PROGRAM STRUCTURE
Statements and declarations may be split between lines at any point except within keywords, names, and multi-character operators. That is, tokens (the smallest units of the syntax) must be wholly contained in a line. Another way to think of this is that C interprets line breaks, like blanks and tabs, as white space.
Overall Structure
As Listing 1-1 shows, C programs have a simple structure. A program is essentially a sequence of comments (line 1), preprocessor directives (lines 2-4), and global declarations (beginning with lines 5, 6, 7, 14, and 18). Notice that the procedural parts of a program--the functions--are considered to be declarations.
Comments are used to clarify the logic of programs and document their operation. They are of no use to the compiler, so the preprocessor squeezes them out of a program before the compiler sees it. Comments are delimited on the left by the /* sequence and on the right by the next */ sequence. There is no limit to the length of comments and they may continue over any number of lines. We can place them anywhere in a program, except within tokens; that is, they are legal anywhere white space is allowed.
The preprocessor is sensitive to the comment delimiters only when they do not occur within quotes or apostrophes. So, we may freely use these character sequences in strings and character constants without fear of confusing the compiler. (The idea of a two-character constant may seem strange, but C accepts them without complaint.)
While some C compilers support comment nesting as an option (it is not a standard feature of C), Small C will definitely be confused by comments within comments. This makes it difficult to comment out of programs sections of logic that are themselves commented. This lack of comment nesting is an unfortunate characteristic of C.
Preprocessor Directives
Although not a part of the C language proper, I included preprocessor directives in the overall structure of C programs (above) because they do actually appear in the source file together with the "pure" C code. We have already seen two examples of these directives in Listing 1-1 (lines 2-4); those and others will be covered in detail in Chapter 11.
However, a few general remarks are in order. First, all C compilers include a preprocessor which prepares the source code for assimilation by the compiler. Small C is unusual in that the preprocessor is integrated into the compiler itself as a low level input routine. This saves one pass of writing and reading the source code. It also means that when we probe the compiler with keyboard input, we may enter preprocessor directives along with regular C code.
The Small C preprocessor supports three capabilities--(1) the inclusion of source code from other files, (2) the definition of symbols that stand for arbitrary substitution text, and (3) the conditional compiling of optional program code.
Each preprocessor directive is written on a line by itself beginning with a # character. The preprocessor lines are eliminated from the program before it goes on the compiler.
Global Declarations
Objects that may be declared in Small C programs include:
Since there is much to be said about function declarations, they are treated more thoroughly below. But first we need to distinguish between two similar terms.
Declarations and Definitions
The terms declaration and definition may seem synonymous, but in the C language they imply different ideas. Declarations only tell the compiler what it needs to know about objects; whereas, definitions also reserve space in memory for them. We can declare something without defining it, but we cannot define it without declaring it.
An object may be said to exist in the file in which it is defined, since compiling the file yields a module containing the object. On the other hand, an object may be declared within a file in which it does not exist. Declarations of this type are preceded by the keyword extern. Thus,
int i;
defines an integer which exists in the present module;
whereas,
extern int ei;
only declares an integer to exist in another, separately
compiled, module. The linker brings these modules together and
connects inter-module references by name. The compiler knows
everything about extern objects except where they are.
The linker is responsible for resolving that discrepancy. The
compiler simply tells the assembler that the objects are in fact
external. And the assembler, in turn, makes this known to the
linker.Function Declarations
A function is simply a subroutine--a common piece of logic which is called from various points in a program, and returns control to the caller when its work is done. Other languages distinguish between functions and procedures (or subroutines). The former being invoked as a term, returning a value, in expressions and the latter being called directly by means of call statements. C eliminates the distinction by accepting a bare expression as a statement. Furthermore, since a function call may comprise an entire expression, functions can be called without the aid of a special statement.
While other block structured languages support the nesting of procedural declarations, C does not. In C all function declarations must occur at the global level--outside of other function declarations. This greatly simplifies the structure of C programs and makes them easier to understand.
A function declaration consists of two parts: a declarator and a body. The declarator states the name of the function and (for use within the function) the names of arguments passed to it. In our sample program, no arguments were passed to the functions that were called, so each of the three function declarators specified a null argument list (empty parentheses). The parentheses are required even when there are no arguments.
The body of a function consists of argument declarations followed by a statement that performs the work. The argument declarations indicate to the function the types of the arguments listed in the declarator. Each argument must be declared and only arguments may be declared at this point. It is the responsibility of the programmer to ensure that the correct types of arguments are passed to the function when it is called. Chapter 28 describes how modern C compilers assuming this responsibility and sugg ests how Small C can be modified to do the same.
The reference above to a statement (singular) may be surprising. After all, Listing 1-1 plainly shows several statements in each function. But the reference is correct since the collected statements following the argument declarations of a function comprise a single compound statement.
Compound Statements
A compound statement (or block) is a sequence of statements, enclosed by braces, that stands in place of a single statement. Simple and compound statements are completely interchangeable as far as the syntax of the C language is concerned. Therefore, the statements that comprise a compound statement may themselves be compound; that is, blocks can be nested. Thus, it is legal to write
{ i=5; { j=6; k=7; } }
in which j=6; and k=7; comprise a block which,
together with i=5;, comprises a larger block. Of course
this offers no advantage over simply writing the three statements
in sequence. As we shall see, however, C does not provide for
sequences of statements except in blocks. So if a sequence of
statements is to be controlled by some condition, we would need
to write something like
if (i < 5) { i=5; j=6; k=7; }
instead of
if (i < 5) i=5; j=6; k=7;
In the latter case, only i=5 is controlled by the
condition I<5. The other assignments happen regardless
of the value of i.Flow of Control
When a function receives control, execution begins with its first statement. If control reaches the end of the body--the closing brace--it then returns to the point from which the function was called and continues onward from there.
C programs always begin execution with an ordinary call to a function named main(). Consequently, there must be a main() function somewhere in the program. A return from that function (for that call) transfers control back to the operating system. Nothing prevents a program from calling main() itself, however. When that occurs, the return from main() transfers control back to the point of the call. Control returns to the operating system from main() only when the in stance of main() called by the operating system returns control.
Global Variables
Variables declared outside of a function, like ch and where in Listing 1-1, are properly called external variables because they are defined outside of any function. While this is the standard term for these variables, it is confusing because there is another class of external variable--one that exists in a separately compiled source file. So there is a need to distinguish between variables that are merely external to functions and those that are also external to the present source file. To avoid this confusion, I have chosen to refer to external variables in the present source file as globals since they are known to all of the functions of the program. By contrast, locals are known only to the functions (more properly, blocks) in which they are declared. Hereafter, the word external is reserved for variables that exist in other, separately compiled, source files; that is, those declared with the keyword extern.
Local Variables
As we shall see later, variables can be declared within a compound statement. We call these local variables since they are known only to the block in which they appear, and to subordinate blocks. For example, we could write
if (i < 5) { int x; x=j; j=k; k=x; }
if we wanted to swap j and k when i < 5.
Notice that the local variable x is declared within the
compound statement. Unlike globals, which are said to be
static, locals are created dynamically when their block is
entered, and they cease to exist when control leaves the block.
Furthermore, local names supersede the names of globals and other
locals declared at higher levels of nesting. Therefore, locals
may be used freely without regard to the names of "outlyi
ng" variables; clashes cannot occur.Source Files
As we noted above, C programs may consist of source code in more than one file. One method of bringing the parts together is to use the #include preprocessor directive which we saw in Listing 1-1. Another is to compile the source files separately, then combine the separate object files as the program is being linked with library modules. This is very convenient since the linker must be used anyway. It permits us to break large programs into parts so that the whole program need not be recomp iled with every change that is made. As we shall see, the Small C compiler consists of four parts which are brought together in this way.
Although truly external variables must be declared as such with the keyword extern, the Small C compiler assumes that undeclared functions are external. In either case, it generates an EXTRN directive telling the assembler to inform the linker to make the necessary connection. Furthermore, each global declaration generates a PUBLIC directive which ensures that it can be seen from other modules.
Summary
In brief, a Small C program comprises one or more source files. Two mechanisms are provided for bringing together the parts. The text from one file may be included into another file by means of the #include directive (Chapter 11); or the parts may be compiled and assembled separately then linked together.
Each source file consists of comments, preprocessor directives, and global declarations for variables, arrays, pointers, and functions. Each function in turn has a declarator and a body. The declarator names the function and gives local names to its arguments. The body declares the types of the arguments and specifies an algorithm in a block of local declarations, executable statements, and other blocks. And these blocks in turn have their own local declarations, statements, and blocks; and so on.
Programs begin execution in a function called main() and other functions receive control only when they are specifically called. A function is called by writing its name followed by a (possibly null) list of arguments enclosed in parentheses.
These concepts are just the foundation for the things that follow. So read on for the good stuff.