mosesdecoder/moses/TranslationModel/PhraseDictionary.cpp
Ulrich Germann e4f5c69109 One step closer to eliminating the requirement to provide num-features=... in the config file.
Some FF (Mmsapt, LexicalReordering, Many single-value FF) provide this number during "registration";
when missing, a default weight vector of uniform 1.0 is automatically generated. This eliminates the
need for the user to figure out what the exact number of features is for each FF, which can get complicated,
e.g. in the case of Mmsapt/PhraseDictionaryBitextSampling.
2015-04-29 20:16:52 +01:00

282 lines
7.7 KiB
C++

// vim:tabstop=2
/***********************************************************************
Moses - factored phrase-based language decoder
Copyright (C) 2006 University of Edinburgh
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 <queue>
#include "moses/TranslationModel/PhraseDictionary.h"
#include "moses/StaticData.h"
#include "moses/InputType.h"
#include "moses/TranslationOption.h"
#include "moses/DecodeStep.h"
#include "moses/DecodeGraph.h"
#include "moses/InputPath.h"
#include "util/exception.hh"
using namespace std;
namespace Moses
{
std::vector<PhraseDictionary*> PhraseDictionary::s_staticColl;
CacheColl::~CacheColl()
{
for (iterator iter = begin(); iter != end(); ++iter) {
std::pair<const TargetPhraseCollection*, clock_t> &key = iter->second;
const TargetPhraseCollection *tps = key.first;
delete tps;
}
}
PhraseDictionary::PhraseDictionary(const std::string &line, bool registerNow)
: DecodeFeature(line, registerNow)
, m_tableLimit(20) // default
, m_maxCacheSize(DEFAULT_MAX_TRANS_OPT_CACHE_SIZE)
{
m_id = s_staticColl.size();
s_staticColl.push_back(this);
}
bool
PhraseDictionary::
ProvidesPrefixCheck() const
{
return false;
}
const TargetPhraseCollection *PhraseDictionary::GetTargetPhraseCollectionLEGACY(const Phrase& src) const
{
const TargetPhraseCollection *ret;
if (m_maxCacheSize) {
CacheColl &cache = GetCache();
size_t hash = hash_value(src);
CacheColl::iterator iter;
iter = cache.find(hash);
if (iter == cache.end()) {
// not in cache, need to look up from phrase table
ret = GetTargetPhraseCollectionNonCacheLEGACY(src);
if (ret) {
ret = new TargetPhraseCollection(*ret);
}
std::pair<const TargetPhraseCollection*, clock_t> value(ret, clock());
cache[hash] = value;
} else {
// in cache. just use it
std::pair<const TargetPhraseCollection*, clock_t> &value = iter->second;
value.second = clock();
ret = value.first;
}
} else {
// don't use cache. look up from phrase table
ret = GetTargetPhraseCollectionNonCacheLEGACY(src);
}
return ret;
}
TargetPhraseCollection const *
PhraseDictionary::
GetTargetPhraseCollectionNonCacheLEGACY(const Phrase& src) const
{
UTIL_THROW(util::Exception, "Legacy method not implemented");
}
TargetPhraseCollectionWithSourcePhrase const*
PhraseDictionary::
GetTargetPhraseCollectionLEGACY(InputType const& src,WordsRange const& range) const
{
UTIL_THROW(util::Exception, "Legacy method not implemented");
}
void
PhraseDictionary::
SetParameter(const std::string& key, const std::string& value)
{
if (key == "cache-size") {
m_maxCacheSize = Scan<size_t>(value);
} else if (key == "path") {
m_filePath = value;
} else if (key == "table-limit") {
m_tableLimit = Scan<size_t>(value);
} else {
DecodeFeature::SetParameter(key, value);
}
}
void
PhraseDictionary::
SetFeaturesToApply()
{
// find out which feature function can be applied in this decode step
const std::vector<FeatureFunction*> &allFeatures = FeatureFunction::GetFeatureFunctions();
for (size_t i = 0; i < allFeatures.size(); ++i) {
FeatureFunction *feature = allFeatures[i];
if (feature->IsUseable(m_outputFactors)) {
m_featuresToApply.push_back(feature);
}
}
}
// tell the Phrase Dictionary that the TargetPhraseCollection is not needed any more
void
PhraseDictionary::
Release(TargetPhraseCollection const* tpc) const
{
// do nothing by default
return;
}
bool
PhraseDictionary::
PrefixExists(ttasksptr const& ttask, Phrase const& phrase) const
{
return true;
}
void
PhraseDictionary::
GetTargetPhraseCollectionBatch(const InputPathList &inputPathQueue) const
{
InputPathList::const_iterator iter;
for (iter = inputPathQueue.begin(); iter != inputPathQueue.end(); ++iter) {
InputPath &inputPath = **iter;
// backoff
if (!SatisfyBackoff(inputPath)) {
continue;
}
const Phrase &phrase = inputPath.GetPhrase();
const TargetPhraseCollection *targetPhrases = this->GetTargetPhraseCollectionLEGACY(phrase);
inputPath.SetTargetPhrases(*this, targetPhrases, NULL);
}
}
// persistent cache handling
// saving presistent cache to disk
//void PhraseDictionary::SaveCache() const
//{
// CacheColl &cache = GetCache();
// for( std::map<size_t, std::pair<const TargetPhraseCollection*,clock_t> >::iterator iter,
// iter != cache.end(),
// iter++ ) {
//
// }
//}
// loading persistent cache from disk
//void PhraseDictionary::LoadCache() const
//{
// CacheColl &cache = GetCache();
// std::map<size_t, std::pair<const TargetPhraseCollection*,clock_t> >::iterator iter;
// iter = cache.begin();
// while( iter != cache.end() ) {
// std::map<size_t, std::pair<const TargetPhraseCollection*,clock_t> >::iterator iterRemove = iter++;
// delete iterRemove->second.first;
// cache.erase(iterRemove);
// }
//}
// reduce presistent cache by half of maximum size
void PhraseDictionary::ReduceCache() const
{
Timer reduceCacheTime;
reduceCacheTime.start();
CacheColl &cache = GetCache();
if (cache.size() <= m_maxCacheSize) return; // not full
// find cutoff for last used time
priority_queue< clock_t > lastUsedTimes;
CacheColl::iterator iter;
iter = cache.begin();
while( iter != cache.end() ) {
lastUsedTimes.push( iter->second.second );
iter++;
}
for( size_t i=0; i < lastUsedTimes.size()-m_maxCacheSize/2; i++ )
lastUsedTimes.pop();
clock_t cutoffLastUsedTime = lastUsedTimes.top();
// remove all old entries
iter = cache.begin();
while( iter != cache.end() ) {
if (iter->second.second < cutoffLastUsedTime) {
CacheColl::iterator iterRemove = iter++;
delete iterRemove->second.first;
cache.erase(iterRemove);
} else iter++;
}
VERBOSE(2,"Reduced persistent translation option cache in " << reduceCacheTime << " seconds." << std::endl);
}
CacheColl &PhraseDictionary::GetCache() const
{
CacheColl *cache;
cache = m_cache.get();
if (cache == NULL) {
cache = new CacheColl;
m_cache.reset(cache);
}
assert(cache);
return *cache;
}
bool PhraseDictionary::SatisfyBackoff(const InputPath &inputPath) const
{
const Phrase &sourcePhrase = inputPath.GetPhrase();
assert(m_container);
const DecodeGraph &decodeGraph = GetDecodeGraph();
size_t backoff = decodeGraph.GetBackoff();
if (backoff == 0) {
// ie. don't backoff. Collect ALL translations
return true;
}
if (sourcePhrase.GetSize() > backoff) {
// source phrase too big
return false;
}
// lookup translation only if no other translations
InputPath::TargetPhrases::const_iterator iter;
for (iter = inputPath.GetTargetPhrases().begin(); iter != inputPath.GetTargetPhrases().end(); ++iter) {
const std::pair<const TargetPhraseCollection*, const void*> &temp = iter->second;
const TargetPhraseCollection *tpCollPrev = temp.first;
if (tpCollPrev && tpCollPrev->GetSize()) {
// already have translation from another pt. Don't create translations
return false;
}
}
return true;
}
} // namespace