From 80686d4e567bae579ea39e009826a2de92cd4ace Mon Sep 17 00:00:00 2001 From: redpony Date: Wed, 11 Aug 2010 02:37:10 +0000 Subject: major refactor, break bad circular deps git-svn-id: https://ws10smt.googlecode.com/svn/trunk@509 ec762483-ff6d-05da-a07a-a48fb63a330f --- utils/dict.h | 66 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 66 insertions(+) create mode 100644 utils/dict.h (limited to 'utils/dict.h') diff --git a/utils/dict.h b/utils/dict.h new file mode 100644 index 00000000..348a97e3 --- /dev/null +++ b/utils/dict.h @@ -0,0 +1,66 @@ +#ifndef DICT_H_ +#define DICT_H_ + + +#include +#include + +#include +#include +#include "hash.h" +#include "wordid.h" + +class Dict { + typedef + HASH_MAP > Map; + public: + Dict() : b0_("") { + HASH_MAP_EMPTY(d_,""); + words_.reserve(1000); + } + + inline int max() const { return words_.size(); } + + inline WordID Convert(const std::string& word, bool frozen = false) { + Map::iterator i = d_.find(word); + if (i == d_.end()) { + if (frozen) + return 0; + words_.push_back(word); + d_[word] = words_.size(); + return words_.size(); + } else { + return i->second; + } + } + + inline WordID Convert(const std::vector& words, bool frozen = false) + { return Convert(toString(words), frozen); } + + static inline std::string toString(const std::vector& words) { + std::string word= ""; + for (std::vector::const_iterator it=words.begin(); + it != words.end(); ++it) { + if (it != words.begin()) word += " ||| "; + word += *it; + } + return word; + } + + inline const std::string& Convert(const WordID& id) const { + if (id == 0) return b0_; + assert(id <= (int)words_.size()); + return words_[id-1]; + } + + void AsVector(const WordID& id, std::vector* results) const; + + void clear() { words_.clear(); d_.clear(); } + + private: + const std::string b0_; + std::vector words_; + Map d_; +}; + +#endif -- cgit v1.2.3