Week 13 Notes for CS 1500, Section 002 -- Spring 2001

• Check schedule
• Class starts at 11:15. I take roll at 11:25
• Announcement(s)
• Discuss file stream information
  • "ios" is the base class for stream I/O operations.
  • "istream" and "ostream" are subclasses of "ios."
    • istream is specifically for input. "cin" is an instance.
    • ostream is specifically for output. "cout" is an instance.
  • "ifstream" is a subclass of istream, tailored for FILE input.
  • "ofstream" is a subclass of ostream, tailored for FILE output.
  • In order to use ifstream or ofstream objects you need this directive:

     #include <fstream.h>

  • These kinds of statements declare and open a file stream:

    ifstream infile("raw.dat") ;
    ofstream outfile("process.dat") ;

  • Declaration and open can be done separately:

    ifstream infile ;
    infile.open ("raw.dat") ;
    
    ofstream outfile ;
    outfile.open("process.dat") ;

  • the value of infile or outfile will be ZERO if the open operation fails, else non-ZERO. One can use this fact to test for failure of the file-open operation.
  • we can use extraction with an ifstream:

    infile >> num ;

  • we can use insertion with an ofstream:

    outfile << num ;

  • An expression such as (infile >> num) or (outfile << num) has a value of ZERO if the operation fails, else non-ZERO. One can use this fact to test for the failure of the operation.
  • One can read individual characters from an ifstream like this:

    char ch ;
    infile >> ch ;
    infile.get(ch) ;

    The first method skips white space, the second does not. infile.get() also returns a value of ZERO or non-ZERO.
  • Formatting works the same with the ifstream and ofstream classes as it does with cin and cout.
  • Closing a file:

    infile.close() ;
    outfile.close() ;

  • After closing, a stream can be re-used to open a different file (or the same one again).

    infile.open("raw2.dat")

• Check schedule
• Class starts at 11:15. I take roll at 11:25
• Announcement(s)
  • The schedule now states what material will be covered by quiz #3
  • I posted study guidance for quiz #3
• Discuss Reference Parameter Information
  • Previously we have used parameters that are passed by value -- a copy of the parameter goes to the function -- not the original. The original is therefore unchanged by anything the function does to the copy.
  • One can also pass a parameter to a function by reference (but only if the parameter is a variable, not a constant or an expression). In this mode, the address of the variable is what the function actually receives. When the function examines or changes the parameter, it goes to the memory location indicated by the address and reads from or writes to that location in the memory. That location is in the memory space accessible to the caller. Therefore if function X calls function Y and passes a parameter z to Y by reference, Y can change the value of z in such a way that the change is visible to X.
  • We can also use pass-by-reference to save the time and space used up to copy an argument.
  • To pass a ifstream or ofstream by value would mean copying what might be a large file. This is not practical.
• Discuss Array Information
  • Arrays make it easy for a program to work with a large collection of items that are all of the same data type -- for example the prices of all the items in a store can be an array of 1000 floats. Here is how that declaration would look:

    float price[1000] ;

    This tells the compiler to reserve a contiguous section of the memory, enough for 1000 float variables, and to associate the name "price" with the entire section.
  • The array has a single name. The programmer does not have to declare and name 1000 different float variables.
  • The slots in the array are numbered from 0 to some upper limit. In the example above, the slots are floating point variables numbered 0 to 999 inclusive. (There are 1000 slots, but because 0 is the lowest, 999 is the highest. There is no slot 1000.)
  • One of the nicer things about arrays is that the index can be a variable or expression. For example the statement

    for (int i=0; i<1000; i++) price[i] = 0.0 ;

    initializes all 1000 of the slots to the value 0.0 through the use of just one for-loop.
  • Also the index to an array can be an input to a program. If there is a line of input like:

    53  6.79

    and if the program statement is:

      cin >> i >> price[i] ;

    then this serves to change price[53] to the value 6.79. If we want our program to make a long series of price changes, we can use C++ code like:

      cout << "Give an index and float value: " ;
      while (cin >> i >> price[i])
      {
        cout << i << "   " << price[i] << endl ;
        cout << "Give an index and float value: " ;
      }

    Then all we have to do is input a series like this to get all the price changes made that we desire:

       42    9.99
      193   18.54
       53   11.98
       17    5.79
      438    0.79 

  • Check out the sample program ex1003.cpp to see how arrays processing may be done in a program, and how arrays are passed as parameters to functions.

• Check schedule
• Class starts at 11:15. I take roll at 11:25
• Announcement(s)
  • If you want to take the make-up final, you must e-mail me and let me know. Otherwise, I won't make a copy of the test for you.
  • I decided to give you some additional help with Solo Problem #6. There is now a "shell" of a solution to the problem in the assignment directory. The name of the file is "batStats.cpp.shell." Use this shell to simplify your task if you wish. Read the file and follow the directions.
• We do lab today -- work on solo #6 if possible.