#ifndef TOK_H
#define TOK_H
/////////////////////////////////////////////////////////////// 
//  tok.h      -  read words from ASCII file                 //
//  ver 1.2                                                  //
//  Language:     Visual C, ver 6.0                          //
//  Platform:     Micron Dual Pentium Pro 200, Win NT 4.0    //
//  Application:  CSE784 Example                             //
//  Author:       Jim Fawcett, CST 2-187, Syracuse Univ.     //
//                (315) 443-3948, fawcett@ecs.syr.edu        //
///////////////////////////////////////////////////////////////
/*

    Module Operations
    =================
    This tok class provides one public function which extracts
    words from an existing file stream.  A word is a sequence of
    characters delimited by whitespace.  punctuator sequences are
    separated from alphanumberic characters; and certain punctuator
    characters are collected as single character tokens, e.g.:
    
        '{', '}', '(', ')', '[', ']', ';'

    The tokenizer is used in the following way:

      ifstream in(filename);
      token toker(in);
      string tok;
      while(getTok(tok)) {

          // do things with collected tokens

      }
      int nl = toker.lines();       // number of '\n' chars
      int sl = toker.scopeLevel();  // net number of '{' and '}' chars

    All newlines, '\n', are counted.  Braces, '{' and '}', are counted
    only if not in comments or quoted strings.

*/
//
/////////////////////////////////////////////////////////////// 
//  build process                                            //
///////////////////////////////////////////////////////////////
//                                                           //
//  Required Files:                                          //
//    tok.h, tok.c                                           //
//                                                           //
//  Compile Command:                                         //
//    cl -GX -DTEST_TOK tok.cpp                              //
//                                                           //
///////////////////////////////////////////////////////////////
/*
    
      Maintenance History
      ===================
      ver 1.2, 11 Sep 1999
      - added counts for lines and braces
      ver 1.1, 26 Apr 1999
      - fixed bug in test stub, e.g., added in.close() and
        in.clear() at end of file loop
      ver 1.0, 20 Sep 1998
      - first release
      
*/

/////////////////////////////////////////////////////////////// 
//  declarations                                             //
/////////////////////////////////////////////////////////////// 

#include <iostream>
#include <string>
using namespace std;

class token {

  public:
    token(istream &input);
    bool getTok(string &buffer);
    unsigned int lines();
    unsigned int scopeLevel();

  private:
    istream &in;                // source of tokens
    unsigned int _lines;        // number of newlines
    unsigned int _scopeLevel;   // net number of open/closed braces
    enum charType { Char, NewLine, White, Punct };
    charType type(char ch);
    int delim(char ch);
    int isComment(const string &buffer);
    int isWhite(const string &buffer);
    int eatWhite(string &buffer);
    int eatComment(string &buffer);
    int isQuote(const string &buffer);
    int collectQuote(string &buffer);
};

inline unsigned int token::lines() { return _lines; }

inline unsigned int token::scopeLevel() { return _scopeLevel; }

#endif