WebSVN - public - Diff - Rev 47 and 55

 #define _STRTOK_H_
 #include <string>
+typedef std::string::size_type SIZE;
 /// Simple string tokenizer class.
 /**
  * This class implements a string tokenizer. It splits a string
  * by a character to a number of elements (tokens) which are
  * provided sequentially.
+ *
- * All operations are made on the original string itself.
+ * All operations are made on a copy of the original string
- * The implementation is not ready to handle any changes of the
+ * (which may be in fact a copy-on-write instance).
- * string.
+ *
  * The original string is left unchanged. All tokens are returned
  * as newly created strings.
+ *
+ * There is possibility to specify a prefix character which
+ * causes the consecutive character is not considered as
+ * a delimiter. If you don't specify this character (or specify
+ * the NUL character, 0x00) this feature is disabled. The mostly
+ * used prefix is a backslash ('\').
+ *
+ * This class is not thread-safe.
+ *
+ * Performance note: This class is currently not intended
+ * to be very fast. Speed optimizations will be done later.
  */
 class StringTokenizer
+{
 public:
   /// Constructor.
   /**
    * Creates a ready-to-use tokenizer.
+   *
    * \param[in] rStr string for tokenizing
    * \param[in] cDelim delimiter (separator) character
+   * \param[in] cPrefix character which is prepended if a
+   *            character must not separate tokens
    */
-  StringTokenizer(const std::string& rStr, char cDelim = ',');
+  StringTokenizer(const std::string& rStr, char cDelim = ',', char cPrefix = '\0');
   /// Destructor.
   ~StringTokenizer() {}
   /// Checks whether the tokenizer can provide more tokens.
     return m_pos < m_len;
+  }
   /// Returns the next token.
   /**
+   * If a prefix is defined it is stripped from the returned
+   * string (e.g. 'abc\ def' is transformed to 'abc def'
+   * while the prefix is '\').
+   *
+   * \param[in] fSkipEmpty skip empty strings (more consecutive delimiters)
+   * \return next token or "" if no more tokens available
+   *
+   * \sa GetNextTokenRaw()
+   */
+  std::string GetNextToken(bool fSkipEmpty = false);
+  /// Returns the next token.
+  /**
+   * This method always returns an unmodified string even
+   * if it contains prefix characters.
+   *
+   * \param[in] fSkipEmpty skip empty strings (more consecutive delimiters)
    * \return next token or "" if no more tokens available
+   *
+   * \sa GetNextToken()
+   */
+  std::string GetNextTokenRaw(bool fSkipEmpty = false);
+  /// Returns the remainder of the source string.
+  /**
+   * This method returns everything what has not been
+   * processed (tokenized) yet and moves the current
+   * position to the end of the string.
+   *
+   * If a prefix is defined it is stripped from
+   * the returned string.
+   *
+   * \return remainder string
    */
-  std::string GetNextToken();
+  std::string GetRemainder();
   /// Sets a delimiter (separator) character.
   /**
    * The new delimiter has effect only to tokens returned later;
    * the position in the string is not affected.
+   *
+   * If you specify a NUL character (0x00) here the prefix
+   * will not be used.
+   *
    * \param[in] cDelim delimiter character
    */
   inline void SetDelimiter(char cDelim)
+  {
     m_cDelim = cDelim;
   inline char GetDelimiter() const
+  {
     return m_cDelim;
+  }
+  /// Sets a prefix character.
+  /**
+   * The new prefix has effect only to tokens returned later;
+   * the position in the string is not affected.
+   *
+   * \param[in] cPrefix prefix character
+   *
+   * \sa SetNoPrefix()
+   */
+  inline void SetPrefix(char cPrefix)
+  {
+    m_cPrefix = cPrefix;
+  }
+  /// Returns the prefix character.
+  /**
+   * \return prefix character
+   */
+  inline char GetPrefix() const
+  {
+    return m_cPrefix;
+  }
+  /// Sets the prefix to 'no prefix'.
+  /**
+   * Calling this method is equivalent to SetPrefix((char) 0).
+   *
+   * \sa SetPrefix()
+   */
+  inline void SetNoPrefix()
+  {
+    SetPrefix('\0');
+  }
   /// Resets the tokenizer.
   /**
    * Re-initializes tokenizing to the start of the string.
    */
   inline void Reset()
+  }
 private:
   std::string m_str;            ///< tokenized string
   char m_cDelim;                ///< delimiter character
+  char m_cPrefix;               ///< prefix character
   std::string::size_type m_pos; ///< current position
   std::string::size_type m_len; ///< string length
+  /// Strips all prefix characters.
+  /**
+   * \param[in] s source string
+   * \param[in] cnt string length
+   * \return modified string
+   */
+  std::string StripPrefix(const char* s, SIZE cnt);
+  /// Extracts the next token (internal method).
+  /**
+   * The extracted token may be empty.
+   *
+   * \param[out] rToken extracted token
+   * \param[in] fStripPrefix strip prefix characters yes/no
+   */
+  void _GetNextToken(std::string& rToken, bool fStripPrefix);
+  /// Extracts the next token (internal method).
+  /**
+   * This method does no checking about the prefix character.
+   *
+   * The extracted token may be empty.
+   *
+   * \param[out] rToken extracted token
+   */
+  void _GetNextTokenNoPrefix(std::string& rToken);
+  /// Extracts the next token (internal method).
+  /**
+   * This method does checking about the prefix character.
+   *
+   * The extracted token may be empty.
+   *
+   * \param[out] rToken extracted token
+   */
+  void _GetNextTokenWithPrefix(std::string& rToken);
 };
 #endif //_STRTOK_H_

Subversion Repositories public

(root)/incron/trunk/strtok.h @ 108 - Rev 47 → 55