few small fixes of alignment tools, add new orthographic similarity feature for word aligner, final naming of directories, libraries in cdec

1 files changed, 40 insertions, 0 deletions

diff --git a/decoder/dict.h b/decoder/dict.h
new file mode 100644
index 00000000..bae9debe
--- /dev/null
+++ b/decoder/dict.h
@@ -0,0 +1,40 @@
+#ifndef DICT_H_
+#define DICT_H_
+
+#include <cassert>
+#include <cstring>
+#include <tr1/unordered_map>
+#include <string>
+#include <vector>
+
+#include <boost/functional/hash.hpp>
+
+#include "wordid.h"
+
+class Dict {
+ typedef std::tr1::unordered_map<std::string, WordID, boost::hash<std::string> > Map;
+ public:
+  Dict() : b0_("<bad0>") { words_.reserve(1000); }
+  inline int max() const { return words_.size(); }
+  inline WordID Convert(const std::string& word) {
+    Map::iterator i = d_.find(word);
+    if (i == d_.end()) {
+      words_.push_back(word);
+      d_[word] = words_.size();
+      return words_.size();
+    } else {
+      return i->second;
+    }
+  }
+  inline const std::string& Convert(const WordID& id) const {
+    if (id == 0) return b0_;
+    assert(id <= words_.size());
+    return words_[id-1];
+  }
+ private:
+  const std::string b0_;
+  std::vector<std::string> words_;
+  Map d_;
+};
+
+#endif