CSCI 255 C vs Java

Small or expected differences for Java programmers

There are no classes

Print with the function printf, not the PrintStream method printf. Also, compute the square root with the function sqrt, not the static method Math.sqrt.

Include files

Almost all C programs will start with a standard set of includes from the C standard library something like the following. 

#include <stdio.h>
#include <stdlib.h>
#include <math.h>
#include <string.h>
#include <stdint.h>

These are similar to the import statement of Java but are not restricted to information about one class.

In general, whenever you use a standard C routine, you should add the appropriate include statements. You can find these by looking at the man page for the routine or by consulting the C standard.

The .h at the end of an include file in a one letter abbreviation for header. On Unix systems, the include files are usually stored in the /usr/include directory.

Scanning input

C has a scanf for reading input. C’s I/O functions use a format descriiptor inspired by FORTRAN’s FORMAT statement. The Java Formatter is an extension of scanf.

In the old days, students were introduced to scanf and printf, along with the mysterious & operator, which allows you to change the value of a parameter, in the first week of their introductory programming class. 

   printf("Enter two numbers to exponentiate\n") ;
   scanf("%f %f", &base, &power) ;
   printf("%f^^%f = %f\n", base, power, pow(base, power)) ;

There is no boolean

The initial versions of C do not have a boolean type. Integers are used represent true and false. The integer 0 is interpreted as false and any non-zero integer is interpreted as true in C control structures. Because C uses integers as booleans, The following code will print the position of the first element of an array containing zero. 

   while (i=0; A[i]; ++i) ;
   printf("The first zero element is at position %d\n", i) ;

In Java, you’d use the test A[i] != 0. For readability, C programmers should include the test.

Using integers as booleans means that the logical operators may be applied to integers, and the relational operators always yields either 0 or 1.

Sometimes that allows you to write succinct, but confusing code, which some C programmers will argue is very efficient. For example, you can count the number of positive elements in an array with the following loop. 

   for (i=0, posCount=0; i<n; posCount += A[i++] >= 0) ;

Of course, it is considered bad style to use the results of logical and relational operators as int’s.

Be careful with equality tests. The following if is legal C. 

   if (i=0) {
      zeroCount = zeroCount + 1 ;
   }

Beginning with C99, C programmers can include stdbool.h to define a bool type with values true and false.

There is no length operator for arrays

C arrays are not objects. Unless you are using pointers (more on that later), they are declared with a fixed length. 

    int X[1000], Y[1000] ;

When writing functions that arrays of varying sizes, it is necessary to use a parameter type in which the size of the array is omitted. and explicitly pass the length of the array. 

    int add_elements(int X[], int n) {
       int i ;
       int sum = 0 ;
       for (i=0; i<n; ++i) {
          sum = sum + X[i] ;
       }
       return sum ;
    }

There is no string type

In C, strings are implemented as null-terminated arrays of characters. Suppose you use the following statement to declare and initialize an array. 

   char[] course = "CSCI 320" ;

This statement actually allocates nine characters for the array. The value of course[7] is '0' and the value of course[8] is '\0', the null character.

Here’s an example of a C routine that counts the number of times the exclamation point appears in a string. 

   int shriek_count(char X[]) {
       int i ;
       int count = 0 ;
       for (i=0; X[i] != '\0'; ++i) {
          if (X[i] == '!') {
             count = count + 1 ;
          }
       }
       return count ;
   }

Arrays are declared differently in C

As you may have noticed, the following declaration is not allowed in C. 

   int[] X ;

The braces must follow the identifier name.

There are signed and unsigned numbers

In C, the integer types can be declared as signed or unsigned. Although unsigned variables are appropriate for some “real world” applications which you know only use positive numbers, you rarely see unsigned in C code.

There are no standard sizes for the number types

In Java, you know that an int is stored in a 32-bit twos-complement number. Consequently, you know that the largest int is 2147483647 and the smallest int is -2147483648.

In C, you know that an int is at least as big as a short and that a int is no longer than a long. Starting with the C99 standard, the include file stdint.h allows the use of types, such as int32_t and uint32_t, that specify the size of integers. This are useful in programming small microcontrollers where, by default, integers are usually 16 bits wide. (The Arduino int is 16 bits.)

Operators

C and Java have the same set of operators with a few exceptions.

Old style declarations

Before C99, C only allowed variables of a function to be declared at the beginning of a code block. If you are accustomed to writing C++/Java-style loop declarations, you must compile your program with C99 enabled.

Large or unexpected differences for Java programmers

Here are some notes I wrote for the Fall 2011 offering of ECE 209 that illustrate some of the more significant differences between C and Java. We spent six weeks on this in ECE 209, but we will speed through in this course.