moses-smt/mosesdecoder
1
1
Fork 0
mirror of https://github.com/moses-smt/mosesdecoder.git synced 2024-11-09 16:04:41 +03:00
Code Issues Packages Projects Releases Wiki Activity

make it compatible with normal extract

Browse Source
This commit is contained in:
Hieu Hoang 2014-01-16 20:14:25 +00:00
parent fef94f27a7
commit 0410ab67ea
1 changed files with 16 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

21
contrib/other-builds/extract-mixed-syntax/extract.cpp
View File

@ -47,7 +47,7 @@ int main(int argc, char* argv[])
Global *global = new Global();
g_global = global;
int sentenceOffset = 0;
if (argc < 5) {
cerr << "syntax: extract-mixed-syntax corpus.target corpus.source corpus.align extract "
@ -167,16 +167,28 @@ int main(int argc, char* argv[])
}
else if (strcmp(argv[i],"--AllowDefaultNonTermEdge") == 0) {
global->allowDefaultNonTermEdge = atoi(argv[++i]);
} else if (strcmp(argv[i], "--GZOutput") == 0) {
}
else if (strcmp(argv[i], "--GZOutput") == 0) {
global->gzOutput = true;
}
else if (strcmp(argv[i],"--MaxSpan") == 0) {
// ignore
++i;
}
else if (strcmp(argv[i],"--SentenceOffset") == 0) {
if (i+1 >= argc || argv[i+1][0] < '0' || argv[i+1][0] > '9') {
cerr << "extract: syntax error, used switch --SentenceOffset without a number" << endl;
exit(1);
}
sentenceOffset = atoi(argv[++i]);
}
else {
cerr << "extract: syntax error, unknown option '" << string(argv[i]) << "'\n";
exit(1);
}
}
// open input files
Moses::InputFileStream tFile(fileNameT);
Moses::InputFileStream sFile(fileNameS);
@ -196,7 +208,7 @@ int main(int argc, char* argv[])
// loop through all sentence pairs
int i=0;
int i = sentenceOffset;
while(true) {
i++;
@ -240,7 +252,6 @@ int main(int argc, char* argv[])
sentencePair.GetLattice().GetRules().Output(extractFile);
sentencePair.GetLattice().GetRules().OutputInv(extractFileInv);
}
}
tFile.Close();