Merge remote branch 'github/master' into miramerge

Conflicts:
	moses/src/AlignmentInfo.cpp
	moses/src/AlignmentInfo.h
	moses/src/ChartHypothesis.cpp
	moses/src/ChartTrellisNode.cpp
	moses/src/LM/Implementation.cpp
	moses/src/LM/Ken.cpp
	moses/src/TargetPhrase.cpp
	moses/src/TargetPhrase.h

2 files changed, 6 insertions, 1 deletions

diff --git a/phrase-extract/extract-ghkm/ExtractGHKM.cpp b/phrase-extract/extract-ghkm/ExtractGHKM.cpp
index dae876116..94e565085 100644
--- a/phrase-extract/extract-ghkm/ExtractGHKM.cpp
+++ b/phrase-extract/extract-ghkm/ExtractGHKM.cpp
@@ -90,7 +90,7 @@ int ExtractGHKM::Main(int argc, char *argv[])
   std::string alignmentLine;
   XmlTreeParser xmlTreeParser(labelSet, topLabelSet);
   ScfgRuleWriter writer(fwdExtractStream, invExtractStream, options);
-  size_t lineNum = 0;
+  size_t lineNum = options.sentenceOffset;
   while (true) {
     std::getline(targetStream, targetLine);
     std::getline(sourceStream, sourceLine);
@@ -289,6 +289,9 @@ void ExtractGHKM::ProcessOptions(int argc, char *argv[],
         "extract minimal rules only")
     ("PCFG",
         "include score based on PCFG scores in target corpus")
+    ("SentenceOffset",
+        po::value(&options.sentenceOffset)->default_value(options.sentenceOffset),
+        "set sentence number offset if processing split corpus")
     ("UnknownWordLabel",
         po::value(&options.unknownWordFile),
         "write unknown word labels to named file")
diff --git a/phrase-extract/extract-ghkm/Options.h b/phrase-extract/extract-ghkm/Options.h
index 362fc95d2..acad2040c 100644
--- a/phrase-extract/extract-ghkm/Options.h
+++ b/phrase-extract/extract-ghkm/Options.h
@@ -38,6 +38,7 @@ struct Options {
       , maxScope(3)
       , minimal(false)
       , pcfg(false)
+      , sentenceOffset(0)
       , unpairedExtractFormat(false) {}
 
   // Positional options
@@ -57,6 +58,7 @@ struct Options {
   int maxScope;
   bool minimal;
   bool pcfg;
+  int sentenceOffset;
   bool unpairedExtractFormat;
   std::string unknownWordFile;
 };