moses-smt/mosesdecoder
1
1
Fork 0
mirror of https://github.com/moses-smt/mosesdecoder.git synced 2024-12-28 14:32:38 +03:00
Code Issues Packages Projects Releases Wiki Activity

output glue rules

Browse Source
This commit is contained in:
Hieu Hoang 2014-01-20 18:37:48 +00:00
parent 4c009e31e8
commit b03e035b5d
1 changed files with 43 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

42
contrib/other-builds/extract-mixed-syntax/extract.cpp
View File

@ -39,6 +39,8 @@
using namespace std;
void writeGlueGrammar(const string &, Global &options, set< string > &targetLabelCollection, map< string, int > &targetTopLabelCollection);
int main(int argc, char* argv[])
{
cerr << "Extract v2.0, written by Philipp Koehn\n"
@ -261,8 +263,48 @@ int main(int argc, char* argv[])
extractFile.Close();
extractFileInv.Close();
if (global->glueGrammarFlag) {
writeGlueGrammar(fileNameGlueGrammar, *global, targetLabelCollection, targetTopLabelCollection);
}
delete global;
}
void writeGlueGrammar( const string & fileName, Global &options, set< string > &targetLabelCollection, map< string, int > &targetTopLabelCollection )
{
ofstream grammarFile;
grammarFile.open(fileName.c_str());
if (!options.targetSyntax) {
grammarFile << "<s> [X] ||| <s> [S] ||| 1 ||| ||| 0" << endl
<< "[X][S] </s> [X] ||| [X][S] </s> [S] ||| 1 ||| 0-0 ||| 0" << endl
<< "[X][S] [X][X] [X] ||| [X][S] [X][X] [S] ||| 2.718 ||| 0-0 1-1 ||| 0" << endl;
} else {
// chose a top label that is not already a label
string topLabel = "QQQQQQ";
for( unsigned int i=1; i<=topLabel.length(); i++) {
if(targetLabelCollection.find( topLabel.substr(0,i) ) == targetLabelCollection.end() ) {
topLabel = topLabel.substr(0,i);
break;
}
}
// basic rules
grammarFile << "<s> [X] ||| <s> [" << topLabel << "] ||| 1 ||| " << endl
<< "[X][" << topLabel << "] </s> [X] ||| [X][" << topLabel << "] </s> [" << topLabel << "] ||| 1 ||| 0-0 " << endl;
// top rules
for( map<string,int>::const_iterator i = targetTopLabelCollection.begin();
i != targetTopLabelCollection.end(); i++ ) {
grammarFile << "<s> [X][" << i->first << "] </s> [X] ||| <s> [X][" << i->first << "] </s> [" << topLabel << "] ||| 1 ||| 1-1" << endl;
}
// glue rules
for( set<string>::const_iterator i = targetLabelCollection.begin();
i != targetLabelCollection.end(); i++ ) {
grammarFile << "[X][" << topLabel << "] [X][" << *i << "] [X] ||| [X][" << topLabel << "] [X][" << *i << "] [" << topLabel << "] ||| 2.718 ||| 0-0 1-1" << endl;
}
grammarFile << "[X][" << topLabel << "] [X][X] [X] ||| [X][" << topLabel << "] [X][X] [" << topLabel << "] ||| 2.718 ||| 0-0 1-1 " << endl; // glue rule for unknown word...
}
grammarFile.close();
}