodePos));

    if (multiBigramMap) {

        return multiBigramMap->getBigramProbability(this /* structurePolicy */, prevWordIds,

                wordId, ptNodeParams.getProbability());

    }

    if (prevWordIds) {

        const int probability = getProbabilityOfWord(prevWordIds, wordId);

        if (probability != NOT_A_PROBABILITY) {

            return probability;

        }

    }

    return getProbability(ptNodeParams.getProbability(), NOT_A_PROBABILITY);

}



int Ver4PatriciaTriePolicy::getProbability(const int unigramProbability,

        const int bigramProbability) const {

    if (mHeaderPolicy->isDecayingDict()) {

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

+3 −0
Original line number Diff line number Diff line
@@ -91,6 +91,9 @@ class Ver4PatriciaTriePolicy : public DictionaryStructureWithBufferPolicy { 


    int getWordId(const CodePointArrayView wordCodePoints, const bool forceLowerCaseSearch) const;



    int getProbabilityOfWordInContext(const int *const prevWordIds, const int wordId,

            MultiBigramMap *const multiBigramMap) const;



    int getProbability(const int unigramProbability, const int bigramProbability) const;



    int getProbabilityOfWord(const int *const prevWordIds, const int wordId) const;