mirror of
https://github.com/moses-smt/mosesdecoder.git
synced 2024-09-11 19:27:11 +03:00
This adds a new string-to-tree decoder, which can be enabled with the -s2t option. It's intended to be faster and simpler than the generic chart decoder, and is designed to support lattice input (still WIP). For a en-de system trained on WMT14 data, it's approximately 40% faster in practice. For background information, see the decoding section of the EMNLP tutorial on syntax-based MT: http://www.emnlp2014.org/tutorials/5_notes.pdf Some features are not implemented yet, including support for internal tree structure and soft source-syntactic constraints.
275 lines
10 KiB
C++
275 lines
10 KiB
C++
// $Id$
|
|
// vim:tabstop=2
|
|
/***********************************************************************
|
|
Moses - factored phrase-based language decoder
|
|
Copyright (C) 2010 Hieu Hoang
|
|
|
|
This library is free software; you can redistribute it and/or
|
|
modify it under the terms of the GNU Lesser General Public
|
|
License as published by the Free Software Foundation; either
|
|
version 2.1 of the License, or (at your option) any later version.
|
|
|
|
This library is distributed in the hope that it will be useful,
|
|
but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
|
|
Lesser General Public License for more details.
|
|
|
|
You should have received a copy of the GNU Lesser General Public
|
|
License along with this library; if not, write to the Free Software
|
|
Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
|
|
***********************************************************************/
|
|
|
|
#include "ChartParser.h"
|
|
#include "ChartParserCallback.h"
|
|
#include "ChartRuleLookupManager.h"
|
|
#include "StaticData.h"
|
|
#include "TreeInput.h"
|
|
#include "Sentence.h"
|
|
#include "DecodeGraph.h"
|
|
#include "moses/FF/UnknownWordPenaltyProducer.h"
|
|
#include "moses/TranslationModel/PhraseDictionary.h"
|
|
|
|
using namespace std;
|
|
using namespace Moses;
|
|
|
|
namespace Moses
|
|
{
|
|
|
|
ChartParserUnknown::ChartParserUnknown() {}
|
|
|
|
ChartParserUnknown::~ChartParserUnknown()
|
|
{
|
|
RemoveAllInColl(m_unksrcs);
|
|
RemoveAllInColl(m_cacheTargetPhraseCollection);
|
|
}
|
|
|
|
void ChartParserUnknown::Process(const Word &sourceWord, const WordsRange &range, ChartParserCallback &to)
|
|
{
|
|
// unknown word, add as trans opt
|
|
const StaticData &staticData = StaticData::Instance();
|
|
const UnknownWordPenaltyProducer &unknownWordPenaltyProducer = UnknownWordPenaltyProducer::Instance();
|
|
|
|
size_t isDigit = 0;
|
|
if (staticData.GetDropUnknown()) {
|
|
const Factor *f = sourceWord[0]; // TODO hack. shouldn't know which factor is surface
|
|
const StringPiece s = f->GetString();
|
|
isDigit = s.find_first_of("0123456789");
|
|
if (isDigit == string::npos)
|
|
isDigit = 0;
|
|
else
|
|
isDigit = 1;
|
|
// modify the starting bitmap
|
|
}
|
|
|
|
Phrase* unksrc = new Phrase(1);
|
|
unksrc->AddWord() = sourceWord;
|
|
Word &newWord = unksrc->GetWord(0);
|
|
newWord.SetIsOOV(true);
|
|
|
|
m_unksrcs.push_back(unksrc);
|
|
|
|
// hack. Once the OOV FF is a phrase table, get rid of this
|
|
PhraseDictionary *firstPt = NULL;
|
|
if (PhraseDictionary::GetColl().size() == 0) {
|
|
firstPt = PhraseDictionary::GetColl()[0];
|
|
}
|
|
|
|
//TranslationOption *transOpt;
|
|
if (! staticData.GetDropUnknown() || isDigit) {
|
|
// loop
|
|
const UnknownLHSList &lhsList = staticData.GetUnknownLHS();
|
|
UnknownLHSList::const_iterator iterLHS;
|
|
for (iterLHS = lhsList.begin(); iterLHS != lhsList.end(); ++iterLHS) {
|
|
const string &targetLHSStr = iterLHS->first;
|
|
float prob = iterLHS->second;
|
|
|
|
// lhs
|
|
//const Word &sourceLHS = staticData.GetInputDefaultNonTerminal();
|
|
Word *targetLHS = new Word(true);
|
|
|
|
targetLHS->CreateFromString(Output, staticData.GetOutputFactorOrder(), targetLHSStr, true);
|
|
UTIL_THROW_IF2(targetLHS->GetFactor(0) == NULL, "Null factor for target LHS");
|
|
|
|
// add to dictionary
|
|
TargetPhrase *targetPhrase = new TargetPhrase(firstPt);
|
|
Word &targetWord = targetPhrase->AddWord();
|
|
targetWord.CreateUnknownWord(sourceWord);
|
|
|
|
// scores
|
|
float unknownScore = FloorScore(TransformScore(prob));
|
|
|
|
targetPhrase->GetScoreBreakdown().Assign(&unknownWordPenaltyProducer, unknownScore);
|
|
targetPhrase->EvaluateInIsolation(*unksrc);
|
|
targetPhrase->SetTargetLHS(targetLHS);
|
|
targetPhrase->SetAlignmentInfo("0-0");
|
|
if (staticData.IsDetailedTreeFragmentsTranslationReportingEnabled() || staticData.PrintNBestTrees() || staticData.GetTreeStructure() != NULL) {
|
|
targetPhrase->SetProperty("Tree","[ " + (*targetLHS)[0]->GetString().as_string() + " "+sourceWord[0]->GetString().as_string()+" ]");
|
|
}
|
|
|
|
// chart rule
|
|
to.AddPhraseOOV(*targetPhrase, m_cacheTargetPhraseCollection, range);
|
|
} // for (iterLHS
|
|
} else {
|
|
// drop source word. create blank trans opt
|
|
float unknownScore = FloorScore(-numeric_limits<float>::infinity());
|
|
|
|
TargetPhrase *targetPhrase = new TargetPhrase(firstPt);
|
|
// loop
|
|
const UnknownLHSList &lhsList = staticData.GetUnknownLHS();
|
|
UnknownLHSList::const_iterator iterLHS;
|
|
for (iterLHS = lhsList.begin(); iterLHS != lhsList.end(); ++iterLHS) {
|
|
const string &targetLHSStr = iterLHS->first;
|
|
//float prob = iterLHS->second;
|
|
|
|
Word *targetLHS = new Word(true);
|
|
targetLHS->CreateFromString(Output, staticData.GetOutputFactorOrder(), targetLHSStr, true);
|
|
UTIL_THROW_IF2(targetLHS->GetFactor(0) == NULL, "Null factor for target LHS");
|
|
|
|
targetPhrase->GetScoreBreakdown().Assign(&unknownWordPenaltyProducer, unknownScore);
|
|
targetPhrase->EvaluateInIsolation(*unksrc);
|
|
|
|
targetPhrase->SetTargetLHS(targetLHS);
|
|
|
|
// chart rule
|
|
to.AddPhraseOOV(*targetPhrase, m_cacheTargetPhraseCollection, range);
|
|
}
|
|
}
|
|
}
|
|
|
|
ChartParser::ChartParser(InputType const &source, ChartCellCollectionBase &cells) :
|
|
m_decodeGraphList(StaticData::Instance().GetDecodeGraphs()),
|
|
m_source(source)
|
|
{
|
|
const StaticData &staticData = StaticData::Instance();
|
|
|
|
staticData.InitializeForInput(source);
|
|
CreateInputPaths(m_source);
|
|
|
|
const std::vector<PhraseDictionary*> &dictionaries = PhraseDictionary::GetColl();
|
|
assert(dictionaries.size() == m_decodeGraphList.size());
|
|
m_ruleLookupManagers.reserve(dictionaries.size());
|
|
for (std::size_t i = 0; i < dictionaries.size(); ++i) {
|
|
const PhraseDictionary *dict = dictionaries[i];
|
|
PhraseDictionary *nonConstDict = const_cast<PhraseDictionary*>(dict);
|
|
std::size_t maxChartSpan = m_decodeGraphList[i]->GetMaxChartSpan();
|
|
ChartRuleLookupManager *lookupMgr = nonConstDict->CreateRuleLookupManager(*this, cells, maxChartSpan);
|
|
m_ruleLookupManagers.push_back(lookupMgr);
|
|
}
|
|
|
|
}
|
|
|
|
ChartParser::~ChartParser()
|
|
{
|
|
RemoveAllInColl(m_ruleLookupManagers);
|
|
StaticData::Instance().CleanUpAfterSentenceProcessing(m_source);
|
|
|
|
InputPathMatrix::const_iterator iterOuter;
|
|
for (iterOuter = m_inputPathMatrix.begin(); iterOuter != m_inputPathMatrix.end(); ++iterOuter) {
|
|
const std::vector<InputPath*> &outer = *iterOuter;
|
|
|
|
std::vector<InputPath*>::const_iterator iterInner;
|
|
for (iterInner = outer.begin(); iterInner != outer.end(); ++iterInner) {
|
|
InputPath *path = *iterInner;
|
|
delete path;
|
|
}
|
|
}
|
|
}
|
|
|
|
void ChartParser::Create(const WordsRange &wordsRange, ChartParserCallback &to)
|
|
{
|
|
assert(m_decodeGraphList.size() == m_ruleLookupManagers.size());
|
|
|
|
std::vector <DecodeGraph*>::const_iterator iterDecodeGraph;
|
|
std::vector <ChartRuleLookupManager*>::const_iterator iterRuleLookupManagers = m_ruleLookupManagers.begin();
|
|
for (iterDecodeGraph = m_decodeGraphList.begin(); iterDecodeGraph != m_decodeGraphList.end(); ++iterDecodeGraph, ++iterRuleLookupManagers) {
|
|
const DecodeGraph &decodeGraph = **iterDecodeGraph;
|
|
assert(decodeGraph.GetSize() == 1);
|
|
ChartRuleLookupManager &ruleLookupManager = **iterRuleLookupManagers;
|
|
size_t maxSpan = decodeGraph.GetMaxChartSpan();
|
|
size_t last = m_source.GetSize()-1;
|
|
if (maxSpan != 0) {
|
|
last = min(last, wordsRange.GetStartPos()+maxSpan);
|
|
}
|
|
if (maxSpan == 0 || wordsRange.GetNumWordsCovered() <= maxSpan) {
|
|
ruleLookupManager.GetChartRuleCollection(wordsRange, last, to);
|
|
}
|
|
}
|
|
|
|
if (wordsRange.GetNumWordsCovered() == 1 && wordsRange.GetStartPos() != 0 && wordsRange.GetStartPos() != m_source.GetSize()-1) {
|
|
bool alwaysCreateDirectTranslationOption = StaticData::Instance().IsAlwaysCreateDirectTranslationOption();
|
|
if (to.Empty() || alwaysCreateDirectTranslationOption) {
|
|
// create unknown words for 1 word coverage where we don't have any trans options
|
|
const Word &sourceWord = m_source.GetWord(wordsRange.GetStartPos());
|
|
m_unknown.Process(sourceWord, wordsRange, to);
|
|
}
|
|
}
|
|
}
|
|
|
|
void ChartParser::CreateInputPaths(const InputType &input)
|
|
{
|
|
size_t size = input.GetSize();
|
|
m_inputPathMatrix.resize(size);
|
|
|
|
UTIL_THROW_IF2(input.GetType() != SentenceInput && input.GetType() != TreeInputType,
|
|
"Input must be a sentence or a tree, not lattice or confusion networks");
|
|
for (size_t phaseSize = 1; phaseSize <= size; ++phaseSize) {
|
|
for (size_t startPos = 0; startPos < size - phaseSize + 1; ++startPos) {
|
|
size_t endPos = startPos + phaseSize -1;
|
|
vector<InputPath*> &vec = m_inputPathMatrix[startPos];
|
|
|
|
WordsRange range(startPos, endPos);
|
|
Phrase subphrase(input.GetSubString(WordsRange(startPos, endPos)));
|
|
const NonTerminalSet &labels = input.GetLabelSet(startPos, endPos);
|
|
|
|
InputPath *node;
|
|
if (range.GetNumWordsCovered() == 1) {
|
|
node = new InputPath(subphrase, labels, range, NULL, NULL);
|
|
vec.push_back(node);
|
|
} else {
|
|
const InputPath &prevNode = GetInputPath(startPos, endPos - 1);
|
|
node = new InputPath(subphrase, labels, range, &prevNode, NULL);
|
|
vec.push_back(node);
|
|
}
|
|
|
|
//m_inputPathQueue.push_back(node);
|
|
}
|
|
}
|
|
}
|
|
|
|
const InputPath &ChartParser::GetInputPath(WordsRange &range) const
|
|
{
|
|
return GetInputPath(range.GetStartPos(), range.GetEndPos());
|
|
}
|
|
|
|
const InputPath &ChartParser::GetInputPath(size_t startPos, size_t endPos) const
|
|
{
|
|
size_t offset = endPos - startPos;
|
|
UTIL_THROW_IF2(offset >= m_inputPathMatrix[startPos].size(),
|
|
"Out of bound: " << offset);
|
|
return *m_inputPathMatrix[startPos][offset];
|
|
}
|
|
|
|
InputPath &ChartParser::GetInputPath(size_t startPos, size_t endPos)
|
|
{
|
|
size_t offset = endPos - startPos;
|
|
UTIL_THROW_IF2(offset >= m_inputPathMatrix[startPos].size(),
|
|
"Out of bound: " << offset);
|
|
return *m_inputPathMatrix[startPos][offset];
|
|
}
|
|
/*
|
|
const Sentence &ChartParser::GetSentence() const {
|
|
const Sentence &sentence = static_cast<const Sentence&>(m_source);
|
|
return sentence;
|
|
}
|
|
*/
|
|
size_t ChartParser::GetSize() const
|
|
{
|
|
return m_source.GetSize();
|
|
}
|
|
|
|
long ChartParser::GetTranslationId() const
|
|
{
|
|
return m_source.GetTranslationId();
|
|
}
|
|
} // namespace Moses
|