diff --git a/phrase-extract/SentenceAlignmentWithSyntax.h b/phrase-extract/SentenceAlignmentWithSyntax.h
index 8b9088770..604b6d0e2 100644
--- a/phrase-extract/SentenceAlignmentWithSyntax.h
+++ b/phrase-extract/SentenceAlignmentWithSyntax.h
@@ -28,7 +28,7 @@
 
 #include "RuleExtractionOptions.h"
 #include "SentenceAlignment.h"
-#include "SyntaxTree.h"
+#include "SyntaxNodeCollection.h"
 
 namespace MosesTraining
 {
@@ -36,8 +36,8 @@ namespace MosesTraining
 class SentenceAlignmentWithSyntax : public SentenceAlignment
 {
 public:
-  SyntaxTree targetTree;
-  SyntaxTree sourceTree;
+  SyntaxNodeCollection targetTree;
+  SyntaxNodeCollection sourceTree;
   std::set<std::string> & m_targetLabelCollection;
   std::set<std::string> & m_sourceLabelCollection;
   std::map<std::string, int> & m_targetTopLabelCollection;
diff --git a/phrase-extract/SyntaxNode.h b/phrase-extract/SyntaxNode.h
new file mode 100644
index 000000000..46e0f456f
--- /dev/null
+++ b/phrase-extract/SyntaxNode.h
@@ -0,0 +1,75 @@
+/***********************************************************************
+  Moses - factored phrase-based language decoder
+  Copyright (C) 2009 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
+ ***********************************************************************/
+
+#pragma once
+
+#include <map>
+#include <sstream>
+#include <string>
+#include <vector>
+
+namespace MosesTraining
+{
+
+class SyntaxNode
+{
+protected:
+  int m_start, m_end;
+  std::string m_label;
+  std::vector< SyntaxNode* > m_children;
+  SyntaxNode* m_parent;
+  float m_pcfgScore;
+public:
+  SyntaxNode( int startPos, int endPos, std::string label )
+    :m_start(startPos)
+    ,m_end(endPos)
+    ,m_label(label)
+    ,m_parent(0)
+    ,m_pcfgScore(0.0f) {
+  }
+  int GetStart() const {
+    return m_start;
+  }
+  int GetEnd() const {
+    return m_end;
+  }
+  std::string GetLabel() const {
+    return m_label;
+  }
+  float GetPcfgScore() const {
+    return m_pcfgScore;
+  }
+  void SetPcfgScore(float score) {
+    m_pcfgScore = score;
+  }
+  SyntaxNode *GetParent() {
+    return m_parent;
+  }
+  void SetParent(SyntaxNode *parent) {
+    m_parent = parent;
+  }
+  void AddChild(SyntaxNode* child) {
+    m_children.push_back(child);
+  }
+  const std::vector< SyntaxNode* > &GetChildren() const {
+    return m_children;
+  }
+};
+
+}  // namespace MosesTraining
diff --git a/phrase-extract/SyntaxTree.cpp b/phrase-extract/SyntaxNodeCollection.cpp
similarity index 80%
rename from phrase-extract/SyntaxTree.cpp
rename to phrase-extract/SyntaxNodeCollection.cpp
index c50693e0d..099a5697f 100644
--- a/phrase-extract/SyntaxTree.cpp
+++ b/phrase-extract/SyntaxNodeCollection.cpp
@@ -1,6 +1,3 @@
-// $Id: SyntaxTree.cpp 1960 2008-12-15 12:52:38Z phkoehn $
-// vim:tabstop=2
-
 /***********************************************************************
   Moses - factored phrase-based language decoder
   Copyright (C) 2009 University of Edinburgh
@@ -21,7 +18,7 @@
  ***********************************************************************/
 
 
-#include "SyntaxTree.h"
+#include "SyntaxNodeCollection.h"
 
 #include <cassert>
 #include <iostream>
@@ -29,12 +26,12 @@
 namespace MosesTraining
 {
 
-SyntaxTree::~SyntaxTree()
+SyntaxNodeCollection::~SyntaxNodeCollection()
 {
   Clear();
 }
 
-void SyntaxTree::Clear()
+void SyntaxNodeCollection::Clear()
 {
   m_top = 0;
   // loop through all m_nodes, delete them
@@ -45,7 +42,8 @@ void SyntaxTree::Clear()
   m_index.clear();
 }
 
-SyntaxNode *SyntaxTree::AddNode( int startPos, int endPos, std::string label )
+SyntaxNode *SyntaxNodeCollection::AddNode(int startPos, int endPos,
+                                          const std::string &label)
 {
   SyntaxNode* newNode = new SyntaxNode( startPos, endPos, label );
   m_nodes.push_back( newNode );
@@ -54,7 +52,7 @@ SyntaxNode *SyntaxTree::AddNode( int startPos, int endPos, std::string label )
   return newNode;
 }
 
-ParentNodes SyntaxTree::Parse()
+ParentNodes SyntaxNodeCollection::Parse()
 {
   ParentNodes parents;
 
@@ -94,12 +92,12 @@ ParentNodes SyntaxTree::Parse()
   return parents;
 }
 
-bool SyntaxTree::HasNode( int startPos, int endPos ) const
+bool SyntaxNodeCollection::HasNode( int startPos, int endPos ) const
 {
   return GetNodes( startPos, endPos).size() > 0;
 }
 
-const std::vector< SyntaxNode* >& SyntaxTree::GetNodes( int startPos, int endPos ) const
+const std::vector< SyntaxNode* >& SyntaxNodeCollection::GetNodes( int startPos, int endPos ) const
 {
   SyntaxTreeIndexIterator startIndex = m_index.find( startPos );
   if (startIndex == m_index.end() )
@@ -112,15 +110,7 @@ const std::vector< SyntaxNode* >& SyntaxTree::GetNodes( int startPos, int endPos
   return endIndex->second;
 }
 
-// for printing out tree
-std::string SyntaxTree::ToString() const
-{
-  std::stringstream out;
-  out << *this;
-  return out.str();
-}
-
-void SyntaxTree::ConnectNodes()
+void SyntaxNodeCollection::ConnectNodes()
 {
   typedef SyntaxTreeIndex2::const_reverse_iterator InnerIterator;
 
@@ -162,27 +152,4 @@ void SyntaxTree::ConnectNodes()
   }
 }
 
-std::ostream& operator<<(std::ostream& os, const SyntaxTree& t)
-{
-  size_t size = t.m_index.size();
-  for(size_t length=1; length<=size; length++) {
-    for(size_t space=0; space<length; space++) {
-      os << "    ";
-    }
-    for(size_t start=0; start<=size-length; start++) {
-
-      if (t.HasNode( start, start+(length-1) )) {
-        std::string label = t.GetNodes( start, start+(length-1) )[0]->GetLabel() + "#######";
-
-        os << label.substr(0,7) << " ";
-      } else {
-        os << "------- ";
-      }
-    }
-    os << std::endl;
-  }
-  return os;
-}
-
-}
-
+}  // namespace MosesTraining
diff --git a/phrase-extract/SyntaxTree.h b/phrase-extract/SyntaxNodeCollection.h
similarity index 62%
rename from phrase-extract/SyntaxTree.h
rename to phrase-extract/SyntaxNodeCollection.h
index 6ffb5da34..70b14206d 100644
--- a/phrase-extract/SyntaxTree.h
+++ b/phrase-extract/SyntaxNodeCollection.h
@@ -1,6 +1,3 @@
-// $Id: SyntaxTree.h 1960 2008-12-15 12:52:38Z phkoehn $
-// vim:tabstop=2
-
 /***********************************************************************
   Moses - factored phrase-based language decoder
   Copyright (C) 2009 University of Edinburgh
@@ -20,66 +17,22 @@
   Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301  USA
  ***********************************************************************/
 
-
 #pragma once
-#include <string>
-#include <vector>
+
 #include <map>
 #include <sstream>
+#include <string>
+#include <vector>
+
+#include "SyntaxNode.h"
 
 namespace MosesTraining
 {
 
-class SyntaxNode
-{
-protected:
-  int m_start, m_end;
-  std::string m_label;
-  std::vector< SyntaxNode* > m_children;
-  SyntaxNode* m_parent;
-  float m_pcfgScore;
-public:
-  SyntaxNode( int startPos, int endPos, std::string label )
-    :m_start(startPos)
-    ,m_end(endPos)
-    ,m_label(label)
-    ,m_parent(0)
-    ,m_pcfgScore(0.0f) {
-  }
-  int GetStart() const {
-    return m_start;
-  }
-  int GetEnd() const {
-    return m_end;
-  }
-  std::string GetLabel() const {
-    return m_label;
-  }
-  float GetPcfgScore() const {
-    return m_pcfgScore;
-  }
-  void SetPcfgScore(float score) {
-    m_pcfgScore = score;
-  }
-  SyntaxNode *GetParent() {
-    return m_parent;
-  }
-  void SetParent(SyntaxNode *parent) {
-    m_parent = parent;
-  }
-  void AddChild(SyntaxNode* child) {
-    m_children.push_back(child);
-  }
-  const std::vector< SyntaxNode* > &GetChildren() const {
-    return m_children;
-  }
-};
-
-
 typedef std::vector< int > SplitPoints;
 typedef std::vector< SplitPoints > ParentNodes;
 
-class SyntaxTree
+class SyntaxNodeCollection
 {
 protected:
   std::vector< SyntaxNode* > m_nodes;
@@ -93,16 +46,14 @@ protected:
   int m_size;
   std::vector< SyntaxNode* > m_emptyNode;
 
-  friend std::ostream& operator<<(std::ostream&, const SyntaxTree&);
-
 public:
-  SyntaxTree()
+  SyntaxNodeCollection()
     : m_top(0)  // m_top doesn't get set unless ConnectNodes is called.
     , m_size(0) {}
 
-  ~SyntaxTree();
+  ~SyntaxNodeCollection();
 
-  SyntaxNode *AddNode( int startPos, int endPos, std::string label );
+  SyntaxNode *AddNode( int startPos, int endPos, const std::string &label );
 
   SyntaxNode *GetTop() {
     return m_top;
@@ -119,10 +70,6 @@ public:
   }
   void ConnectNodes();
   void Clear();
-  std::string ToString() const;
 };
 
-std::ostream& operator<<(std::ostream&, const SyntaxTree&);
-
-}
-
+}  // namespace MosesTraining
diff --git a/phrase-extract/XmlTree.cpp b/phrase-extract/XmlTree.cpp
index 6efa1bf5c..0f068fca7 100644
--- a/phrase-extract/XmlTree.cpp
+++ b/phrase-extract/XmlTree.cpp
@@ -1,6 +1,3 @@
-// $Id: XmlOption.cpp 1960 2008-12-15 12:52:38Z phkoehn $
-// vim:tabstop=2
-
 /***********************************************************************
   Moses - factored phrase-based language decoder
   Copyright (C) 2006 University of Edinburgh
@@ -27,7 +24,8 @@
 #include <iostream>
 #include <cstdlib>
 #include <sstream>
-#include "SyntaxTree.h"
+
+#include "SyntaxNodeCollection.h"
 #include "XmlException.h"
 
 using namespace std;
@@ -228,7 +226,10 @@ vector<string> TokenizeXml(const string& str)
 	parse because we don't have the completed source parsed until after this function
 	removes all the markup from it (CreateFromString in Sentence::Read).
 */
-bool ProcessAndStripXMLTags(string &line, SyntaxTree &tree, set< string > &labelCollection, map< string, int > &topLabelCollection, bool unescapeSpecialChars )
+bool ProcessAndStripXMLTags(string &line, SyntaxNodeCollection &nodeCollection,
+                            set< string > &labelCollection,
+                            map< string, int > &topLabelCollection,
+                            bool unescapeSpecialChars )
 {
   //parse XML markup in translation line
 
@@ -374,7 +375,7 @@ bool ProcessAndStripXMLTags(string &line, SyntaxTree &tree, set< string > &label
           cerr << "XML TAG LABEL IS: '" << label << "'" << endl;
           cerr << "XML SPAN IS: " << startPos << "-" << (endPos-1) << endl;
         }
-        SyntaxNode *node = tree.AddNode( startPos, endPos-1, label );
+        SyntaxNode *node = nodeCollection.AddNode( startPos, endPos-1, label );
         node->SetPcfgScore(pcfgScore);
       }
     }
@@ -386,7 +387,7 @@ bool ProcessAndStripXMLTags(string &line, SyntaxTree &tree, set< string > &label
   }
 
   // collect top labels
-  const vector< SyntaxNode* >& topNodes = tree.GetNodes( 0, wordPos-1 );
+  const vector< SyntaxNode* >& topNodes = nodeCollection.GetNodes( 0, wordPos-1 );
   for( vector< SyntaxNode* >::const_iterator node = topNodes.begin(); node != topNodes.end(); node++ ) {
     SyntaxNode *n = *node;
     const string &label = n->GetLabel();
diff --git a/phrase-extract/XmlTree.h b/phrase-extract/XmlTree.h
index 50b1c0acc..3b5afd4dd 100644
--- a/phrase-extract/XmlTree.h
+++ b/phrase-extract/XmlTree.h
@@ -1,6 +1,3 @@
-// $Id: XmlOption.cpp 1960 2008-12-15 12:52:38Z phkoehn $
-// vim:tabstop=2
-
 /***********************************************************************
   Moses - factored phrase-based language decoder
   Copyright (C) 2006 University of Edinburgh
@@ -21,11 +18,13 @@
  ***********************************************************************/
 
 #pragma once
+
 #include <string>
 #include <vector>
 #include <set>
 #include <map>
-#include "SyntaxTree.h"
+
+#include "SyntaxNodeCollection.h"
 
 namespace MosesTraining
 {
@@ -35,9 +34,8 @@ std::string Trim(const std::string& str, const std::string dropChars = " \t\n\r"
 std::string TrimXml(const std::string& str);
 bool isXmlTag(const std::string& tag);
 std::vector<std::string> TokenizeXml(const std::string& str);
-bool ProcessAndStripXMLTags(std::string &line, SyntaxTree &tree, std::set< std::string > &labelCollection, std::map< std::string, int > &topLabelCollection, bool unescape = true);
+bool ProcessAndStripXMLTags(std::string &line, SyntaxNodeCollection &tree, std::set< std::string > &labelCollection, std::map< std::string, int > &topLabelCollection, bool unescape = true);
 std::string unescape(const std::string &str);
 
 
-} // namespace
-
+} // namespace MosesTraining
diff --git a/phrase-extract/extract-ghkm/ExtractGHKM.cpp b/phrase-extract/extract-ghkm/ExtractGHKM.cpp
index bc687ec6b..937d88030 100644
--- a/phrase-extract/extract-ghkm/ExtractGHKM.cpp
+++ b/phrase-extract/extract-ghkm/ExtractGHKM.cpp
@@ -33,7 +33,8 @@
 #include "Span.h"
 #include "StsgRule.h"
 #include "StsgRuleWriter.h"
-#include "SyntaxTree.h"
+#include "SyntaxNode.h"
+#include "SyntaxNodeCollection.h"
 #include "tables-core.h"
 #include "XmlException.h"
 #include "XmlTree.h"
@@ -172,7 +173,7 @@ int ExtractGHKM::Main(int argc, char *argv[])
 
 
     // Parse source tree and construct a SyntaxTree object.
-    MosesTraining::SyntaxTree sourceSyntaxTree;
+    MosesTraining::SyntaxNodeCollection sourceSyntaxTree;
     MosesTraining::SyntaxNode *sourceSyntaxTreeRoot=NULL;
 
     if (options.sourceLabels) {
@@ -196,7 +197,7 @@ int ExtractGHKM::Main(int argc, char *argv[])
     // Read source tokens.
     std::vector<std::string> sourceTokens(ReadTokens(sourceLine));
 
-    // Construct a source ParseTree object from the SyntaxTree object.
+    // Construct a source ParseTree object from the SyntaxNodeCollection object.
     std::auto_ptr<ParseTree> sourceParseTree;
 
     if (options.sourceLabels) {
diff --git a/phrase-extract/extract-ghkm/ScfgRule.cpp b/phrase-extract/extract-ghkm/ScfgRule.cpp
index 01178b72c..918c88eeb 100644
--- a/phrase-extract/extract-ghkm/ScfgRule.cpp
+++ b/phrase-extract/extract-ghkm/ScfgRule.cpp
@@ -19,11 +19,12 @@
 
 #include "ScfgRule.h"
 
+#include <algorithm>
+
 #include "Node.h"
 #include "Subgraph.h"
-#include "SyntaxTree.h"
-
-#include <algorithm>
+#include "SyntaxNode.h"
+#include "SyntaxNodeCollection.h"
 
 namespace Moses
 {
@@ -31,7 +32,7 @@ namespace GHKM
 {
 
 ScfgRule::ScfgRule(const Subgraph &fragment,
-                   const MosesTraining::SyntaxTree *sourceSyntaxTree)
+                   const MosesTraining::SyntaxNodeCollection *sourceSyntaxTree)
   : m_graphFragment(fragment)
   , m_sourceLHS("X", NonTerminal)
   , m_targetLHS(fragment.GetRoot()->GetLabel(), NonTerminal)
@@ -133,9 +134,9 @@ ScfgRule::ScfgRule(const Subgraph &fragment,
   }
 }
 
-void ScfgRule::PushSourceLabel(const MosesTraining::SyntaxTree *sourceSyntaxTree,
-                               const Node *node,
-                               const std::string &nonMatchingLabel)
+void ScfgRule::PushSourceLabel(
+    const MosesTraining::SyntaxNodeCollection *sourceSyntaxTree,
+    const Node *node, const std::string &nonMatchingLabel)
 {
   ContiguousSpan span = Closure(node->GetSpan());
   if (sourceSyntaxTree->HasNode(span.first,span.second)) { // does a source constituent match the span?
diff --git a/phrase-extract/extract-ghkm/ScfgRule.h b/phrase-extract/extract-ghkm/ScfgRule.h
index 94ee7b82e..c8b76114a 100644
--- a/phrase-extract/extract-ghkm/ScfgRule.h
+++ b/phrase-extract/extract-ghkm/ScfgRule.h
@@ -19,16 +19,16 @@
 
 #pragma once
 
-#include "Alignment.h"
-#include "Rule.h"
-#include "SyntaxTree.h"
-
 #include <string>
 #include <vector>
 #include <list>
 #include <memory>
 #include <iostream>
 
+#include "Alignment.h"
+#include "Rule.h"
+#include "SyntaxNodeCollection.h"
+
 namespace Moses
 {
 namespace GHKM
@@ -41,7 +41,7 @@ class ScfgRule : public Rule
 {
 public:
   ScfgRule(const Subgraph &fragment,
-           const MosesTraining::SyntaxTree *sourceSyntaxTree = 0);
+           const MosesTraining::SyntaxNodeCollection *sourceSyntaxTree = 0);
 
   const Subgraph &GetGraphFragment() const {
     return m_graphFragment;
@@ -78,9 +78,9 @@ public:
   }
 
 private:
-  void PushSourceLabel(const MosesTraining::SyntaxTree *sourceSyntaxTree,
-                       const Node *node,
-                       const std::string &nonMatchingLabel);
+  void PushSourceLabel(
+      const MosesTraining::SyntaxNodeCollection *sourceSyntaxTree,
+      const Node *node, const std::string &nonMatchingLabel);
 
   const Subgraph& m_graphFragment;
   Symbol m_sourceLHS;
@@ -95,4 +95,3 @@ private:
 
 }  // namespace GHKM
 }  // namespace Moses
-
diff --git a/phrase-extract/extract-ghkm/XmlTreeParser.h b/phrase-extract/extract-ghkm/XmlTreeParser.h
index ff0baeace..db9fa8bf2 100644
--- a/phrase-extract/extract-ghkm/XmlTreeParser.h
+++ b/phrase-extract/extract-ghkm/XmlTreeParser.h
@@ -23,14 +23,15 @@
 
 #include "Exception.h"
 
-#include "SyntaxTree.h"
-
 #include <map>
 #include <memory>
 #include <set>
 #include <string>
 #include <vector>
 
+#include "SyntaxNode.h"
+#include "SyntaxNodeCollection.h"
+
 namespace Moses
 {
 namespace GHKM
@@ -58,7 +59,7 @@ private:
   std::set<std::string> &m_labelSet;
   std::map<std::string, int> &m_topLabelSet;
   std::string m_line;
-  MosesTraining::SyntaxTree m_tree;
+  MosesTraining::SyntaxNodeCollection m_tree;
   std::vector<std::string> m_words;
 };
 
diff --git a/phrase-extract/extract-rules-main.cpp b/phrase-extract/extract-rules-main.cpp
index 50baa4e0d..825f12d89 100644
--- a/phrase-extract/extract-rules-main.cpp
+++ b/phrase-extract/extract-rules-main.cpp
@@ -41,7 +41,7 @@
 #include "HoleCollection.h"
 #include "RuleExist.h"
 #include "SentenceAlignmentWithSyntax.h"
-#include "SyntaxTree.h"
+#include "SyntaxNode.h"
 #include "tables-core.h"
 #include "XmlTree.h"
 #include "InputFileStream.h"
diff --git a/phrase-extract/pcfg-common/xml_tree_parser.h b/phrase-extract/pcfg-common/xml_tree_parser.h
index 675a112d8..8605c0691 100644
--- a/phrase-extract/pcfg-common/xml_tree_parser.h
+++ b/phrase-extract/pcfg-common/xml_tree_parser.h
@@ -28,7 +28,8 @@
 #include <vector>
 
 #include "pcfg_tree.h"
-#include "SyntaxTree.h"
+#include "SyntaxNode.h"
+#include "SyntaxNodeCollection.h"
 
 namespace MosesTraining {
 namespace Syntax {
@@ -47,7 +48,7 @@ class XmlTreeParser {
   std::set<std::string> m_labelSet;
   std::map<std::string, int> m_topLabelSet;
   std::string m_line;
-  MosesTraining::SyntaxTree m_tree;
+  MosesTraining::SyntaxNodeCollection m_tree;
   std::vector<std::string> m_words;
 };
 
diff --git a/phrase-extract/relax-parse-main.cpp b/phrase-extract/relax-parse-main.cpp
index 5c9daa7ae..5bca886bf 100644
--- a/phrase-extract/relax-parse-main.cpp
+++ b/phrase-extract/relax-parse-main.cpp
@@ -43,7 +43,7 @@ int main(int argc, char* argv[])
     // process into syntax tree representation
     set< string > labelCollection;         // set of labels, not used
     map< string, int > topLabelCollection; // count of top labels, not used
-    SyntaxTree tree;
+    SyntaxNodeCollection tree;
     ProcessAndStripXMLTags( inBufferString, tree, labelCollection, topLabelCollection, false );
     const vector< string > inWords = util::tokenize( inBufferString );
 
@@ -105,7 +105,7 @@ void init(int argc, char* argv[])
   }
 }
 
-void store( SyntaxTree &tree, const vector< string > &words )
+void store( SyntaxNodeCollection &tree, const vector< string > &words )
 {
   // output words
   for( size_t i=0; i<words.size(); i++ ) {
@@ -126,7 +126,7 @@ void store( SyntaxTree &tree, const vector< string > &words )
   cout << endl;
 }
 
-void LeftBinarize( SyntaxTree &tree, ParentNodes &parents )
+void LeftBinarize( SyntaxNodeCollection &tree, ParentNodes &parents )
 {
   for(ParentNodes::const_iterator p = parents.begin(); p != parents.end(); p++) {
     const SplitPoints &point = *p;
@@ -143,7 +143,7 @@ void LeftBinarize( SyntaxTree &tree, ParentNodes &parents )
   }
 }
 
-void RightBinarize( SyntaxTree &tree, ParentNodes &parents )
+void RightBinarize( SyntaxNodeCollection &tree, ParentNodes &parents )
 {
   for(ParentNodes::const_iterator p = parents.begin(); p != parents.end(); p++) {
     const SplitPoints &point = *p;
@@ -161,11 +161,11 @@ void RightBinarize( SyntaxTree &tree, ParentNodes &parents )
   }
 }
 
-void SAMT( SyntaxTree &tree, ParentNodes &parents )
+void SAMT( SyntaxNodeCollection &tree, ParentNodes &parents )
 {
   int numWords = tree.GetNumWords();
 
-  SyntaxTree newTree; // to store new nodes
+  SyntaxNodeCollection newTree; // to store new nodes
 
   // look through parents to combine children
   for(ParentNodes::const_iterator p = parents.begin(); p != parents.end(); p++) {
diff --git a/phrase-extract/relax-parse.h b/phrase-extract/relax-parse.h
index 9bd0bfb23..a00aa6deb 100644
--- a/phrase-extract/relax-parse.h
+++ b/phrase-extract/relax-parse.h
@@ -28,7 +28,7 @@
 #include <algorithm>
 #include <cstring>
 
-#include "SyntaxTree.h"
+#include "SyntaxNodeCollection.h"
 #include "XmlTree.h"
 
 #define LINE_MAX_LENGTH 1000000
@@ -39,8 +39,8 @@ char SAMTLevel = 0;
 
 // functions
 void init(int argc, char* argv[]);
-void store( MosesTraining::SyntaxTree &tree, const std::vector<std::string> &words );
-void LeftBinarize( MosesTraining::SyntaxTree &tree, MosesTraining::ParentNodes &parents );
-void RightBinarize( MosesTraining::SyntaxTree &tree, MosesTraining::ParentNodes &parents );
-void SAMT( MosesTraining::SyntaxTree &tree, MosesTraining::ParentNodes &parents );
+void store( MosesTraining::SyntaxNodeCollection &tree, const std::vector<std::string> &words );
+void LeftBinarize( MosesTraining::SyntaxNodeCollection &tree, MosesTraining::ParentNodes &parents );
+void RightBinarize( MosesTraining::SyntaxNodeCollection &tree, MosesTraining::ParentNodes &parents );
+void SAMT( MosesTraining::SyntaxNodeCollection &tree, MosesTraining::ParentNodes &parents );
 
diff --git a/phrase-extract/syntax-common/xml_tree_parser.cc b/phrase-extract/syntax-common/xml_tree_parser.cc
index c6e3cd3c3..2f8a904fa 100644
--- a/phrase-extract/syntax-common/xml_tree_parser.cc
+++ b/phrase-extract/syntax-common/xml_tree_parser.cc
@@ -13,17 +13,17 @@ namespace Syntax {
 
 StringTree *XmlTreeParser::Parse(const std::string &line) {
   line_ = line;
-  tree_.Clear();
+  node_collection_.Clear();
   try {
-    if (!ProcessAndStripXMLTags(line_, tree_, label_set_, top_label_set_,
-                                false)) {
+    if (!ProcessAndStripXMLTags(line_, node_collection_, label_set_,
+                                top_label_set_, false)) {
       throw Exception("");
     }
   } catch (const XmlException &e) {
     throw Exception(e.getMsg());
   }
-  tree_.ConnectNodes();
-  SyntaxNode *root = tree_.GetTop();
+  node_collection_.ConnectNodes();
+  SyntaxNode *root = node_collection_.GetTop();
   assert(root);
   words_ = util::tokenize(line_);
   return ConvertTree(*root, words_);
diff --git a/phrase-extract/syntax-common/xml_tree_parser.h b/phrase-extract/syntax-common/xml_tree_parser.h
index a5563f63a..c84ea25ec 100644
--- a/phrase-extract/syntax-common/xml_tree_parser.h
+++ b/phrase-extract/syntax-common/xml_tree_parser.h
@@ -5,7 +5,8 @@
 #include <string>
 #include <vector>
 
-#include "SyntaxTree.h"
+#include "SyntaxNode.h"
+#include "SyntaxNodeCollection.h"
 
 #include "exception.h"
 #include "string_tree.h"
@@ -26,7 +27,7 @@ class XmlTreeParser {
   std::set<std::string> label_set_;
   std::map<std::string, int> top_label_set_;
   std::string line_;
-  MosesTraining::SyntaxTree tree_;
+  MosesTraining::SyntaxNodeCollection node_collection_;
   std::vector<std::string> words_;
 };
 
diff --git a/scripts/OSM/OSM-Train.perl b/scripts/OSM/OSM-Train.perl
index 895a821db..07ad71f68 100755
--- a/scripts/OSM/OSM-Train.perl
+++ b/scripts/OSM/OSM-Train.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/OSM/extract-singletons.perl b/scripts/OSM/extract-singletons.perl
index 5a1665a8c..6295edfad 100755
--- a/scripts/OSM/extract-singletons.perl
+++ b/scripts/OSM/extract-singletons.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 #use strict;
 use warnings;
diff --git a/scripts/OSM/flipAlignment.perl b/scripts/OSM/flipAlignment.perl
index b896c0a23..57a1e9bb0 100755
--- a/scripts/OSM/flipAlignment.perl
+++ b/scripts/OSM/flipAlignment.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/Transliteration/clean.pl b/scripts/Transliteration/clean.pl
index ccc364fc9..7a08271da 100755
--- a/scripts/Transliteration/clean.pl
+++ b/scripts/Transliteration/clean.pl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 #input hindi word urdu word, delete all those entries that have number on any side
 use warnings;
@@ -314,4 +317,4 @@ sub charFreqFilter{
 		}
 	}
   }
-}
\ No newline at end of file
+}
diff --git a/scripts/Transliteration/corpusCreator.pl b/scripts/Transliteration/corpusCreator.pl
index 4c62449df..ac67f5d74 100755
--- a/scripts/Transliteration/corpusCreator.pl
+++ b/scripts/Transliteration/corpusCreator.pl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/Transliteration/in-decoding-transliteration.pl b/scripts/Transliteration/in-decoding-transliteration.pl
index c3cc31f26..e8130db02 100755
--- a/scripts/Transliteration/in-decoding-transliteration.pl
+++ b/scripts/Transliteration/in-decoding-transliteration.pl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/Transliteration/post-decoding-transliteration.pl b/scripts/Transliteration/post-decoding-transliteration.pl
index 60c3200f6..2c7908085 100755
--- a/scripts/Transliteration/post-decoding-transliteration.pl
+++ b/scripts/Transliteration/post-decoding-transliteration.pl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/Transliteration/prepare-transliteration-phrase-table.pl b/scripts/Transliteration/prepare-transliteration-phrase-table.pl
index df3b1ceca..0a9f554c5 100755
--- a/scripts/Transliteration/prepare-transliteration-phrase-table.pl
+++ b/scripts/Transliteration/prepare-transliteration-phrase-table.pl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/Transliteration/threshold.pl b/scripts/Transliteration/threshold.pl
index bf6657742..3baa8e0a7 100755
--- a/scripts/Transliteration/threshold.pl
+++ b/scripts/Transliteration/threshold.pl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use utf8;
diff --git a/scripts/Transliteration/train-transliteration-module.pl b/scripts/Transliteration/train-transliteration-module.pl
index 35e4ee396..b1d4d0ff5 100755
--- a/scripts/Transliteration/train-transliteration-module.pl
+++ b/scripts/Transliteration/train-transliteration-module.pl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use utf8;
diff --git a/scripts/analysis/bootstrap-hypothesis-difference-significance.pl b/scripts/analysis/bootstrap-hypothesis-difference-significance.pl
index 8e6a6255a..9a3f63d69 100755
--- a/scripts/analysis/bootstrap-hypothesis-difference-significance.pl
+++ b/scripts/analysis/bootstrap-hypothesis-difference-significance.pl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 use utf8;
 
 ###############################################
diff --git a/scripts/analysis/extract-target-trees.py b/scripts/analysis/extract-target-trees.py
index 3a92fdc4d..7166211d9 100755
--- a/scripts/analysis/extract-target-trees.py
+++ b/scripts/analysis/extract-target-trees.py
@@ -1,9 +1,13 @@
 #!/usr/bin/env python
-
-# Usage: extract-target-trees.py [FILE]
 #
-# Reads moses-chart's -T output from FILE or standard input and writes trees to
-# standard output in Moses' XML tree format.
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
+
+"""Usage: extract-target-trees.py [FILE]
+
+Reads moses-chart's -T output from FILE or standard input and writes trees to
+standard output in Moses' XML tree format.
+"""
 
 import re
 import sys
diff --git a/scripts/analysis/nontranslated_words.pl b/scripts/analysis/nontranslated_words.pl
index 51a4f9d20..7213deb76 100755
--- a/scripts/analysis/nontranslated_words.pl
+++ b/scripts/analysis/nontranslated_words.pl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 # $Id$
 # Reads a source and hypothesis file and counts equal tokens. Some of these
diff --git a/scripts/analysis/oov.pl b/scripts/analysis/oov.pl
index 052c9994d..9756887c9 100755
--- a/scripts/analysis/oov.pl
+++ b/scripts/analysis/oov.pl
@@ -1,6 +1,9 @@
 #!/usr/bin/env perl
 # Display OOV rate of a test set against a training corpus or a phrase table.
 # Ondrej Bojar
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use strict;
 use warnings;
diff --git a/scripts/analysis/sentence-by-sentence.pl b/scripts/analysis/sentence-by-sentence.pl
index 72b70dc72..b9eb6e56d 100755
--- a/scripts/analysis/sentence-by-sentence.pl
+++ b/scripts/analysis/sentence-by-sentence.pl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 # $Id$
 #sentence-by-sentence: take in a system output, with any number of factors, and a reference translation, also maybe with factors, and show each sentence and its errors
diff --git a/scripts/analysis/sg2dot.perl b/scripts/analysis/sg2dot.perl
index e9c1639ed..5f9a5ea1d 100755
--- a/scripts/analysis/sg2dot.perl
+++ b/scripts/analysis/sg2dot.perl
@@ -3,6 +3,8 @@
 # Author : Loic BARRAULT
 # Script to convert MOSES searchgraph to DOT format
 #
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/analysis/show-phrases-used.pl b/scripts/analysis/show-phrases-used.pl
index 522e6d3ff..9428ea9b8 100755
--- a/scripts/analysis/show-phrases-used.pl
+++ b/scripts/analysis/show-phrases-used.pl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 # $Id$
 #show-phrases-used: display all source and target phrases for each sentence in a corpus, and give average phrase length used
diff --git a/scripts/analysis/smtgui/Corpus.pm b/scripts/analysis/smtgui/Corpus.pm
index f050a9f6d..2391a6c15 100644
--- a/scripts/analysis/smtgui/Corpus.pm
+++ b/scripts/analysis/smtgui/Corpus.pm
@@ -1,5 +1,8 @@
 #package Corpus: hold a bunch of sentences in any language, with translation factors and stats about individual sentences and the corpus as a whole
 #Evan Herbst, 7 / 25 / 06
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 package Corpus;
 BEGIN
diff --git a/scripts/analysis/smtgui/filter-phrase-table.pl b/scripts/analysis/smtgui/filter-phrase-table.pl
index 55f2619c0..cd0f6b91b 100755
--- a/scripts/analysis/smtgui/filter-phrase-table.pl
+++ b/scripts/analysis/smtgui/filter-phrase-table.pl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 # $Id$
 #by Philipp Koehn, de-augmented by Evan Herbst
diff --git a/scripts/analysis/smtgui/newsmtgui.cgi b/scripts/analysis/smtgui/newsmtgui.cgi
index 32ad3a948..034ee265e 100755
--- a/scripts/analysis/smtgui/newsmtgui.cgi
+++ b/scripts/analysis/smtgui/newsmtgui.cgi
@@ -1,4 +1,7 @@
 #!/usr/bin/perl -w
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 # $Id$
 use strict;
diff --git a/scripts/analysis/suspicious_tokenization.pl b/scripts/analysis/suspicious_tokenization.pl
index 3ea15154e..f807153d9 100755
--- a/scripts/analysis/suspicious_tokenization.pl
+++ b/scripts/analysis/suspicious_tokenization.pl
@@ -2,6 +2,9 @@
 # Collects and prints all n-grams that appear in the given corpus both
 # tokenized as well as untokenized.
 # Ondrej Bojar
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use strict;
 use warnings;
diff --git a/scripts/analysis/weight-scan-summarize.sh b/scripts/analysis/weight-scan-summarize.sh
index 237182736..2fccb6470 100755
--- a/scripts/analysis/weight-scan-summarize.sh
+++ b/scripts/analysis/weight-scan-summarize.sh
@@ -1,4 +1,8 @@
 #!/bin/bash
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
+
 # Hackish summarization of weight-scan.pl results, heavily relies on tools by
 # Ondrej Bojar (bojar@ufal.mff.cuni.cz), some of which need Mercury; beware.
 
diff --git a/scripts/analysis/weight-scan.pl b/scripts/analysis/weight-scan.pl
index b33360694..b51a6bcd1 100755
--- a/scripts/analysis/weight-scan.pl
+++ b/scripts/analysis/weight-scan.pl
@@ -1,4 +1,8 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
+
 # runs Moses many times changing the values of one weight, all others fixed
 # nbest lists are always produced to allow for comparison of real and
 # 'projected' BLEU (BLEU estimated from n-best lists collected at a neighouring
diff --git a/scripts/ems/experiment.perl b/scripts/ems/experiment.perl
index ece110fbc..a3f5310a5 100755
--- a/scripts/ems/experiment.perl
+++ b/scripts/ems/experiment.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 # Experiment Management System
 # Documentation at http://www.statmt.org/moses/?n=FactoredTraining.EMS
diff --git a/scripts/ems/fix-info.perl b/scripts/ems/fix-info.perl
index abe58fe83..6659027b2 100755
--- a/scripts/ems/fix-info.perl
+++ b/scripts/ems/fix-info.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/ems/support/analysis.perl b/scripts/ems/support/analysis.perl
index f4d5a55b4..01bb21773 100755
--- a/scripts/ems/support/analysis.perl
+++ b/scripts/ems/support/analysis.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/ems/support/berkeley-process.sh b/scripts/ems/support/berkeley-process.sh
index e68056c96..347ebba3c 100755
--- a/scripts/ems/support/berkeley-process.sh
+++ b/scripts/ems/support/berkeley-process.sh
@@ -1,4 +1,7 @@
 #!/bin/sh
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 if [ $# -lt 8 ]
 	then
diff --git a/scripts/ems/support/berkeley-train.sh b/scripts/ems/support/berkeley-train.sh
index 96f6b648c..530cf978f 100755
--- a/scripts/ems/support/berkeley-train.sh
+++ b/scripts/ems/support/berkeley-train.sh
@@ -1,4 +1,7 @@
 #!/bin/sh
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 if [ $# -lt 6 ]
 	then
diff --git a/scripts/ems/support/build-domain-file-from-subcorpora.perl b/scripts/ems/support/build-domain-file-from-subcorpora.perl
index 085fd2629..f45b5ba2a 100755
--- a/scripts/ems/support/build-domain-file-from-subcorpora.perl
+++ b/scripts/ems/support/build-domain-file-from-subcorpora.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/ems/support/build-sparse-features.perl b/scripts/ems/support/build-sparse-features.perl
index 79fc1e394..b134cee69 100755
--- a/scripts/ems/support/build-sparse-features.perl
+++ b/scripts/ems/support/build-sparse-features.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/ems/support/consolidate-training-data.perl b/scripts/ems/support/consolidate-training-data.perl
index 4ab7f82cf..2a732be77 100755
--- a/scripts/ems/support/consolidate-training-data.perl
+++ b/scripts/ems/support/consolidate-training-data.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 # $Id: consolidate-training-data.perl 928 2009-09-02 02:58:01Z philipp $
 
diff --git a/scripts/ems/support/defaultconfig.py b/scripts/ems/support/defaultconfig.py
index a118e96b3..53913da08 100644
--- a/scripts/ems/support/defaultconfig.py
+++ b/scripts/ems/support/defaultconfig.py
@@ -1,4 +1,7 @@
 #!/usr/bin/env python2
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 """Version of ConfigParser which accepts default values."""
 
diff --git a/scripts/ems/support/fast-align-in-parts.perl b/scripts/ems/support/fast-align-in-parts.perl
index f777d7e52..bc340a50f 100755
--- a/scripts/ems/support/fast-align-in-parts.perl
+++ b/scripts/ems/support/fast-align-in-parts.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 #######################
 # Revision history
diff --git a/scripts/ems/support/generic-multicore-parallelizer.perl b/scripts/ems/support/generic-multicore-parallelizer.perl
index 0f7910603..d821aa114 100755
--- a/scripts/ems/support/generic-multicore-parallelizer.perl
+++ b/scripts/ems/support/generic-multicore-parallelizer.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/ems/support/generic-parallelizer.perl b/scripts/ems/support/generic-parallelizer.perl
index 811a99bde..087498ccf 100755
--- a/scripts/ems/support/generic-parallelizer.perl
+++ b/scripts/ems/support/generic-parallelizer.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/ems/support/input-from-sgm.perl b/scripts/ems/support/input-from-sgm.perl
index 18000581a..eb6a2e3a1 100755
--- a/scripts/ems/support/input-from-sgm.perl
+++ b/scripts/ems/support/input-from-sgm.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/ems/support/interpolate-lm.perl b/scripts/ems/support/interpolate-lm.perl
index 7d52fd877..4d9a513f6 100755
--- a/scripts/ems/support/interpolate-lm.perl
+++ b/scripts/ems/support/interpolate-lm.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/ems/support/lmplz-wrapper.perl b/scripts/ems/support/lmplz-wrapper.perl
index df503754f..89b2847d6 100755
--- a/scripts/ems/support/lmplz-wrapper.perl
+++ b/scripts/ems/support/lmplz-wrapper.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/ems/support/mml-filter.perl b/scripts/ems/support/mml-filter.perl
index 51bc4cda5..32bca335b 100755
--- a/scripts/ems/support/mml-filter.perl
+++ b/scripts/ems/support/mml-filter.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/ems/support/mml-score.perl b/scripts/ems/support/mml-score.perl
index 6f7b724ea..f88021818 100755
--- a/scripts/ems/support/mml-score.perl
+++ b/scripts/ems/support/mml-score.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/ems/support/mml-train.perl b/scripts/ems/support/mml-train.perl
index dcc998711..bdf6c1c1a 100755
--- a/scripts/ems/support/mml-train.perl
+++ b/scripts/ems/support/mml-train.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/ems/support/prepare-fast-align.perl b/scripts/ems/support/prepare-fast-align.perl
index 80fec36b2..68b1f0189 100755
--- a/scripts/ems/support/prepare-fast-align.perl
+++ b/scripts/ems/support/prepare-fast-align.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/ems/support/reference-from-sgm.perl b/scripts/ems/support/reference-from-sgm.perl
index ebb9ae4ae..b8e1d108d 100755
--- a/scripts/ems/support/reference-from-sgm.perl
+++ b/scripts/ems/support/reference-from-sgm.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/ems/support/remove-segmentation-markup.perl b/scripts/ems/support/remove-segmentation-markup.perl
index a0bd61fff..3b02bceaf 100755
--- a/scripts/ems/support/remove-segmentation-markup.perl
+++ b/scripts/ems/support/remove-segmentation-markup.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/ems/support/report-experiment-scores.perl b/scripts/ems/support/report-experiment-scores.perl
index b649951ce..c859508cb 100755
--- a/scripts/ems/support/report-experiment-scores.perl
+++ b/scripts/ems/support/report-experiment-scores.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 # $Id: report-experiment-scores.perl 407 2008-11-10 14:43:31Z philipp $
 
diff --git a/scripts/ems/support/run-command-on-multiple-refsets.perl b/scripts/ems/support/run-command-on-multiple-refsets.perl
index 1e914b44b..41823b4ee 100755
--- a/scripts/ems/support/run-command-on-multiple-refsets.perl
+++ b/scripts/ems/support/run-command-on-multiple-refsets.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/ems/support/run-wade.perl b/scripts/ems/support/run-wade.perl
index 175948b98..dfdb8e59d 100755
--- a/scripts/ems/support/run-wade.perl
+++ b/scripts/ems/support/run-wade.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/ems/support/split-sentences.perl b/scripts/ems/support/split-sentences.perl
index 02a1e2315..f72767054 100755
--- a/scripts/ems/support/split-sentences.perl
+++ b/scripts/ems/support/split-sentences.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 # Based on Preprocessor written by Philipp Koehn
 
diff --git a/scripts/ems/support/submit-grid.perl b/scripts/ems/support/submit-grid.perl
index a0967f9a5..ff43cd123 100755
--- a/scripts/ems/support/submit-grid.perl
+++ b/scripts/ems/support/submit-grid.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/ems/support/substitute-filtered-tables-and-weights.perl b/scripts/ems/support/substitute-filtered-tables-and-weights.perl
index 13be52c6b..2e6908ab4 100755
--- a/scripts/ems/support/substitute-filtered-tables-and-weights.perl
+++ b/scripts/ems/support/substitute-filtered-tables-and-weights.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/ems/support/substitute-filtered-tables.perl b/scripts/ems/support/substitute-filtered-tables.perl
index c5ebabded..548982592 100755
--- a/scripts/ems/support/substitute-filtered-tables.perl
+++ b/scripts/ems/support/substitute-filtered-tables.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 
diff --git a/scripts/ems/support/substitute-weights.perl b/scripts/ems/support/substitute-weights.perl
index b692f3f85..efa9338ca 100755
--- a/scripts/ems/support/substitute-weights.perl
+++ b/scripts/ems/support/substitute-weights.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 
diff --git a/scripts/ems/support/symmetrize-fast-align.perl b/scripts/ems/support/symmetrize-fast-align.perl
index 9f7fec248..4ed3e087d 100755
--- a/scripts/ems/support/symmetrize-fast-align.perl
+++ b/scripts/ems/support/symmetrize-fast-align.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/ems/support/thot-lm-wrapper.perl b/scripts/ems/support/thot-lm-wrapper.perl
index 59d483e65..ffbcb50e2 100755
--- a/scripts/ems/support/thot-lm-wrapper.perl
+++ b/scripts/ems/support/thot-lm-wrapper.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/ems/support/tree-converter-wrapper.perl b/scripts/ems/support/tree-converter-wrapper.perl
index aae55991a..ae7e2c5a6 100755
--- a/scripts/ems/support/tree-converter-wrapper.perl
+++ b/scripts/ems/support/tree-converter-wrapper.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/ems/support/wrap-xml.perl b/scripts/ems/support/wrap-xml.perl
index 52190309a..09ea2a2f8 100755
--- a/scripts/ems/support/wrap-xml.perl
+++ b/scripts/ems/support/wrap-xml.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/ems/web/analysis.php b/scripts/ems/web/analysis.php
index 57776dd22..5e5f707f6 100644
--- a/scripts/ems/web/analysis.php
+++ b/scripts/ems/web/analysis.php
@@ -1,5 +1,10 @@
 <?php
 
+/*
+This file is part of moses.  Its use is licensed under the GNU Lesser General
+Public License version 2.1 or, at your option, any later version.
+*/
+
 # main page frame, triggers the loading of parts
 function show_analysis() {
   global $task,$user,$setup,$id,$set;
diff --git a/scripts/ems/web/analysis_diff.php b/scripts/ems/web/analysis_diff.php
index 214ae1592..9840de7c2 100644
--- a/scripts/ems/web/analysis_diff.php
+++ b/scripts/ems/web/analysis_diff.php
@@ -1,5 +1,9 @@
 <?php
 
+/*
+This file is part of moses.  Its use is licensed under the GNU Lesser General
+Public License version 2.1 or, at your option, any later version.
+*/
 function diff_analysis() {
   global $task,$user,$setup,$id,$id2,$set;
   global $comment,$dir;
diff --git a/scripts/ems/web/diff.php b/scripts/ems/web/diff.php
index f440d3240..c290e7c02 100644
--- a/scripts/ems/web/diff.php
+++ b/scripts/ems/web/diff.php
@@ -1,5 +1,10 @@
 <?php
 
+/*
+This file is part of moses.  Its use is licensed under the GNU Lesser General
+Public License version 2.1 or, at your option, any later version.
+*/
+
 function diff() {
   global $experiment;
   $display = $_GET["run"];
diff --git a/scripts/ems/web/hierarchical-segmentation.js b/scripts/ems/web/hierarchical-segmentation.js
index fad5a16d4..7f0df85ff 100644
--- a/scripts/ems/web/hierarchical-segmentation.js
+++ b/scripts/ems/web/hierarchical-segmentation.js
@@ -1,3 +1,7 @@
+/*
+This file is part of moses.  Its use is licensed under the GNU Lesser General
+Public License version 2.1 or, at your option, any later version.
+*/
 var nodeIn = [];
 var nodeOut = [];
 var nodeChildren = [];
diff --git a/scripts/ems/web/index.php b/scripts/ems/web/index.php
index 9c918a96a..757c20d72 100644
--- a/scripts/ems/web/index.php
+++ b/scripts/ems/web/index.php
@@ -1,5 +1,10 @@
 <?php
 
+/*
+This file is part of moses.  Its use is licensed under the GNU Lesser General
+Public License version 2.1 or, at your option, any later version.
+*/
+
 require("lib.php");
 require("overview.php");
 require("analysis.php");
diff --git a/scripts/ems/web/lib.php b/scripts/ems/web/lib.php
index 68c58860b..df3286ebb 100644
--- a/scripts/ems/web/lib.php
+++ b/scripts/ems/web/lib.php
@@ -1,5 +1,10 @@
 <?php
 
+/*
+This file is part of moses.  Its use is licensed under the GNU Lesser General
+Public License version 2.1 or, at your option, any later version.
+*/
+
 function load_experiment_info() {
   global $dir,$task,$user,$setup;
   global $evalset;
diff --git a/scripts/ems/web/overview.php b/scripts/ems/web/overview.php
index 534c7d8c0..c59970b67 100644
--- a/scripts/ems/web/overview.php
+++ b/scripts/ems/web/overview.php
@@ -1,5 +1,9 @@
 <?php
 
+/*
+This file is part of moses.  Its use is licensed under the GNU Lesser General
+Public License version 2.1 or, at your option, any later version.
+*/
 
 function setup() {
   $setup = file("setup");
diff --git a/scripts/ems/web/progress.perl b/scripts/ems/web/progress.perl
index fa2ce9e8f..f5acfc64f 100755
--- a/scripts/ems/web/progress.perl
+++ b/scripts/ems/web/progress.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/ems/web/sgviz.js b/scripts/ems/web/sgviz.js
index 13d45f8c5..3926f9513 100644
--- a/scripts/ems/web/sgviz.js
+++ b/scripts/ems/web/sgviz.js
@@ -1,3 +1,7 @@
+/*
+This file is part of moses.  Its use is licensed under the GNU Lesser General
+Public License version 2.1 or, at your option, any later version.
+*/
 var xmlns="http://www.w3.org/2000/svg";
 var RECOMBINED = 0;
 var FROM = 1;
diff --git a/scripts/ems/web/sgviz.php b/scripts/ems/web/sgviz.php
index 9fccadf60..85778caab 100644
--- a/scripts/ems/web/sgviz.php
+++ b/scripts/ems/web/sgviz.php
@@ -1,4 +1,10 @@
 <?php
+
+/*
+This file is part of moses.  Its use is licensed under the GNU Lesser General
+Public License version 2.1 or, at your option, any later version.
+*/
+
 function sgviz($sentence) {
   global $setup,$dir,$id,$set;
 ?><html><head><title>Search Graph Visualization, Sentence <?php $sentence ?></title>
diff --git a/scripts/fuzzy-match/create_xml.perl b/scripts/fuzzy-match/create_xml.perl
index 4ab281eae..97025d62a 100755
--- a/scripts/fuzzy-match/create_xml.perl
+++ b/scripts/fuzzy-match/create_xml.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 binmode( STDIN,  ":utf8" );
 binmode( STDOUT, ":utf8" );
diff --git a/scripts/generic/bsbleu.py b/scripts/generic/bsbleu.py
index 12d2201de..296900b18 100755
--- a/scripts/generic/bsbleu.py
+++ b/scripts/generic/bsbleu.py
@@ -1,6 +1,9 @@
 #!/usr/bin/env python
 # compute Bleu scores with confidence intervals via boostrap resampling
 # written by Ulrich Germann
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 from argparse import ArgumentParser
 import math
diff --git a/scripts/generic/compound-splitter.perl b/scripts/generic/compound-splitter.perl
index b39d4d660..2ece80a60 100755
--- a/scripts/generic/compound-splitter.perl
+++ b/scripts/generic/compound-splitter.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/generic/extract-factors.pl b/scripts/generic/extract-factors.pl
index 38cf97bd4..2b1c51cd1 100755
--- a/scripts/generic/extract-factors.pl
+++ b/scripts/generic/extract-factors.pl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 # $Id$
 #extract-factors.pl: extract only the desired factors from a factored corpus
diff --git a/scripts/generic/extract-parallel.perl b/scripts/generic/extract-parallel.perl
index be30ff652..3240f24eb 100755
--- a/scripts/generic/extract-parallel.perl
+++ b/scripts/generic/extract-parallel.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 # example
 #  ./extract-parallel.perl 8 ./coreutils-8.9/src/split "./coreutils-8.9/src/sort --batch-size=253" ./extract ./corpus.5.en ./corpus.5.ar ./align.ar-en.grow-diag-final-and ./extracted 7 --NoFileLimit orientation --GZOutput
diff --git a/scripts/generic/fsa2fsal.pl b/scripts/generic/fsa2fsal.pl
index 7dc7751ee..28ec28a26 100755
--- a/scripts/generic/fsa2fsal.pl
+++ b/scripts/generic/fsa2fsal.pl
@@ -4,6 +4,9 @@
 # ' ' to delimit nodes (i.e. original lines).
 # Some rudimentary sanity checks are done on the fly.
 # Ondrej Bojar, bojar@ufal.mff.cuni.cz
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/generic/fsa2plf.pl b/scripts/generic/fsa2plf.pl
index 07c8a4cc1..4b9474d5a 100755
--- a/scripts/generic/fsa2plf.pl
+++ b/scripts/generic/fsa2plf.pl
@@ -7,6 +7,9 @@
 # final nodes.
 # Note that the output format may not contain any spaces.
 # Ondrej Bojar, bojar@ufal.mff.cuni.cz
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/generic/fsal2fsa.pl b/scripts/generic/fsal2fsa.pl
index a21305dad..158dab5b3 100755
--- a/scripts/generic/fsal2fsa.pl
+++ b/scripts/generic/fsal2fsa.pl
@@ -1,6 +1,9 @@
 #!/usr/bin/env perl
 # A very simple script that converts fsal back to fsa format (openfst lattices)
 # Ondrej Bojar, bojar@ufal.mff.cuni.cz
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/generic/generic-parallel.perl b/scripts/generic/generic-parallel.perl
index a9bc73d85..07f6a210a 100755
--- a/scripts/generic/generic-parallel.perl
+++ b/scripts/generic/generic-parallel.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/generic/giza-parallel.perl b/scripts/generic/giza-parallel.perl
index 9a6516a8f..a9921a992 100755
--- a/scripts/generic/giza-parallel.perl
+++ b/scripts/generic/giza-parallel.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 # example
 # ~/giza-parallel.perl 10 split ~/workspace/sourceforge/trunk/scripts/training/train-model.perl ar en train align
diff --git a/scripts/generic/lopar2pos.pl b/scripts/generic/lopar2pos.pl
index 2b9245e0f..fc2c35c7f 100755
--- a/scripts/generic/lopar2pos.pl
+++ b/scripts/generic/lopar2pos.pl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 # $Id$
 #lopar2pos: extract POSs from LOPAR output
diff --git a/scripts/generic/moses-parallel.pl b/scripts/generic/moses-parallel.pl
index eb51daa98..144b7d6b2 100755
--- a/scripts/generic/moses-parallel.pl
+++ b/scripts/generic/moses-parallel.pl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 # $Id$
 #######################
diff --git a/scripts/generic/moses_sim_pe.py b/scripts/generic/moses_sim_pe.py
index 32f785961..3497ca558 100755
--- a/scripts/generic/moses_sim_pe.py
+++ b/scripts/generic/moses_sim_pe.py
@@ -1,20 +1,25 @@
 #!/usr/bin/env python
 
 # Written by Michael Denkowski
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
-# This script parallelizes decoding with simulated post-editing via moses XML
-# input (XML entities need to be escaped in tokenization).  Memory mapped
-# dynamic phrase tables (Ulrich Germann,
-# www.statmt.org/moses/?n=Moses.AdvancedFeatures#ntoc40) and language models
-# (Kenneth Heafield,
-# http://www.statmt.org/moses/?n=FactoredTraining.BuildingLanguageModel#ntoc19)
-# facilitate memory efficient multi process decoding.  Input is divided into
-# batches, each of which is decoded sequentially.  Each batch pre-loads the
-# data from previous batches.
+"""Parallelize decoding with simulated post-editing via moses XML input.
 
-# To use in tuning, run mert-moses.pl with --sim-pe=SYMAL where SYMAL is the
-# alignment from input to references.  Specify the number of jobs with
-# --decoder-flags="-threads N".
+(XML entities need to be escaped in tokenization).  Memory mapped
+dynamic phrase tables (Ulrich Germann,
+www.statmt.org/moses/?n=Moses.AdvancedFeatures#ntoc40) and language models
+(Kenneth Heafield,
+http://www.statmt.org/moses/?n=FactoredTraining.BuildingLanguageModel#ntoc19)
+facilitate memory efficient multi process decoding.  Input is divided into
+batches, each of which is decoded sequentially.  Each batch pre-loads the
+data from previous batches.
+
+To use in tuning, run mert-moses.pl with --sim-pe=SYMAL where SYMAL is the
+alignment from input to references.  Specify the number of jobs with
+--decoder-flags="-threads N".
+"""
 
 import gzip
 import itertools
diff --git a/scripts/generic/mteval-v12.pl b/scripts/generic/mteval-v12.pl
index 2666c8012..b4dfbf83a 100755
--- a/scripts/generic/mteval-v12.pl
+++ b/scripts/generic/mteval-v12.pl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/generic/mteval-v13a.pl b/scripts/generic/mteval-v13a.pl
index 41a88800a..bdc2d9479 100755
--- a/scripts/generic/mteval-v13a.pl
+++ b/scripts/generic/mteval-v13a.pl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/generic/multi-bleu.perl b/scripts/generic/multi-bleu.perl
index 344f58c6f..61de10d45 100755
--- a/scripts/generic/multi-bleu.perl
+++ b/scripts/generic/multi-bleu.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 # $Id$
 use warnings;
diff --git a/scripts/generic/ph_numbers.perl b/scripts/generic/ph_numbers.perl
index 612263249..f0ae1f851 100755
--- a/scripts/generic/ph_numbers.perl
+++ b/scripts/generic/ph_numbers.perl
@@ -6,6 +6,9 @@ package ph_numbers;
 # and decoder input
 #
 # (c) 2013 TAUS
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/generic/qsub-wrapper.pl b/scripts/generic/qsub-wrapper.pl
index ac3d0900a..ef9938e07 100755
--- a/scripts/generic/qsub-wrapper.pl
+++ b/scripts/generic/qsub-wrapper.pl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 # $Id$
 use warnings;
diff --git a/scripts/generic/reverse-alignment.perl b/scripts/generic/reverse-alignment.perl
index 681b3221e..f01acf5b0 100755
--- a/scripts/generic/reverse-alignment.perl
+++ b/scripts/generic/reverse-alignment.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/generic/score-parallel.perl b/scripts/generic/score-parallel.perl
index 81bc6f7d0..625b449c0 100755
--- a/scripts/generic/score-parallel.perl
+++ b/scripts/generic/score-parallel.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 # example
 # ./score-parallel.perl 8 "gsort --batch-size=253" ./score ./extract.2.sorted.gz ./lex.2.f2e ./phrase-table.2.half.f2e  --GoodTuring ./phrase-table.2.coc 0
diff --git a/scripts/generic/strip-xml.perl b/scripts/generic/strip-xml.perl
index c993421f0..a5dbbaa37 100755
--- a/scripts/generic/strip-xml.perl
+++ b/scripts/generic/strip-xml.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/generic/trainlm-irst2.perl b/scripts/generic/trainlm-irst2.perl
index f664e96ee..8af372fac 100755
--- a/scripts/generic/trainlm-irst2.perl
+++ b/scripts/generic/trainlm-irst2.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 # Compatible with sri LM-creating script, eg.
 #    ngram-count -order 5 -interpolate -wbdiscount -unk -text corpus.txt -lm lm.txt
diff --git a/scripts/other/beautify.py b/scripts/other/beautify.py
index f03a58ce7..56df24bc8 100755
--- a/scripts/other/beautify.py
+++ b/scripts/other/beautify.py
@@ -1,4 +1,9 @@
 #! /usr/bin/env python
+#
+# Originally written in 2015 by Jeroen Vermeulen (Precision Translation Tools).
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 """Reformat project source code, and/or check for style errors ("lint").
 
@@ -38,6 +43,17 @@ BEAUTIFY_IGNORE = '.beautify-ignore'
 
 class LintCheckFailure(Exception):
     """Lint was found, or the lint checker otherwise returned failure."""
+    exit_code = 1
+
+
+class ProgramFailure(Exception):
+    """The program failed, but it's not a bug.  No traceback."""
+    exit_code = 2
+
+
+class CommandLineError(Exception):
+    """Something wrong with the command-line arguments."""
+    exit_code = 3
 
 
 def read_ignore_file(root_dir):
@@ -52,7 +68,7 @@ def read_ignore_file(root_dir):
             ignore_contents = ignore_file.read()
     except IOError as error:
         if error.errno == ENOENT:
-            raise Exception(
+            raise ProgramFailure(
                 "No .gitignore file found in %s.  "
                 "Is it really the project's root directory?"
                 % root_dir)
@@ -200,7 +216,7 @@ def check_astyle_version(verbose=False):
         ['astyle', '--version'], verbose=verbose, env={'LC_ALL': 'C'})
     version = version.strip()
     if version != EXPECTED_ASTYLE_VERSION:
-        raise Exception(
+        raise ProgramFailure(
             "Wrong astyle version.  "
             "Expected '%s', but got version string '%s'."
             % (EXPECTED_ASTYLE_VERSION, version))
@@ -226,8 +242,15 @@ def run_perltidy(source_files, verbose=False, dry_run=False):
         # Write "} else {", with 'else' on the same line as the braces.
         '--cuddled-else',
     ]
-    _, stderr = run_command(
-        command_line + source_files, verbose=verbose, dry_run=dry_run)
+    try:
+        _, stderr = run_command(
+            command_line + source_files, verbose=verbose, dry_run=dry_run)
+    except OSError as error:
+        if error.errno == ENOENT:
+            raise ProgramFailure(
+                "Could not run 'perltidy'.  Make sure that it is installed.")
+        else:
+            raise
     if stderr != '':
         sys.stderr.write(stderr)
 
@@ -386,7 +409,7 @@ def main():
     """Find and format source files."""
     args = parse_arguments()
     if not args.format and not args.lint:
-        raise Exception("Select action: --format, --lint, or both.")
+        raise CommandLineError("Select action: --format, --lint, or both.")
 
     ignore = read_ignore_file(args.root_dir)
 
@@ -409,8 +432,8 @@ def main():
 if __name__ == '__main__':
     try:
         main()
-    except LintCheckFailure as error:
+    except (CommandLineError, LintCheckFailure, ProgramFailure) as error:
         # This is a failure, but not a bug.  Print a friendly error
         # message, not a traceback.
         sys.stderr.write('%s\n' % error)
-        sys.exit(1)
+        sys.exit(error.exit_code)
diff --git a/scripts/other/convert-pt.perl b/scripts/other/convert-pt.perl
index e087126f1..60c8cbdb2 100755
--- a/scripts/other/convert-pt.perl
+++ b/scripts/other/convert-pt.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 # $Id$
 # convert a phrase-table with alignment in Moses' dead-end format
diff --git a/scripts/other/delete-scores.perl b/scripts/other/delete-scores.perl
index ffb788867..ebaf277fa 100755
--- a/scripts/other/delete-scores.perl
+++ b/scripts/other/delete-scores.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/other/gacha_filter.py b/scripts/other/gacha_filter.py
index 0deb45761..af5921d41 100644
--- a/scripts/other/gacha_filter.py
+++ b/scripts/other/gacha_filter.py
@@ -1,4 +1,7 @@
 #!/usr/bin/env python3 -*- coding: utf-8 -*-
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 """
 The Gacha filter cleans out sentence pairs that have global character mean
diff --git a/scripts/other/get_many_translations_from_google.perl b/scripts/other/get_many_translations_from_google.perl
index 0b1436c20..ac2933296 100755
--- a/scripts/other/get_many_translations_from_google.perl
+++ b/scripts/other/get_many_translations_from_google.perl
@@ -5,6 +5,9 @@
 # Expects one sentence per line, not tokenized!
 #
 # Ondrej Bojar, bojar@ufal.mff.cuni.cz
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/other/retain-lines.perl b/scripts/other/retain-lines.perl
index f04a8ebad..c789f96c7 100755
--- a/scripts/other/retain-lines.perl
+++ b/scripts/other/retain-lines.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 #retain lines in clean.lines-retained.1
 use strict;
diff --git a/scripts/other/translate_by_microsoft_bing.perl b/scripts/other/translate_by_microsoft_bing.perl
index c9b1b31de..d4222878e 100755
--- a/scripts/other/translate_by_microsoft_bing.perl
+++ b/scripts/other/translate_by_microsoft_bing.perl
@@ -2,6 +2,9 @@
 
 # Script implemented by Pranava Swaroop Madhyastha (a student at Charles
 # University, UFAL)
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use strict;
 use warnings;
diff --git a/scripts/recaser/detruecase.perl b/scripts/recaser/detruecase.perl
index b882852a0..66ca24fa2 100755
--- a/scripts/recaser/detruecase.perl
+++ b/scripts/recaser/detruecase.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/recaser/recase.perl b/scripts/recaser/recase.perl
index 52cec36ea..b951ca764 100755
--- a/scripts/recaser/recase.perl
+++ b/scripts/recaser/recase.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 # $Id$
 use warnings;
diff --git a/scripts/recaser/train-recaser.perl b/scripts/recaser/train-recaser.perl
index dce388bca..cb3388c38 100755
--- a/scripts/recaser/train-recaser.perl
+++ b/scripts/recaser/train-recaser.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 # $Id$
 use warnings;
diff --git a/scripts/recaser/train-truecaser.perl b/scripts/recaser/train-truecaser.perl
index 753183324..7f8909082 100755
--- a/scripts/recaser/train-truecaser.perl
+++ b/scripts/recaser/train-truecaser.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 # $Id: train-recaser.perl 1326 2007-03-26 05:44:27Z bojar $
 
diff --git a/scripts/recaser/truecase.perl b/scripts/recaser/truecase.perl
index 544b79c47..aab185ce9 100755
--- a/scripts/recaser/truecase.perl
+++ b/scripts/recaser/truecase.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 # $Id: train-recaser.perl 1326 2007-03-26 05:44:27Z bojar $
 
diff --git a/scripts/regression-testing/MosesScriptsRegressionTesting.pm b/scripts/regression-testing/MosesScriptsRegressionTesting.pm
index d8b0590c8..acc134d70 100644
--- a/scripts/regression-testing/MosesScriptsRegressionTesting.pm
+++ b/scripts/regression-testing/MosesScriptsRegressionTesting.pm
@@ -1,3 +1,6 @@
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
+
 package MosesScriptsRegressionTesting;
 
 use strict;
diff --git a/scripts/regression-testing/compare-results.pl b/scripts/regression-testing/compare-results.pl
index 572431951..8f1461cec 100755
--- a/scripts/regression-testing/compare-results.pl
+++ b/scripts/regression-testing/compare-results.pl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/regression-testing/create_localized_moses_ini.pl b/scripts/regression-testing/create_localized_moses_ini.pl
index 1d03e5ab8..3e2b6f37f 100755
--- a/scripts/regression-testing/create_localized_moses_ini.pl
+++ b/scripts/regression-testing/create_localized_moses_ini.pl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/regression-testing/modify-pars.pl b/scripts/regression-testing/modify-pars.pl
index de2df2919..7726af9e6 100755
--- a/scripts/regression-testing/modify-pars.pl
+++ b/scripts/regression-testing/modify-pars.pl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/regression-testing/moses-virtual.pl b/scripts/regression-testing/moses-virtual.pl
index 3af3c79e4..3b23b525a 100755
--- a/scripts/regression-testing/moses-virtual.pl
+++ b/scripts/regression-testing/moses-virtual.pl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/regression-testing/run-single-test.pl b/scripts/regression-testing/run-single-test.pl
index e8307da36..037de8285 100755
--- a/scripts/regression-testing/run-single-test.pl
+++ b/scripts/regression-testing/run-single-test.pl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/regression-testing/run-test-suite.pl b/scripts/regression-testing/run-test-suite.pl
index b384f8b98..a12938e61 100755
--- a/scripts/regression-testing/run-test-suite.pl
+++ b/scripts/regression-testing/run-test-suite.pl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/server/moses.py b/scripts/server/moses.py
index 7cf152187..e825ab39e 100644
--- a/scripts/server/moses.py
+++ b/scripts/server/moses.py
@@ -1,5 +1,8 @@
 #!/usr/bin/env python
 # -*- coding: utf-8 -*-
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 """
 Python utilities for moses
diff --git a/scripts/server/sim-pe.py b/scripts/server/sim-pe.py
index 5f1407524..6f76bf46d 100755
--- a/scripts/server/sim-pe.py
+++ b/scripts/server/sim-pe.py
@@ -2,8 +2,14 @@
 # -*- coding: utf-8 -*-
 
 # Written by Ulrich Germann on the basis of contrib/server/client.py.
-# This script simulates post-editing of MT output and incrementally
-# updates the dynamic phrase tables in the moses server.
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
+
+"""Simulate post-editing of MT output.
+
+Incrementally updates the dynamic phrase tables in the moses server.
+"""
 
 import argparse
 import os
diff --git a/scripts/tokenizer/deescape-special-chars-PTB.perl b/scripts/tokenizer/deescape-special-chars-PTB.perl
index f9601924f..ad2529b21 100755
--- a/scripts/tokenizer/deescape-special-chars-PTB.perl
+++ b/scripts/tokenizer/deescape-special-chars-PTB.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/tokenizer/deescape-special-chars.perl b/scripts/tokenizer/deescape-special-chars.perl
index 002955e62..b9d1ad74c 100755
--- a/scripts/tokenizer/deescape-special-chars.perl
+++ b/scripts/tokenizer/deescape-special-chars.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/tokenizer/detokenizer.perl b/scripts/tokenizer/detokenizer.perl
index 3a92bd024..881b93dd1 100755
--- a/scripts/tokenizer/detokenizer.perl
+++ b/scripts/tokenizer/detokenizer.perl
@@ -4,6 +4,9 @@
 # Sample De-Tokenizer
 # written by Josh Schroeder, based on code by Philipp Koehn
 # further modifications by Ondrej Bojar
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 binmode(STDIN, ":utf8");
 binmode(STDOUT, ":utf8");
diff --git a/scripts/tokenizer/escape-special-chars.perl b/scripts/tokenizer/escape-special-chars.perl
index fbbbae292..143e85490 100755
--- a/scripts/tokenizer/escape-special-chars.perl
+++ b/scripts/tokenizer/escape-special-chars.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/tokenizer/lowercase.perl b/scripts/tokenizer/lowercase.perl
index e5c41bbed..bc75e5e5c 100755
--- a/scripts/tokenizer/lowercase.perl
+++ b/scripts/tokenizer/lowercase.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/tokenizer/normalize-punctuation.perl b/scripts/tokenizer/normalize-punctuation.perl
index 13e9fd3fc..7dab7543a 100755
--- a/scripts/tokenizer/normalize-punctuation.perl
+++ b/scripts/tokenizer/normalize-punctuation.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/tokenizer/pre-tok-clean.perl b/scripts/tokenizer/pre-tok-clean.perl
index 900e992ee..064f7b187 100755
--- a/scripts/tokenizer/pre-tok-clean.perl
+++ b/scripts/tokenizer/pre-tok-clean.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use strict;
 
diff --git a/scripts/tokenizer/pre-tokenizer.perl b/scripts/tokenizer/pre-tokenizer.perl
index 514d8da8d..541ce77fb 100755
--- a/scripts/tokenizer/pre-tokenizer.perl
+++ b/scripts/tokenizer/pre-tokenizer.perl
@@ -3,6 +3,9 @@
 # script for preprocessing language data prior to tokenization
 # Start by Ulrich Germann, after noticing systematic preprocessing errors
 # in some of the English Europarl data.
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/tokenizer/pre_tokenize_cleaning.py b/scripts/tokenizer/pre_tokenize_cleaning.py
index 096a45dc4..c03af8f66 100644
--- a/scripts/tokenizer/pre_tokenize_cleaning.py
+++ b/scripts/tokenizer/pre_tokenize_cleaning.py
@@ -1,4 +1,7 @@
 #!/usr/bin/env python -*- coding: utf-8 -*-
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 """
 The Gacha filter cleans out sentence pairs that have global character mean
diff --git a/scripts/tokenizer/remove-non-printing-char.perl b/scripts/tokenizer/remove-non-printing-char.perl
index 9125b7691..92f6ade16 100755
--- a/scripts/tokenizer/remove-non-printing-char.perl
+++ b/scripts/tokenizer/remove-non-printing-char.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use utf8;
diff --git a/scripts/tokenizer/replace-unicode-punctuation.perl b/scripts/tokenizer/replace-unicode-punctuation.perl
index cda69ddf7..c2c7088d6 100755
--- a/scripts/tokenizer/replace-unicode-punctuation.perl
+++ b/scripts/tokenizer/replace-unicode-punctuation.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/tokenizer/tokenizer.perl b/scripts/tokenizer/tokenizer.perl
index a5d4fadd3..e08bac941 100755
--- a/scripts/tokenizer/tokenizer.perl
+++ b/scripts/tokenizer/tokenizer.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 
diff --git a/scripts/tokenizer/tokenizer_PTB.perl b/scripts/tokenizer/tokenizer_PTB.perl
index 6fff8d7f7..46b14775c 100755
--- a/scripts/tokenizer/tokenizer_PTB.perl
+++ b/scripts/tokenizer/tokenizer_PTB.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 # Sample Tokenizer
 ### Version 1.1
diff --git a/scripts/training/LexicalTranslationModel.pm b/scripts/training/LexicalTranslationModel.pm
index c5dad60fb..3adc45f5e 100644
--- a/scripts/training/LexicalTranslationModel.pm
+++ b/scripts/training/LexicalTranslationModel.pm
@@ -1,3 +1,6 @@
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
+
 package LexicalTranslationModel;
 
 use strict;
diff --git a/scripts/training/absolutize_moses_model.pl b/scripts/training/absolutize_moses_model.pl
index bb7085895..27eccd8c7 100755
--- a/scripts/training/absolutize_moses_model.pl
+++ b/scripts/training/absolutize_moses_model.pl
@@ -5,6 +5,9 @@
 # paths with absolute paths.
 #
 # Ondrej Bojar.
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 
diff --git a/scripts/training/analyse_moses_model.pl b/scripts/training/analyse_moses_model.pl
index 656f4a59b..7a5c2e701 100755
--- a/scripts/training/analyse_moses_model.pl
+++ b/scripts/training/analyse_moses_model.pl
@@ -4,6 +4,9 @@
 # given a moses.ini file, checks the translation and generation tables and reports
 # statistics on ambiguity
 # Ondrej Bojar
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/training/bilingual-lm/averageNullEmbedding.py b/scripts/training/bilingual-lm/averageNullEmbedding.py
index 891595aff..54c9a1bc4 100755
--- a/scripts/training/bilingual-lm/averageNullEmbedding.py
+++ b/scripts/training/bilingual-lm/averageNullEmbedding.py
@@ -1,4 +1,7 @@
 #!/usr/bin/env python2
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 import sys
 import numpy
 import argparse
diff --git a/scripts/training/bilingual-lm/extract.py b/scripts/training/bilingual-lm/extract.py
index f620edb5d..876fba9ee 100755
--- a/scripts/training/bilingual-lm/extract.py
+++ b/scripts/training/bilingual-lm/extract.py
@@ -1,4 +1,7 @@
 #!/usr/bin/env python
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 from collections import Counter
 import logging
diff --git a/scripts/training/bilingual-lm/extract_test.py b/scripts/training/bilingual-lm/extract_test.py
index 3c9a03b85..8cade1e04 100755
--- a/scripts/training/bilingual-lm/extract_test.py
+++ b/scripts/training/bilingual-lm/extract_test.py
@@ -1,4 +1,7 @@
 #!/usr/bin/env python
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 """Create a test corpus, using a previously pruned vocabulary."""
 
diff --git a/scripts/training/bilingual-lm/extract_training.py b/scripts/training/bilingual-lm/extract_training.py
index bd3538188..e39a70318 100755
--- a/scripts/training/bilingual-lm/extract_training.py
+++ b/scripts/training/bilingual-lm/extract_training.py
@@ -1,4 +1,7 @@
 #!/usr/bin/env python
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 from collections import Counter
 import logging
diff --git a/scripts/training/bilingual-lm/reduce_ngrams.py b/scripts/training/bilingual-lm/reduce_ngrams.py
index 3442fb302..4db41378d 100755
--- a/scripts/training/bilingual-lm/reduce_ngrams.py
+++ b/scripts/training/bilingual-lm/reduce_ngrams.py
@@ -1,6 +1,9 @@
 #!/usr/bin/env python3
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
-"""Reduces an ngrams file for training nplm to a smaller version of it.
+"""Reduce an ngrams file for training nplm to a smaller version of it.
 
 The smaller version will have fewer ngrams.
 """
diff --git a/scripts/training/bilingual-lm/test_nplm.py b/scripts/training/bilingual-lm/test_nplm.py
index 737266bc3..3a59fd344 100755
--- a/scripts/training/bilingual-lm/test_nplm.py
+++ b/scripts/training/bilingual-lm/test_nplm.py
@@ -1,4 +1,7 @@
 #!/usr/bin/env python
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 import logging
 import optparse
diff --git a/scripts/training/bilingual-lm/train_nplm.py b/scripts/training/bilingual-lm/train_nplm.py
index 7bc74429e..572076006 100755
--- a/scripts/training/bilingual-lm/train_nplm.py
+++ b/scripts/training/bilingual-lm/train_nplm.py
@@ -1,4 +1,7 @@
 #!/usr/bin/env python
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 from __future__ import print_function, unicode_literals
 
@@ -36,7 +39,8 @@ parser.add_argument("--input-words-file", dest="input_words_file")
 parser.add_argument("--output-words-file", dest="output_words_file")
 parser.add_argument("--input_vocab_size", dest="input_vocab_size", type=int)
 parser.add_argument("--output_vocab_size", dest="output_vocab_size", type=int)
-
+parser.add_argument("--mmap", dest="mmap", action="store_true",
+    help="Use memory-mapped file (for lower memory consumption).")
 
 parser.set_defaults(
     working_dir="working",
@@ -110,6 +114,11 @@ def main(options):
         options.working_dir,
         os.path.basename(options.corpus_stem) + ".numberized")
 
+    mmap_command = []
+    if options.mmap:
+        in_file += '.mmap'
+        mmap_command = ['--mmap_file', '1']
+
     model_prefix = os.path.join(
         options.output_dir, options.output_model + ".model.nplm")
     train_args = [
@@ -124,9 +133,9 @@ def main(options):
         "--input_embedding_dimension", str(options.input_embedding),
         "--output_embedding_dimension", str(options.output_embedding),
         "--num_threads", str(options.threads),
-        "--activation_function",
-        options.activation_fn,
-    ] + validations_command + vocab_command
+        "--activation_function", options.activation_fn,
+        "--ngram_size", str(options.ngram_size),
+    ] + validations_command + vocab_command + mmap_command
     print("Train model command: ")
     print(', '.join(train_args))
 
diff --git a/scripts/training/binarize-model.perl b/scripts/training/binarize-model.perl
index 0239f5fc8..0131d2222 100755
--- a/scripts/training/binarize-model.perl
+++ b/scripts/training/binarize-model.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 #
 # Binarize a Moses model
diff --git a/scripts/training/build-generation-table.perl b/scripts/training/build-generation-table.perl
index 435f7f58e..14176908a 100755
--- a/scripts/training/build-generation-table.perl
+++ b/scripts/training/build-generation-table.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 # $Id$
 use warnings;
diff --git a/scripts/training/build-mmsapt.perl b/scripts/training/build-mmsapt.perl
index 00cbd09d6..d0c5b818e 100755
--- a/scripts/training/build-mmsapt.perl
+++ b/scripts/training/build-mmsapt.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/training/clean-corpus-n.perl b/scripts/training/clean-corpus-n.perl
index cee4c76a2..76a09e539 100755
--- a/scripts/training/clean-corpus-n.perl
+++ b/scripts/training/clean-corpus-n.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 # $Id: clean-corpus-n.perl 3633 2010-10-21 09:49:27Z phkoehn $
 use warnings;
diff --git a/scripts/training/clone_moses_model.pl b/scripts/training/clone_moses_model.pl
index bf6708fca..18dc4aa41 100755
--- a/scripts/training/clone_moses_model.pl
+++ b/scripts/training/clone_moses_model.pl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 # $Id$
 # given a moses.ini file, creates a fresh version of it
diff --git a/scripts/training/combine_factors.pl b/scripts/training/combine_factors.pl
index fa6f15db2..fcc9ab3f5 100755
--- a/scripts/training/combine_factors.pl
+++ b/scripts/training/combine_factors.pl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 # $Id$
 # given a list of files, combines them to a single corpus (sent to stdout)
diff --git a/scripts/training/convert-moses-ini-to-v2.perl b/scripts/training/convert-moses-ini-to-v2.perl
index e091a710d..3fdfa53a6 100755
--- a/scripts/training/convert-moses-ini-to-v2.perl
+++ b/scripts/training/convert-moses-ini-to-v2.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/training/convert-moses-ini-v2-to-v1.py b/scripts/training/convert-moses-ini-v2-to-v1.py
index 44f192efe..3ef7d7c0d 100755
--- a/scripts/training/convert-moses-ini-v2-to-v1.py
+++ b/scripts/training/convert-moses-ini-v2-to-v1.py
@@ -1,5 +1,8 @@
 #! /usr/bin/env python
 # -*- coding: utf8 -*-
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 3 or, at your option, any later version.
 
 
 from __future__ import (
diff --git a/scripts/training/corpus-sizes.perl b/scripts/training/corpus-sizes.perl
index 30ae67ebb..1a6db669b 100755
--- a/scripts/training/corpus-sizes.perl
+++ b/scripts/training/corpus-sizes.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 # $Id: consolidate-training-data.perl 928 2009-09-02 02:58:01Z philipp $
 
diff --git a/scripts/training/create_count_tables.py b/scripts/training/create_count_tables.py
index 2288c034a..12499b1d7 100755
--- a/scripts/training/create_count_tables.py
+++ b/scripts/training/create_count_tables.py
@@ -1,6 +1,9 @@
 #!/usr/bin/env python
 # -*- coding: utf-8 -*-
 # Author: Rico Sennrich <sennrich [AT] cl.uzh.ch>
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 # This script creates tables that store phrase pair frequencies rather than
 # probabilities.
diff --git a/scripts/training/exodus.perl b/scripts/training/exodus.perl
index bb8616007..579056ff0 100755
--- a/scripts/training/exodus.perl
+++ b/scripts/training/exodus.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 # $Id$
 
diff --git a/scripts/training/filter-model-given-input.pl b/scripts/training/filter-model-given-input.pl
index e3a34c40b..a44d9c193 100755
--- a/scripts/training/filter-model-given-input.pl
+++ b/scripts/training/filter-model-given-input.pl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 # $Id$
 # Given a moses.ini file and an input text prepare minimized translation
diff --git a/scripts/training/filter-rule-table.py b/scripts/training/filter-rule-table.py
index 14736fe1f..d28fa0c89 100755
--- a/scripts/training/filter-rule-table.py
+++ b/scripts/training/filter-rule-table.py
@@ -1,25 +1,29 @@
 #!/usr/bin/env python
 
 # Author: Phil Williams
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
-# Usage: filter-rule-table.py [--min-non-initial-rule-count=N] INPUT
-#
-# Given a rule table (on stdin) and an input text, filter out rules that
-# couldn't be used in parsing the input and write the resulting rule table
-# to stdout.  The input text is assumed to contain the same factors as
-# the rule table and is assumed to be small (not more than a few thousand
-# sentences): the current algorithm won't scale well to large input sets.
-#
-# The filtering algorithm considers a source RHS to be a sequence of
-# words and gaps, which must match a sequence of words in one of the
-# input sentences, with at least one input word per gap.  The NT labels
-# are ignored, so for example a rule with the source RHS "the JJ dog"
-# would be allowed if the sequence "the slobbering dog" occurs in one of
-# the input sentences, even if there's no rule to derive a JJ from
-# "slobbering."  (If "slobbering" were an unknown word, the 'unknown-lhs'
-# decoder option would allow it to take a number of NT labels, likely
-# including JJ, with varying probabilities, so removing the rule would
-# be a bad idea.)
+"""Usage: filter-rule-table.py [--min-non-initial-rule-count=N] INPUT
+
+Given a rule table (on stdin) and an input text, filter out rules that
+couldn't be used in parsing the input and write the resulting rule table
+to stdout.  The input text is assumed to contain the same factors as
+the rule table and is assumed to be small (not more than a few thousand
+sentences): the current algorithm won't scale well to large input sets.
+
+The filtering algorithm considers a source RHS to be a sequence of
+words and gaps, which must match a sequence of words in one of the
+input sentences, with at least one input word per gap.  The NT labels
+are ignored, so for example a rule with the source RHS "the JJ dog"
+would be allowed if the sequence "the slobbering dog" occurs in one of
+the input sentences, even if there's no rule to derive a JJ from
+"slobbering."  (If "slobbering" were an unknown word, the 'unknown-lhs'
+decoder option would allow it to take a number of NT labels, likely
+including JJ, with varying probabilities, so removing the rule would
+be a bad idea.)
+"""
 
 import optparse
 import sys
diff --git a/scripts/training/flexibility_score.py b/scripts/training/flexibility_score.py
index 496184616..56d4f9425 100755
--- a/scripts/training/flexibility_score.py
+++ b/scripts/training/flexibility_score.py
@@ -2,6 +2,9 @@
 # -*- coding: utf-8 -*-
 
 # author: Rico Sennrich
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 """Add flexibility scores to a phrase table half.
 
diff --git a/scripts/training/giza2bal.pl b/scripts/training/giza2bal.pl
index 27ba9d659..ad9edb584 100755
--- a/scripts/training/giza2bal.pl
+++ b/scripts/training/giza2bal.pl
@@ -6,6 +6,9 @@
 #produced by giza containing the frequency of each traning sentence.
 
 #Copyright Marcello Federico, November 2004
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 #use warnings;
 
diff --git a/scripts/training/mert-moses.pl b/scripts/training/mert-moses.pl
index 92e1a79ff..c73e75a87 100755
--- a/scripts/training/mert-moses.pl
+++ b/scripts/training/mert-moses.pl
@@ -1,4 +1,8 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
+
 # $Id$
 # Usage:
 # mert-moses.pl <foreign> <english> <decoder-executable> <decoder-config>
diff --git a/scripts/training/postprocess-lopar.perl b/scripts/training/postprocess-lopar.perl
index 44be9c26c..05a56a3b5 100755
--- a/scripts/training/postprocess-lopar.perl
+++ b/scripts/training/postprocess-lopar.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 # $Id$
 
diff --git a/scripts/training/rdlm/average_null_embedding.py b/scripts/training/rdlm/average_null_embedding.py
index 28abc9508..899b402c1 100755
--- a/scripts/training/rdlm/average_null_embedding.py
+++ b/scripts/training/rdlm/average_null_embedding.py
@@ -1,6 +1,9 @@
 #!/usr/bin/env python
 # -*- coding: utf-8 -*-
 # Author: Rico Sennrich
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 """Average embeddings of special null words for RDLM.
 
diff --git a/scripts/training/rdlm/extract_syntactic_ngrams.py b/scripts/training/rdlm/extract_syntactic_ngrams.py
index 1292e90f2..be4ed2335 100755
--- a/scripts/training/rdlm/extract_syntactic_ngrams.py
+++ b/scripts/training/rdlm/extract_syntactic_ngrams.py
@@ -1,6 +1,9 @@
 #!/usr/bin/env python
 # -*- coding: utf-8 -*-
 # Author: Rico Sennrich
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 """
 Extract syntactic n-grams from dependency treebank in Moses XML format for
diff --git a/scripts/training/rdlm/extract_vocab.py b/scripts/training/rdlm/extract_vocab.py
index ed9266fd9..48e5215c3 100755
--- a/scripts/training/rdlm/extract_vocab.py
+++ b/scripts/training/rdlm/extract_vocab.py
@@ -1,6 +1,9 @@
 #!/usr/bin/env python
 # -*- coding: utf-8 -*-
 # Author: Rico Sennrich
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 # extract 5 vocabulary files from parsed corpus in moses XML format
 
diff --git a/scripts/training/rdlm/train_rdlm.py b/scripts/training/rdlm/train_rdlm.py
index 639c1b32c..289ab405c 100755
--- a/scripts/training/rdlm/train_rdlm.py
+++ b/scripts/training/rdlm/train_rdlm.py
@@ -1,5 +1,8 @@
 #!/usr/bin/env python
 # -*- coding: utf-8 -*-
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 from __future__ import print_function, unicode_literals
 
@@ -91,11 +94,14 @@ parser.add_argument(
     "--output-words-file", dest="output_words_file", metavar="PATH",
     help="Output vocabulary (default: %(default)s).")
 parser.add_argument(
-    "--input_vocab_size", dest="input_vocab_size", type=int, metavar="INT",
+    "--input-vocab-size", dest="input_vocab_size", type=int, metavar="INT",
     help="Input vocabulary size (default: %(default)s).")
 parser.add_argument(
     "--output-vocab-size", dest="output_vocab_size", type=int, metavar="INT",
     help="Output vocabulary size (default: %(default)s).")
+parser.add_argument(
+    "--mmap", dest="mmap", action="store_true",
+    help="Use memory-mapped file (for lower memory consumption).")
 
 
 parser.set_defaults(
@@ -192,11 +198,14 @@ def main(options):
             "extracting vocabulary from training text.\n")
         prepare_vocabulary(options)
 
+    numberized_file = os.path.basename(options.corpus_stem) + '.numberized'
+    train_file = numberized_file
+    if options.mmap:
+        train_file += '.mmap'
+
     extract_options = extract_syntactic_ngrams.create_parser().parse_args([
         '--input', options.corpus_stem,
-        '--output', os.path.join(
-            options.working_dir,
-            os.path.basename(options.corpus_stem) + '.numberized'),
+        '--output', os.path.join(options.working_dir, numberized_file),
         '--vocab', options.input_words_file,
         '--output_vocab', options.output_words_file,
         '--right_context', str(options.right_context_size),
@@ -219,6 +228,23 @@ def main(options):
     else:
         options.validation_file = None
 
+    if options.mmap:
+        try:
+            os.remove(os.path.join(options.working_dir, train_file))
+        except OSError:
+            pass
+        mmap_cmd = [os.path.join(options.nplm_home, 'src', 'createMmap'),
+                    '--input_file',
+                    os.path.join(options.working_dir, numberized_file),
+                    '--output_file',
+                    os.path.join(options.working_dir, train_file)
+                    ]
+        sys.stderr.write('creating memory-mapped file\n')
+        sys.stderr.write('executing: ' + ', '.join(mmap_cmd) + '\n')
+        ret = subprocess.call(mmap_cmd)
+        if ret:
+            raise Exception("creating memory-mapped file failed")
+
     sys.stderr.write('training neural network\n')
     train_nplm.main(options)
 
@@ -231,7 +257,7 @@ def main(options):
             options.output_model + '.model.nplm.' + str(options.epochs)),
         os.path.join(
             options.working_dir,
-            os.path.basename(options.corpus_stem) + '.numberized'),
+            numberized_file),
         os.path.join(options.output_dir, options.output_model + '.model.nplm')
         ])
     if ret:
diff --git a/scripts/training/reduce-factors.perl b/scripts/training/reduce-factors.perl
index 09f9c7f2b..82aed4355 100755
--- a/scripts/training/reduce-factors.perl
+++ b/scripts/training/reduce-factors.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/training/reduce-topt-count.pl b/scripts/training/reduce-topt-count.pl
index f760051c4..85ce0d6d9 100755
--- a/scripts/training/reduce-topt-count.pl
+++ b/scripts/training/reduce-topt-count.pl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 # given a moses.ini, filter the phrase tables to contain
 # only ttable-limit options per source phrase
diff --git a/scripts/training/reduce_combine.pl b/scripts/training/reduce_combine.pl
index a7614f73e..2055bed5b 100755
--- a/scripts/training/reduce_combine.pl
+++ b/scripts/training/reduce_combine.pl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 # $Id$
 # given a pathname to a factored corpus, a list of (numeric) factors to keep
diff --git a/scripts/training/remove-orphan-phrase-pairs-from-reordering-table.perl b/scripts/training/remove-orphan-phrase-pairs-from-reordering-table.perl
index eda529393..25c5cc028 100755
--- a/scripts/training/remove-orphan-phrase-pairs-from-reordering-table.perl
+++ b/scripts/training/remove-orphan-phrase-pairs-from-reordering-table.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/training/threshold-filter.perl b/scripts/training/threshold-filter.perl
index 3e42ca795..0aed67d25 100755
--- a/scripts/training/threshold-filter.perl
+++ b/scripts/training/threshold-filter.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/training/train-global-lexicon-model.perl b/scripts/training/train-global-lexicon-model.perl
index d3c55789d..528bfbd72 100755
--- a/scripts/training/train-global-lexicon-model.perl
+++ b/scripts/training/train-global-lexicon-model.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/training/train-model.perl b/scripts/training/train-model.perl
index 5a304c2f9..b693d774d 100755
--- a/scripts/training/train-model.perl
+++ b/scripts/training/train-model.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/training/train-neurallm.py b/scripts/training/train-neurallm.py
index 2d2f12015..ae77a42af 100755
--- a/scripts/training/train-neurallm.py
+++ b/scripts/training/train-neurallm.py
@@ -1,8 +1,12 @@
 #!/usr/bin/env python
 # -*- coding: utf-8 -*-
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
-""" train feed-forward neural network LM with NPLM tool
-resulting model can be used in Moses as feature function NeuralLM
+"""Train feed-forward neural network LM with NPLM tool.
+
+The resulting model can be used in Moses as feature function NeuralLM.
 """
 
 from __future__ import print_function, unicode_literals
@@ -83,6 +87,9 @@ parser.add_argument(
 parser.add_argument(
     "--vocab-size", dest="vocab_size", type=int, metavar="INT",
     help="Vocabulary size (default: %(default)s).")
+parser.add_argument(
+    "--mmap", dest="mmap", action="store_true",
+    help="Use memory-mapped file (for lower memory consumption).")
 
 parser.set_defaults(
     working_dir="working",
@@ -117,20 +124,43 @@ def main(options):
         if not os.path.exists(options.output_dir):
             os.makedirs(options.output_dir)
 
+    numberized_file = os.path.basename(options.corpus_stem) + '.numberized'
+    train_file = numberized_file
+    if options.mmap:
+        train_file += '.mmap'
+
     extraction_cmd = [os.path.join(options.nplm_home, 'src', 'prepareNeuralLM'),
                       '--train_text', options.corpus_stem,
                       '--ngramize', '1',
                       '--ngram_size', str(options.ngram_size),
                       '--vocab_size', str(options.vocab_size),
                       '--write_words_file', os.path.join(options.working_dir, options.words_file),
-                      '--train_file', os.path.join(options.working_dir, os.path.basename(options.corpus_stem) + '.numberized')
+                      '--train_file', os.path.join(options.working_dir, numberized_file)
                       ]
 
     sys.stderr.write('extracting n-grams\n')
+    sys.stderr.write('executing: ' + ', '.join(extraction_cmd) + '\n')
     ret = subprocess.call(extraction_cmd)
     if ret:
         raise Exception("preparing neural LM failed")
-    
+
+    if options.mmap:
+        try:
+            os.remove(os.path.join(options.working_dir, train_file))
+        except OSError:
+            pass
+        mmap_cmd = [os.path.join(options.nplm_home, 'src', 'createMmap'),
+                    '--input_file',
+                    os.path.join(options.working_dir, numberized_file),
+                    '--output_file',
+                    os.path.join(options.working_dir, train_file)
+                    ]
+        sys.stderr.write('creating memory-mapped file\n')
+        sys.stderr.write('executing: ' + ', '.join(mmap_cmd) + '\n')
+        ret = subprocess.call(mmap_cmd)
+        if ret:
+            raise Exception("creating memory-mapped file failed")
+
     if options.validation_corpus:
 
         extraction_cmd = [os.path.join(options.nplm_home, 'src', 'prepareNeuralLM'),
@@ -143,6 +173,7 @@ def main(options):
                           ]
 
         sys.stderr.write('extracting n-grams (validation file)\n')
+        sys.stderr.write('executing: ' + ', '.join(extraction_cmd) + '\n')
         ret = subprocess.call(extraction_cmd)
         if ret:
             raise Exception("preparing neural LM failed")
@@ -162,7 +193,7 @@ def main(options):
     average_options = averageNullEmbedding.parser.parse_args(
         ['-i', os.path.join(options.output_dir, options.output_model + '.model.nplm.' + str(options.epochs)),
          '-o', os.path.join(options.output_dir, options.output_model + '.model.nplm'),
-         '-t', os.path.join(options.working_dir, os.path.basename(options.corpus_stem) + '.numberized'),
+         '-t', os.path.join(options.working_dir, numberized_file),
          '-p', os.path.join(options.nplm_home, 'python')])
     averageNullEmbedding.main(average_options)
 
diff --git a/scripts/training/wrappers/adam-suffix-array/suffix-array-create.sh b/scripts/training/wrappers/adam-suffix-array/suffix-array-create.sh
index 238a53349..5db5e9aa9 100755
--- a/scripts/training/wrappers/adam-suffix-array/suffix-array-create.sh
+++ b/scripts/training/wrappers/adam-suffix-array/suffix-array-create.sh
@@ -1,4 +1,7 @@
 #!/bin/bash
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 # execute: ~/workspace/bin/moses-smt/scripts/training/wrappers/suffix-array-create.sh $SA_EXEC_DIR $SOURCE_CORPUS $TARGET_CORPUS $ALIGNMENT $SA_OUTPUT
 
diff --git a/scripts/training/wrappers/adam-suffix-array/suffix-array-extract.sh b/scripts/training/wrappers/adam-suffix-array/suffix-array-extract.sh
index 8c255b1b6..128ccaa9e 100755
--- a/scripts/training/wrappers/adam-suffix-array/suffix-array-extract.sh
+++ b/scripts/training/wrappers/adam-suffix-array/suffix-array-extract.sh
@@ -1,4 +1,7 @@
 #!/bin/bash
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 # execute: ~/workspace/bin/moses-smt/scripts/training/wrappers/adam-suffix-array/suffix-array-extract.sh $SA_EXEC_DIR $MODEL_DIR $INPUT_FILE $OUTPUT_DIR
 
diff --git a/scripts/training/wrappers/berkeleyparsed2mosesxml.perl b/scripts/training/wrappers/berkeleyparsed2mosesxml.perl
index 232cfefab..9c376200c 100755
--- a/scripts/training/wrappers/berkeleyparsed2mosesxml.perl
+++ b/scripts/training/wrappers/berkeleyparsed2mosesxml.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/training/wrappers/berkeleyparsed2mosesxml_PTB.perl b/scripts/training/wrappers/berkeleyparsed2mosesxml_PTB.perl
index 9e8c30d42..b8ba146c9 100755
--- a/scripts/training/wrappers/berkeleyparsed2mosesxml_PTB.perl
+++ b/scripts/training/wrappers/berkeleyparsed2mosesxml_PTB.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/training/wrappers/conll2mosesxml.py b/scripts/training/wrappers/conll2mosesxml.py
index 761037488..6473166d9 100755
--- a/scripts/training/wrappers/conll2mosesxml.py
+++ b/scripts/training/wrappers/conll2mosesxml.py
@@ -1,6 +1,9 @@
 #!/usr/bin/env python
 # -*- coding: utf-8 -*-
 # Author: Rico Sennrich
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 """
 Takes a file in the CoNLL dependency format (from the CoNLL-X shared task on
diff --git a/scripts/training/wrappers/filter-excluded-lines.perl b/scripts/training/wrappers/filter-excluded-lines.perl
index dff104dba..508ab8a06 100755
--- a/scripts/training/wrappers/filter-excluded-lines.perl
+++ b/scripts/training/wrappers/filter-excluded-lines.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/training/wrappers/find-unparseable.perl b/scripts/training/wrappers/find-unparseable.perl
index 00009e2e9..fd0664f1d 100755
--- a/scripts/training/wrappers/find-unparseable.perl
+++ b/scripts/training/wrappers/find-unparseable.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/training/wrappers/mada-wrapper.perl b/scripts/training/wrappers/mada-wrapper.perl
index f2cf14f40..d4124e34c 100755
--- a/scripts/training/wrappers/mada-wrapper.perl
+++ b/scripts/training/wrappers/mada-wrapper.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/training/wrappers/madamira-tok.perl b/scripts/training/wrappers/madamira-tok.perl
index 37e70079e..e9f19d53a 100755
--- a/scripts/training/wrappers/madamira-tok.perl
+++ b/scripts/training/wrappers/madamira-tok.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/training/wrappers/madamira-wrapper.perl b/scripts/training/wrappers/madamira-wrapper.perl
index 6535b6187..05ec44d7d 100755
--- a/scripts/training/wrappers/madamira-wrapper.perl
+++ b/scripts/training/wrappers/madamira-wrapper.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/training/wrappers/make-factor-brown-cluster-mkcls.perl b/scripts/training/wrappers/make-factor-brown-cluster-mkcls.perl
index 1e3a1ce3f..a8ce5f24e 100755
--- a/scripts/training/wrappers/make-factor-brown-cluster-mkcls.perl
+++ b/scripts/training/wrappers/make-factor-brown-cluster-mkcls.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/training/wrappers/make-factor-de-lemma.perl b/scripts/training/wrappers/make-factor-de-lemma.perl
index db978317e..0b93002a9 100755
--- a/scripts/training/wrappers/make-factor-de-lemma.perl
+++ b/scripts/training/wrappers/make-factor-de-lemma.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/perl -w
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use strict;
 use Encode;
diff --git a/scripts/training/wrappers/make-factor-de-morph.perl b/scripts/training/wrappers/make-factor-de-morph.perl
index 366a5a76d..d09196745 100755
--- a/scripts/training/wrappers/make-factor-de-morph.perl
+++ b/scripts/training/wrappers/make-factor-de-morph.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/training/wrappers/make-factor-de-pos.perl b/scripts/training/wrappers/make-factor-de-pos.perl
index 495517352..585323bd4 100755
--- a/scripts/training/wrappers/make-factor-de-pos.perl
+++ b/scripts/training/wrappers/make-factor-de-pos.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/training/wrappers/make-factor-en-porter.perl b/scripts/training/wrappers/make-factor-en-porter.perl
index 749dc1318..7ae5fd0b3 100755
--- a/scripts/training/wrappers/make-factor-en-porter.perl
+++ b/scripts/training/wrappers/make-factor-en-porter.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/perl -w
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use strict;
 use FindBin qw($RealBin);
diff --git a/scripts/training/wrappers/make-factor-en-pos.mxpost.perl b/scripts/training/wrappers/make-factor-en-pos.mxpost.perl
index 4aa66bac6..2bff8e329 100755
--- a/scripts/training/wrappers/make-factor-en-pos.mxpost.perl
+++ b/scripts/training/wrappers/make-factor-en-pos.mxpost.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/training/wrappers/make-factor-pos.tree-tagger.perl b/scripts/training/wrappers/make-factor-pos.tree-tagger.perl
index 0ad04d4de..1e8ccd0ee 100755
--- a/scripts/training/wrappers/make-factor-pos.tree-tagger.perl
+++ b/scripts/training/wrappers/make-factor-pos.tree-tagger.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/training/wrappers/make-factor-stem.perl b/scripts/training/wrappers/make-factor-stem.perl
index 662f1d882..9bde7648f 100755
--- a/scripts/training/wrappers/make-factor-stem.perl
+++ b/scripts/training/wrappers/make-factor-stem.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/training/wrappers/make-factor-suffix.perl b/scripts/training/wrappers/make-factor-suffix.perl
index 6a59254e4..015df3874 100755
--- a/scripts/training/wrappers/make-factor-suffix.perl
+++ b/scripts/training/wrappers/make-factor-suffix.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/training/wrappers/morfessor-wrapper.perl b/scripts/training/wrappers/morfessor-wrapper.perl
index c65a2cebc..0269045a0 100755
--- a/scripts/training/wrappers/morfessor-wrapper.perl
+++ b/scripts/training/wrappers/morfessor-wrapper.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/training/wrappers/mosesxml2berkeleyparsed.perl b/scripts/training/wrappers/mosesxml2berkeleyparsed.perl
index e929658ff..02bc7b88e 100755
--- a/scripts/training/wrappers/mosesxml2berkeleyparsed.perl
+++ b/scripts/training/wrappers/mosesxml2berkeleyparsed.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/training/wrappers/mosesxml2brackets.py b/scripts/training/wrappers/mosesxml2brackets.py
index 6ff1d20c9..6b90aa256 100755
--- a/scripts/training/wrappers/mosesxml2brackets.py
+++ b/scripts/training/wrappers/mosesxml2brackets.py
@@ -1,8 +1,11 @@
 #!/usr/bin/env python
 # -*- coding: utf-8 -*-
 # Author: Rico Sennrich
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
-# convert trees in moses XML format to PTB-style bracketed format
+"""Convert trees in moses XML format to PTB-style bracketed format."""
 
 from __future__ import print_function, unicode_literals
 import sys
diff --git a/scripts/training/wrappers/parse-de-berkeley.perl b/scripts/training/wrappers/parse-de-berkeley.perl
index 596fb3eff..f605a37ae 100755
--- a/scripts/training/wrappers/parse-de-berkeley.perl
+++ b/scripts/training/wrappers/parse-de-berkeley.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/training/wrappers/parse-de-bitpar.perl b/scripts/training/wrappers/parse-de-bitpar.perl
index 1bbcf5329..0d5346058 100755
--- a/scripts/training/wrappers/parse-de-bitpar.perl
+++ b/scripts/training/wrappers/parse-de-bitpar.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/training/wrappers/parse-en-collins.perl b/scripts/training/wrappers/parse-en-collins.perl
index 252d3d2b7..c9a960912 100755
--- a/scripts/training/wrappers/parse-en-collins.perl
+++ b/scripts/training/wrappers/parse-en-collins.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/training/wrappers/parse-en-egret.perl b/scripts/training/wrappers/parse-en-egret.perl
index 9f434063b..e97bc1ae0 100755
--- a/scripts/training/wrappers/parse-en-egret.perl
+++ b/scripts/training/wrappers/parse-en-egret.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/training/wrappers/parse-en-senna.perl b/scripts/training/wrappers/parse-en-senna.perl
index f271633ea..2df46284b 100755
--- a/scripts/training/wrappers/parse-en-senna.perl
+++ b/scripts/training/wrappers/parse-en-senna.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl 
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use strict;
 use warnings;
diff --git a/scripts/training/wrappers/parse-en-stanford.py b/scripts/training/wrappers/parse-en-stanford.py
index 7d8be4bcf..06b027e55 100755
--- a/scripts/training/wrappers/parse-en-stanford.py
+++ b/scripts/training/wrappers/parse-en-stanford.py
@@ -1,11 +1,17 @@
 #!/usr/bin/python
 # -*- coding: utf-8 -*-
 # Author: Rico Sennrich
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
-# (hacky) wrapper around Stanford CoreNLP to produce CoNLL dependency format.
-# assumes tokenized and sentence-split text.
+"""
+(Hacky) wrapper around Stanford CoreNLP to produce CoNLL dependency format.
+Assumes tokenized and sentence-split text.
 
-# to get Moses XML format, first projectivize the trees, then use conll2mosesxml.py.
+To get Moses XML format, first projectivize the trees, then use
+conll2mosesxml.py.
+"""
 
 from __future__ import print_function, unicode_literals
 import os
diff --git a/scripts/training/wrappers/senna2brackets.py b/scripts/training/wrappers/senna2brackets.py
index 4fc71ed44..a81100277 100755
--- a/scripts/training/wrappers/senna2brackets.py
+++ b/scripts/training/wrappers/senna2brackets.py
@@ -1,19 +1,24 @@
 #!/usr/bin/env python
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
-# Read SENNA output (from stdin), extract the parse trees, and write them in
-# PTB-style bracketed format (to stdout).
-#
-# The SENNA output is assumed to contain tokens in the first column, POS tags
-# in the second column, and PSG fragments in the final column.
-#
-# It is also assumed that SENNA was run through the parse-en-senna.perl wrapper,
-# which:
-#
-#   - Substitutes the special "SENTENCE_TOO_LONG" token for sentences that
-#     exceed SENNA's hardcoded limit.
-#
-#   - Replaces the bracket-like tokens "-LRB-", "-RRB-", etc. with "(", ")",
-#     etc.
+"""
+Read SENNA output (from stdin), extract the parse trees, and write them in
+PTB-style bracketed format (to stdout).
+
+The SENNA output is assumed to contain tokens in the first column, POS tags
+in the second column, and PSG fragments in the final column.
+
+It is also assumed that SENNA was run through the parse-en-senna.perl wrapper,
+which:
+
+  - Substitutes the special "SENTENCE_TOO_LONG" token for sentences that
+    exceed SENNA's hardcoded limit.
+
+  - Replaces the bracket-like tokens "-LRB-", "-RRB-", etc. with "(", ")",
+    etc.
+"""
 
 import optparse
 import os
diff --git a/scripts/training/wrappers/syntax-hyphen-splitting.perl b/scripts/training/wrappers/syntax-hyphen-splitting.perl
index 653b410d0..1a260df10 100755
--- a/scripts/training/wrappers/syntax-hyphen-splitting.perl
+++ b/scripts/training/wrappers/syntax-hyphen-splitting.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;
diff --git a/scripts/training/wrappers/tagger-german-chunk.perl b/scripts/training/wrappers/tagger-german-chunk.perl
index c57031889..0b707a579 100755
--- a/scripts/training/wrappers/tagger-german-chunk.perl
+++ b/scripts/training/wrappers/tagger-german-chunk.perl
@@ -1,4 +1,7 @@
 #!/usr/bin/env perl
+#
+# This file is part of moses.  Its use is licensed under the GNU Lesser General
+# Public License version 2.1 or, at your option, any later version.
 
 use warnings;
 use strict;