File size: 5,244 Bytes
158b61b |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 |
// $Id$
/***********************************************************************
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
***********************************************************************/
#ifndef moses_LanguageModelReloading_h
#define moses_LanguageModelReloading_h
#include <string>
#include "moses/LM/Base.h"
#include "moses/LM/Ken.h"
#include "util/tokenize_piece.hh"
#include "util/string_stream.hh"
#include <iostream>
namespace Moses
{
class FFState;
template <class Model> class ReloadingLanguageModel : public LanguageModelKen<Model>
{
public:
ReloadingLanguageModel(const std::string &line, const std::string &file, FactorType factorType, bool lazy) : LanguageModelKen<Model>(line, file, factorType, lazy ? util::LAZY : util::POPULATE_OR_READ), m_file(file), m_lazy(lazy) {
VERBOSE(1, "ReloadingLM constructor: " << m_file << std::endl);
}
virtual void InitializeForInput(ttasksptr const& ttask) {
VERBOSE(1, "ReloadingLM InitializeForInput" << std::endl);
// The context scope object for this translation task
// contains a map of translation task-specific data
boost::shared_ptr<Moses::ContextScope> contextScope = ttask->GetScope();
// The key to the map is this object
void const* key = static_cast<void const*>(this);
// The value stored in the map is a string representing a phrase table
boost::shared_ptr<string> value = contextScope->get<string>(key);
// Create a stream to read the phrase table data
stringstream strme(*(value.get()));
ofstream tmp;
tmp.open(m_file.c_str());
// Read the phrase table data, one line at a time
string line;
while (getline(strme, line)) {
tmp << line << "\n";
}
tmp.close();
LanguageModelKen<Model>::LoadModel(m_file, m_lazy ? util::LAZY : util::POPULATE_OR_READ);
};
protected:
using LanguageModelKen<Model>::m_ngram;
using LanguageModelKen<Model>::m_lmIdLookup;
using LanguageModelKen<Model>::m_beginSentenceFactor;
const std::string m_file;
bool m_lazy;
};
LanguageModel *ConstructReloadingLM(const std::string &line, const std::string &file, FactorType factorType, bool lazy)
{
lm::ngram::ModelType model_type;
if (lm::ngram::RecognizeBinary(file.c_str(), model_type)) {
switch(model_type) {
case lm::ngram::PROBING:
return new ReloadingLanguageModel<lm::ngram::ProbingModel>(line, file, factorType, lazy);
case lm::ngram::REST_PROBING:
return new ReloadingLanguageModel<lm::ngram::RestProbingModel>(line, file, factorType, lazy);
case lm::ngram::TRIE:
return new ReloadingLanguageModel<lm::ngram::TrieModel>(line, file, factorType, lazy);
case lm::ngram::QUANT_TRIE:
return new ReloadingLanguageModel<lm::ngram::QuantTrieModel>(line, file, factorType, lazy);
case lm::ngram::ARRAY_TRIE:
return new ReloadingLanguageModel<lm::ngram::ArrayTrieModel>(line, file, factorType, lazy);
case lm::ngram::QUANT_ARRAY_TRIE:
return new ReloadingLanguageModel<lm::ngram::QuantArrayTrieModel>(line, file, factorType, lazy);
default:
UTIL_THROW2("Unrecognized kenlm model type " << model_type);
}
} else {
return new ReloadingLanguageModel<lm::ngram::ProbingModel>(line, file, factorType, lazy);
}
}
LanguageModel *ConstructReloadingLM(const std::string &lineOrig)
{
FactorType factorType = 0;
std::string filePath;
bool lazy = false;
util::TokenIter<util::SingleCharacter, true> argument(lineOrig, ' ');
++argument; // KENLM
util::StringStream line;
line << "KENLM";
for (; argument; ++argument) {
const char *equals = std::find(argument->data(), argument->data() + argument->size(), '=');
UTIL_THROW_IF2(equals == argument->data() + argument->size(),
"Expected = in ReloadingLM argument " << *argument);
StringPiece name(argument->data(), equals - argument->data());
StringPiece value(equals + 1, argument->data() + argument->size() - equals - 1);
if (name == "factor") {
factorType = boost::lexical_cast<FactorType>(value);
} else if (name == "order") {
// Ignored
} else if (name == "path") {
filePath.assign(value.data(), value.size());
} else if (name == "lazyken") {
lazy = boost::lexical_cast<bool>(value);
} else {
// pass to base class to interpret
line << " " << name << "=" << value;
}
}
return ConstructReloadingLM(line.str(), filePath, factorType, lazy);
}
} // namespace Moses
#endif
|