Steve Summit
Copyright 1997,1998
[DISCLAIMER: This is, on balance, a very difficult test.
DO NOT use it in an actual interview situation; you would in all
likelihood only embarrass yourself and insult your interviewee.
In particular, you would obviously not want to ask a question
which you yourself did not know the answer to, but the answers
to many of these questions are not obvious or well-known.
There are several trick questions here, as well as a number which
are explicitly marked "poor". The poor questions are, alas, not
uncommon in actual interviews, and are presented here only so
that their faults and wrong answers can be presented.
This test is intended for study purposes only. Most of the
answers can be found in the comp.lang.c FAQ list.
The answers here were prepared before I read a draft of the
impending C9X standard. I would word several of them
differently today in anticipation of that standard, and
a few of them will become wrong when that standard is adopted.
Acknowledgments: This test is an expanded version of
one I prepared for a training class held at the request
of Tony McNamara, at a now-defunct company called
SCS/Compute (no relation).]
1.1: How can you print a literal % with printf?
A: %%
1.2: Why doesn't \% print a literal % with printf?
A: Backslash sequences are interpreted by the compiler
(\n, \", \0, etc.), and \% is not one of the recognized
backslash sequences. It's not clear what the compiler
would do with a \% sequence -- it might delete it, or
replace it with a single %, or perhaps pass it through as
\ %. But it's printf's behavior we're trying to change,
and printf's special character is %. So it's a
%-sequence we should be looking for to print a literal %,
and printf defines the one we want as %%.
1.3: Are the parentheses in a return statement mandatory?
A: No. The formal syntax of a return statement is
return expression ;
But it's legal to put parentheses around any expression,
of course, whether they're needed or not.
1.4: How can %f work for type double in printf if %lf is
required in scanf?
A: In variable-length argument lists such as printf's, the
old "default argument promotions" apply, and type float
is implicitly converted to double. So printf always
receives doubles, and defines %f to be the sequence that
works whether you had passed a float or a double.
(Strictly speaking, %lf is *not* a valid printf format
specifier, although most versions of printf quietly
accept it.)
scanf, on the other hand, always accepts pointers, and
the types pointer-to-float and pointer-to-double are very
different (especially when you're using them for storing
values). No implicit promotions apply.
1.5: If a machine uses some nonzero internal bit pattern for
null pointers, how should the NULL macro be defined?
A: As 0 (or (char *)0), as usual. The *compiler* is
responsible for translating null pointer constants into
internal null pointer representations, not the
preprocessor.
1.6: If p is a pointer, is the test
if(p)
valid? What if a machine uses some nonzero internal bit
pattern for null pointers?
A: The test is always valid. Since the definition of "true"
in C is "not equal to 0," the test is equivalent to
if(p != 0)
and the compiler then translates the 0 into the
appropriate internal representation of a null pointer.
1.7: What is the ANSI Standard definition of a null pointer
constant?
A: "An integral constant expression with the value 0, or
such an expression cast to type (void *)".
1.8: What does the auto keyword mean? When is it needed?
A: auto is a storage-class specifier, just like extern and
static. But since automatic duration is the default for
local variables (and meaningless, in fact illegal, for
global variables), the keyword is never needed. (It's a
relic from the dawn of C.)
1.9: What does *p++ increment?
A: The pointer p. To increment what p points to, use (*p)++
or ++*p.
1.10: What's the value of the expression 5["abcdef"] ?
A: 'f'.
(The string literal "abcdef" is an array, and the
expression is equivalent to "abcdef"[5]. Why is the
inside-out expression equivalent? Because a[b] is
equivalent to *(a + b) which is equivalent to *(b + a)
which is equivalent to b[a].
1.11: [POOR QUESTION] How can you swap two integer variables
without using a temporary?
A: The reason that this question is poor is that the answer
ceased to be interesting when we came down out of the
trees and stopped using assembly language.
The "classic" solution, expressed in C, is
a ^= b;
b ^= a;
a ^= b;
Due to the marvels of the exclusive-OR operator, after
these three operations, a's and b's values will be
swapped.
However, it is exactly as many lines, and (if we can
spare one measly word on the stack) is likely to be more
efficient, to write the obvious
int t = a;
a = b;
b = t;
No, this doesn't meet the stipulation of not using a
temporary. But the whole reason we're using C and not
assembly language (well, one reason, anyway) is that
we're not interested in keeping track of how many
registers we have.
If the processor happens to have an EXCH instruction, the
compiler is more likely to recognize the possibility of
using it if we use the three-assignment idiom, rather
than the three-XOR.
By the way, the even more seductively concise rendition
of the "classic" trick in C, namely
a ^= b ^= a ^= b
is, strictly speaking, undefined, because it modifies a
twice between sequence points. Also, if an attempt is
made to use the idiom (in any form) in a function which
is supposed to swap the locations pointed to by two
pointers, as in
swap(int *p1, *p2)
{
*p1 ^= *p2;
*p2 ^= *p1;
*p1 ^= *p2;
}
then the function will fail if it is ever asked to swap a
value with itself, as in
swap(&a, &a);
or
swap(&a[i], &a[j]);
when i == j. (The latter case is not uncommon in sorting
algorithms. The effect when p1 == p2 is that the pointed-
to value is set to 0.)
1.12: What is sizeof('A') ?
A: The same as sizeof(int). Character constants have type
int in C. (This is one area in which C++ differs.)
1.13: According to the ANSI Standard, how many bits are there
in an int? A char? A short int? A long int? In other
words, what is sizeof(int) ? sizeof(char) ?
sizeof(short int) ? sizeof(long int) ?
A: ANSI guarantees that the range of a signed char is at
least +-127, of a short int is at least +-32767, of an int
at least +-32767, and a long int at least +-2147483648. So
we can deduce that a char must be at least 8 bits, an int
or a short int must be at least 16 bits, and a long int
must be at least 32 bits. The only guarantees about
sizeof are that
1 = sizeof(char) <= sizeof(short) <= sizeof(int) <= sizeof(long)
1.14: If arr is an array, in an ordinary expression, what's
the difference between arr and &arr ?
A: If the array is of type T, the expression arr yields a
pointer of type pointer-to-T pointing to the array's
first element. The expression &arr, on the other hand,
yields a pointer of type pointer-to-array-of-T pointing
to the entire array. (The two pointers will likely have
the same "value," but the types are distinct. The
difference would be visible if you assigned or
incremented the resulting pointer.)
1.15: What's the difference between
char *p = malloc(n);
and
char *p = malloc(n * sizeof(char));
?
A: There is little or no difference, since sizeof(char) is
by definition exactly 1.
1.16: What's the difference between these three declarations?
char *a = "abc";
char b[] = "abc";
char c[3] = "abc";
A: The first declares a pointer-to-char, initialized to
point to a four-character array somewhere in (possibly
read-only) memory containing the four characters
a b c \0. The second declares an array (a writable
array) of 4 characters, initially containing the
characters a b c \0. The third declares an array of 3
characters, initially containing a b c. (The third array
is therefore not an immediately valid string.)
1.17: The first line of a source file contains the line
extern int f(struct x *);
The compiler warns about "struct x declared inside
parameter list". What is the compiler worried about?
A: For two structures to be compatible, they must not only
have the same tag name but be defined in the same scope.
A function prototype, however, introduces a new, nested
scope for its parameters. Therefore, the structure tag x
is defined in this narrow scope, which almost immediately
disappears. No other struct x pointer in this
translation unit can therefore be compatible with f's
first parameter, so it will be impossible to call f
correctly (at least, without drawing more warnings). The
warning alluded to in the question is trying to tell you
that you shouldn't mention struct tags for the first time
in function prototypes.
(The warning message in the question is actually produced
by gcc, and the message runs on for two more lines,
explaining that the scope of the structure declared "is
only this definition or declaration, which is probably
not what you want.")
1.18: List several ways for a function to safely return a
string.
A: It can return a pointer to a static array, or it can
return a pointer obtained from malloc, or it can fill in
a buffer supplied by the caller.
1.19: [hard] How would you implement the va_arg() macro in
A: A straightforward implementation, assuming a conventional
stack-based architecture, is
#define va_arg(argp, type) (((type *)(argp += sizeof(type)))[-1])
This assumes that type va_list is char *.
1.20: Under what circumstances is the declaration
typedef xxx int16;
where xxx is replaced with an appropriate type for a
particular machine, useful?
A: It is potentially useful if the int16 typedef is used to
declare variables or structures which will be read from
or written to some external data file or stream in some
fixed, "binary" format. (However, the typedef can at
most ensure that the internal type is the same size as
the external representation; it cannot correct for any
byte order discrepancies.)
Such a typedef may also be useful for allowing
precompiled object files or libraries to be used with
different compilers (compilers which define basic types
such as int differently), without recompilation.
1.21: Suppose that you declare
struct x *xp;
without any definition of struct x. Is this legal?
Under what circumstances would it be useful?
A: It is perfectly legal to refer to a structure which has
not been "fleshed out," as long as the compiler is never
asked to compute the size of the structure or generate
offsets to any members. Passing around pointers to
otherwise undefined structures is quite acceptable, and
is a good way of implementing "opaque" data types in C.
1.22: What's the difference between
struct x1 { ... };
typedef struct { ... } x2;
?
A: The first declaration declares a structure tag x1; the
second declares a typedef name x2. The difference
becomes clear when you declare actual variables of the
two structure types:
struct x1 a, b;
but
x2 a, b;
(This distinction is insignificant in C++, where all
structure and class tags automatically become full-
fledged types, as if via typedef.)
1.23: What do these declarations mean?
int **a();
int (*b)();
int (*c[3])();
int (*d)[10];
A: declare a as function returning pointer to pointer to int
declare b as pointer to function returning int
declare c as array of 3 pointers to functions returning int
declare d as pointer to array of 10 ints
The way to read these is "inside out," remembering that
[] and () bind more tightly than *, unless overridden by
explicit parentheses.
1.24: State the declaration for a pointer to a function
returning a pointer to char.
A: char *(*f)();
1.25: If sizeof(long int) is 4, why might sizeof report the
size of the structure
struct x {char c; long int i;};
as 8 instead of 5?
A: The compiler will typically allocate invisible padding
between the two members of the structure, to keep i
aligned on a longword boundary.
1.26: If sizeof(long int) is 4, why might sizeof report the
size of the structure
struct y {long int i; char c;};
as 8 instead of 5?
A: The compiler will typically allocate invisible padding at
the end of structure, so that if an array of these
structures is allocated, the i's will all be aligned.
1.27: [POOR QUESTION] If i starts out as 1, what does the
expression
i++ + i++
evaluate to? What is i's final value?
A: This is a poor question because it has no answer. The
expression attempts to modify i twice between sequence
points (not to mention modifying and inspecting i's
value, where the inspection is for purposes other than
determining the value to be stored), so the expression is
undefined. Different compilers can (and do) generate
different results, and none of them is "wrong."
1.28: Consider these definitions:
#define Push(val) (*stackp++ = (val))
#define Pop() (*--stackp)
int stack[100];
int *stackp = stack;
Now consider the expression
Push(Pop() + Pop())
1. What is the expression trying to do? In what sort of
program might such an expression be found?
2. What are some deficiencies of this implementation?
Under what circumstances might it fail?
A: The expression is apparently intended to pop two values
from a stack, add them, and push the result. This code
might be found in a calculator program, or in the
evaluation loop of the engine for a stack-based language.
The implementation has at least four problems, however.
The Push macro does not check for stack overflow; if more
than 100 values are pushed, the results will be
unpredictable. Similarly, the the Pop macro does not
check for stack underflow; an attempt to pop a value when
the stack is empty will likewise result in undefined
behavior.
On a stylistic note, the stackp variable is global as far
as the Push and Pop macros are concerned. If it is
certain that, in a particular program, only one stack
will be used, this assumption may be a reasonable one,
as it allows considerably more succinct invocations.
If multiple stacks are a possibility, however, it might
be preferable to pass the stack pointer as an argument
to the Push and Pop macros.
Finally, the most serious problem is that the "add"
operation as shown above is *not* guaranteed to work!
After macro expansion, it becomes
(*stackp++ = ((*--stackp) + (*--stackp)))
This expression modifies a single object more than once
between sequence points; specifically, it modifies stackp
three times. It is not guaranteed to work; moreover,
there are popular compilers (one is gcc) under which it
*will* *not* work as expected. (The extra parentheses
do nothing to affect the evaluation order; in particular,
they do not make it any more defined.)
1.29: [POOR QUESTION] Write a small function to sort an array
of integers.
A: This is a poor question because no one writes small
functions to sort arrays of integers any more, except as
pedagogical exercises. If you have an array of integers
that needs sorting, the thing to do is call your library
sort routine -- in C, qsort(). So here is my "small
function":
static int intcmp(const void *, const void *);
sortints(int a[], int n)
{
qsort(a, n, sizeof(int), intcmp);
}
static int intcmp(const void *p1, const void *p2)
{
int i1 = *(const int *)p1;
int i2 = *(const int *)p2;
if(i1 < i2)
return -1;
else if(i1 > i2)
return 1;
else return 0;
}
(The reason for using two comparisons and three explicit
return statements rather than the "more obvious"
return i1 - i2; is that i1 - i2 can underflow, with
unpredictable results.)
1.30: State the ANSI rules for determining whether an
expression is defined or undefined.
A: An expression is undefined if, between sequence points,
it attempts to modify the same location twice, or if it
attempts to both read from and write to the same
location. It's permissible to read and write the same
location only if the laws of causality (a higher
authority even than X3.159) prove that the read must
unfailingly precede the write, that is, if the write is
of a value which was computed from the value which was
read. This exception means that old standbys such as
i = i + 1 are still legal.
Sequence points occur at the ends of full expressions
(expression statements, and the expressions in if, while,
for, do/while, switch, and return statements, and
initializers), at the &&, ||, and comma operators, at the
end of the first expression in a ?: expression, and just
before the call of a function (after the arguments have
all been evaluated).
(The actual language from the ANSI Standard is
Between the previous and next sequence point an
object shall have its stored value modified at
most once by the evaluation of an expression.
Furthermore, the prior value shall be accessed
only to determine the value to be stored.
)
1.30a: What's the difference between these two declarations?
extern char x[];
extern char *x;
A: The first is an external declaration for an array of char
named x, defined elsewhere. The second is an external
declaration for a pointer to char named x, also defined
elsewhere. These declarations could not both appear in
the same program, because they specify incompatible types
for x.
1.31: What's the difference between these two declarations?
int f1();
extern int f2();
A: There is no difference; the extern keyword is essentially
optional in external function declarations.
1.32: What's the difference between these two declarations?
extern int f1();
extern int f2(void);
A: The first is an old-style function declaration declaring
f1 as a function taking an unspecified (but fixed) number
of arguments; the second is a prototype declaration
declaring f2 as a function taking precisely zero
arguments.
1.33: What's the difference between these two definitions?
int f1()
{
}
int f2(void)
{
}
A: There is no difference, other than that the first uses
the old definition style and the second uses the
prototype style. Both functions take zero arguments.
1.34: How does operator precedence influence order of
evaluation?
A: Only partially. Precedence affects the binding of
operators to operands, but it does *not* control (or even
influence) the order in which the operands themselves are
evaluated. For example, in
a() + b() * c()
we have no idea what order the three functions will be
called in. (The compiler might choose to call a first,
even though its result will be needed last.)
1.35: Will the expression in
if(n != 0 && sum / n != 0)
ever divide by 0?
A: No. The "short circuiting" behavior of the && operator
guarantees that sum / n will not be evaluated if n is 0
(because n != 0 is false).
1.36: Will the expression
x = ((n == 0) ? 0 : sum / n)
ever divide by 0?
A: No. Only one of the pair of controlled expressions in a
?: expression is evaluated. In this example, if n is 0,
the third expression will not be evaluated at all.
1.37: Explain these three fragments:
if((p = malloc(10)) != NULL) ...
if((fp = fopen(filename, "r")) == NULL) ...
while((c = getc(fp)) != EOF) ...
A: The first calls malloc, assigns the result to p, and does
something if the just-assigned result is not NULL.
The second calls fopen, assigns the result to fp, and
does something if the just-assigned result is NULL.
The third repeatedly calls getc, assigns the results in
turn to c, and does something as long as each just-
assigned result is not EOF.
1.38: What's the difference between these two statements?
++i;
i++;
A: There is no difference. The only difference between the
prefix and postfix forms of the autoincrement operator is
the value passed on to the surrounding expression, but
since the expressions in the question stand alone as
expression statements, the value is discarded, and each
expression merely serves to increment i.
1.39: Why might a compiler warn about conversions or
assignments from char * to int * ?
A: In general, compilers complain about assignments between
pointers of different types (and are required by the
Standard to so complain) because such assignments do not
make sense. A pointer to type T1 is supposed to point to
objects of type T1, and presumably the only reason for
assigning the pointer to a pointer of a different type,
say pointer-to-T2, would be to try to access the pointed-
to object as a value of type T2, but if the pointed-to
object is of type T2, why were we pointing at it with a
pointer-to-T1 in the first place?
In the particular example cited in the question, the
warning also implies the possibility of unaligned access.
For example, this code:
int a[2] = {0, 1};
char *p = &a; /* suspicious */
int *ip = p + 1; /* even more suspicious */
printf("%d\n", *ip);
is likely to crash (perhaps with a "Bus Error") because
the programmer has contrived to make ip point to an odd,
unaligned address.
When it is desired to use pointers of the "wrong" type,
explicit casts must generally be used. One class of
exceptions is exemplified by malloc: the memory it
allocates, and hence the pointers it returns, are
supposed to be usable as any type the programmer wishes,
so malloc's return value will almost always be the
"wrong" type. To avoid the need for so much explicit,
dangerous casting, ANSI invented the void * type, which
quietly interconverts (i.e. without warning) between
other pointer types. Pointers of type void * are
therefore used as containers to hold "generic" pointers
which are known to be safely usable as pointers to other,
more specific types.
1.40: When do ANSI function prototype declarations *not*
provide argument type checking, or implicit conversions?
A: In the variable-length part of variable-length argument
lists, and (perhaps obviously, perhaps not) when no
prototype is in scope at all. The point is that it is
not safe to assume that since prototypes have been
invented, programmers don't have to be careful about
matching function-call arguments any more. Care must
still be exercised in variable-length argument lists, and
if prototypes are to take care of the rest, care must be
exercised to use prototypes correctly.
1.41: State the rule(s) underlying the "equivalence" of arrays
and pointers in C.
A: Rule 1: When an array appears in an expression where its
value is needed, the value generated is a pointer to the
array's first element. Rule 2: Array-like subscripts
(integer expressions in brackets) may be used to
subscript pointers as well as arrays; the expression p[i]
is by definition equivalent to *(p+i). (Actually, by
rule 1, subscripts *always* find themselves applied to
pointers, never arrays.)
1.42: What's the difference between these two declarations?
extern int f2(char []);
extern int f1(char *);
A: There is no difference. The compiler always quietly
rewrites function declarations so that any array
parameters are actually declared as pointers, because
(by the equivalence of arrays and pointers) a pointer
is what the function will actually receive.
1.43: Rewrite the parameter declaration in
f(int x[5][7])
{
}
to explicitly show the pointer type which the compiler
will assume for x.
A: f(int (*x)[7])
{
}
Note that the type int (*)[7] is *not* the same as
int **.
1.44: A program uses a fixed-size array, and in response to
user complaints you have been asked to replace it with a
dynamically-allocated "array," obtained from malloc.
Which parts of the program will need attention? What
"gotchas" must you be careful of?
A: Ideally, you will merely have to change the declaration
of the array from an array to a pointer, and add one call
to malloc (with a check for a null return, of course) to
initialize the pointer to a dynamically-allocated
"array." All of the code which accesses the array can
remain unchanged, because expressions of the form x[i]
are valid whether x is an array or a pointer.
The only thing to be careful of is that if the existing
code ever used the sizeof operator to determine the size
of the array, that determination becomes grossly invalid,
because after the change, sizeof will return only the
size of the pointer.
1.45: A program which uses a dynamically allocated array is
still running into problems because the initial
allocation is not always big enough. Your task is now
to use realloc to make the "array" bigger, if need be.
What must you be careful of?
A: The actual call to realloc is straightforward enough, to
request that the base pointer now point at a larger block
of memory. The problem is that the larger block of
memory may be in a different place; the base pointer may
move. Therefore, you must reassign not only the base
pointer, but also any copies of the base pointer you may
have made, and also any pointers which may have been set
to point anywhere into the middle of the array. (For
pointers into the array, you must in general convert them
temporarily into offsets from the base pointer, then call
realloc, then recompute new pointers based on the offsets
and the new base pointer. See also question 2.10.)
1.46: How can you you use sizeof to determine the number of
elements in an array?
A: The standard idiom is
sizeof(array) / sizeof(array[0])
(or, equivalently, sizeof(array) / sizeof(*array) ).
1.47: When sizeof doesn't work (when the array is declared
extern, or is a parameter to a function), what are some
strategies for determining the size of an array?
A: Use a sentinel value as the last element of the array;
pass the size around in a separate variable or as a
separate function parameter; use a preprocessor macro
to define the size.
1.48: Why might explicit casts on malloc's return value, as in
int *ip = (int *)malloc(10 * sizeof(int));
be a bad idea?
A: Although such casts used to be required (before the
void * type, which converts quietly and implicitly, was
invented), they can now be considered poor style, because
they will probably muzzle the compiler's attempts to warn
you on those occasions when you forget to #include
will be incorrectly assumed to be a function returning
int.
1.49: How are Boolean true/false values defined in C? What
values can the == (and other logical and comparison
operators) yield?
A: The value 0 is considered "false," and any nonzero value
is considered "true." The relational and logical
operators all yield 0 for false, 1 for true.
1.50: x is an integer, having some value. What is the value of
the expression
0 <= !x && !!x < 2
?
A: 1.
1.51: In your opinion, is it acceptable for a header file to
contain #include directives for other header files?
A: The argument in favor of "nested #include files" is that
they allow each header to arrange to have any subsidiary
definitions, upon which its own definitions depend, made
automatically. (For example, a file containing a
prototype for a function that accepts an argument of type
FILE * could #include
alternative is to potentially require everyone who
includes a particular header file to include one or
several others first, or risk cryptic errors.
The argument against is that nested headers can be
confusing, can make definitions difficult to find,
and can in some circumstances even make it difficult
to determine which file(s) is/are being included.
1.52: How can a header file be protected against being
included multiple times (perhaps due to nested
#include directives)?
A: The standard trick is to place lines like
#ifndef headerfilename_H
#define headerfilename_H
at the beginning of the file, and an extra #endif at the
end.
1.53: A source file contains as its first two lines:
#include "a.h"
int i;
The compiler complains about an invalid declaration on
line 2. What's probably happening?
A: It's likely that the last declaration in a.h is missing
its trailing semicolon, causing that declaration to merge
into "int i", with meaningless results. (That is, the
merged declaration is probably something along the lines
of
extern int f() int i;
or
struct x { int y; } int i;
.)
1.54: What's the difference between a header file and a
library?
A: A header file typically contains declarations and
definitions, but it never contains executable code.
(A header file arguably shouldn't even contain any
function bodies which would compile into executable
code.) A library, on the other hand, contains only
compiled, executable code and data.
A third-party library is often delivered as a library and
a header file. Both pieces are important. The header
file is included during compilation, and the library is
included during linking.
1.55: What are the acceptable declaration(s) for main()? A: The most common declarations, all legal, are:
main()
int main()
int main(void)
int main(int argc, char **argv)
int main(int argc, char *argv[])
int main(argc, argv) int argc; char *argv[];
(Basically: the return type must be an implicit or
explicit int; the parameter list must either be empty, or
void, or one int plus one array of strings; and the
function may be declared using either old-style or
prototyped syntax. The actual names of the two
parameters are arbitrary, although of course argc and
argv are traditional.)
1.56: You wish to use ANSI function prototypes to guard against
errors due to accidentally calling functions with
incorrect arguments. Where should you place the
prototype declarations? How can you ensure that the
prototypes will be maximally effective?
A: The prototype for a global function should be placed in a
header file, and the header file should be included in
all files where the function is called, *and* in the file
where the function is defined. The prototype for a
static function should be placed at the top of the file
where the function is defined.
Since following these rules is only slightly less hard
than getting all function calls right by hand (i.e.
without the aid of prototypes), the compiler should be
configured to warn about functions called without
prototypes in scope, *and* about functions defined
without prototypes in scope.
1.57: Why must the variable used to hold getchar's return value
be declared as int?
A: Because getchar can return, besides all values of type
char, the additional "out of band" value EOF, and there
obviously isn't room in a variable of type char to hold
one more than the number of values which can be
unambiguously stored in a variable of type char.
1.58: You must write code to read and write "binary" data
files. How do you proceed? How will you actually open,
read, and write the files?
A: When calling fopen, the files must be opened using the b
modifier (e.g. "rb", "wb"). Binary data files are
generally read and written a byte at a time using getc
and putc, or a data structure at a time using fread and
fwrite.
1.59: Write the function
void error(const char *message, ...);
which accepts a message string, possibly containing %
sequences, along with optional extra arguments
corresponding to the % sequences, and prints the string
"error: ", followed by the message as printf would print
it, followed by a newline, all to stderr.
A: #include
#include
void error(char *fmt, ...)
{
va_list argp;
fprintf(stderr, "error: ");
va_start(argp, fmt);
vfprintf(stderr, fmt, argp);
va_end(argp);
fprintf(stderr, "\n");
}
1.60: Write the function
char *vstrcat(char *, ...);
which accepts a variable number of strings and
concatenates them all together into a block of malloc'ed
memory just big enough for the result. The end of the
list of strings will be indicated with a null pointer.
For example, the call
char *p = vstrcat("Hello, ", "world!", (char *)NULL);
should return the string "Hello, world!".
A: #include
#include
#include
char *vstrcat(char *first, ...)
{
size_t len;
char *retbuf;
va_list argp;
char *p;
if(first == NULL)
return NULL;
len = strlen(first);
va_start(argp, first);
while((p = va_arg(argp, char *)) != NULL)
len += strlen(p);
va_end(argp);
retbuf = malloc(len + 1); /* +1 for trailing \0 */
if(retbuf == NULL)
return NULL; /* error */
(void)strcpy(retbuf, first);
va_start(argp, first); /* restart; 2nd scan */
while((p = va_arg(argp, char *)) != NULL)
(void)strcat(retbuf, p);
va_end(argp);
return retbuf;
}
1.61: Write a stripped-down version of printf which accepts
only the %c, %d, %o, %s, %x, and %% format specifiers.
(Do not worry about width, precision, flags, or length
modifiers.)
A: [Although this question is obviously supposed to test one's
familiarity with the va_ macros, a significant nuisance in
composing a working answer is performing the sub-task of
converting integers to digit strings. For some reason,
back when I composed this test, I felt it appropriate to
defer that task to an "itoa" function; perhaps I had just
presented an implementation of itoa to the same class for
whom I first prepared this test.]
#include
#include
extern char *itoa(int, char *, int);
myprintf(const char *fmt, ...)
{
const char *p;
va_list argp;
int i;
char *s;
char fmtbuf[256];
va_start(argp, fmt);
for(p = fmt; *p != '\0'; p++)
{
if(*p != '%')
{
putchar(*p);
continue;
}
switch(*++p)
{
case 'c':
i = va_arg(argp, int);
putchar(i);
break;
case 'd':
i = va_arg(argp, int);
s = itoa(i, fmtbuf, 10);
fputs(s, stdout);
break;
case 's':
s = va_arg(argp, char *);
fputs(s, stdout);
break;
case 'x':
i = va_arg(argp, int);
s = itoa(i, fmtbuf, 16);
fputs(s, stdout);
break;
case '%':
putchar('%');
break;
}
}
va_end(argp);
}
1.62: You are to write a program which accepts single
keystrokes from the user, without waiting for the RETURN
key. You are to restrict yourself only to features
guaranteed by the ANSI/ISO C Standard. How do you
proceed?
A: You proceed by pondering the sorrow of your fate, and
perhaps by complaining to your boss/professor/psychologist
that you've been given an impossible task. There is no
ANSI Standard function for reading one keystroke from the
user without waiting for the RETURN key. You'll have to
use facilities specific to your operating system; you
won't be able to write the code strictly portably.
1.63: [POOR QUESTION] How do you convert an integer to binary
or hexadecimal?
A: The question is poor because an integer is a *number*; it
doesn't make much sense to ask what base it's in. If I'm
holding eleven apples, what base is that in?
(Of course, internal to the computer, an integer is
almost certainly represented in binary, although it's not
at all unreasonable to think of it as being hexadecimal,
or decimal for that matter.)
The only time the base of a number matters is when it's
being read from or written to the outside world as a
string of digits. In those cases, and depending on just
what you're doing, you can specify the base by picking
the correct printf or scanf format specifier (%d, %o, or
%x), or by picking the third argument to strtol. (There
isn't a Standard function to convert an integer to a
string using an arbitrary base. For that task, it's a
straightforward exercise to write a function to do the
conversion. Some versions of the nonstandard itoa
function also accept a base or radix argument.)
1.64: You're trying to discover the sizes of the basic types
under a certain compiler. You write the code
printf("sizeof(char) = %d\n", sizeof(char));
printf("sizeof(short) = %d\n", sizeof(short));
printf("sizeof(int) = %d\n", sizeof(int));
printf("sizeof(long) = %d\n", sizeof(long));
However, all four values are printed as 0. What have you
learned?
A: You've learned that this compiler defines size_t, the
type returned by sizeof, as an unsigned long int, and that
the compiler also defines long integers as having a
larger size than plain int. (Furthermore, you've learned
that the machine probably uses big-endian byte order.)
Finally, you may have learned that the code you should
have written is along the lines of either
printf("sizeof(int) = %u\n", (unsigned)sizeof(int));
or
printf("sizeof(int) = %lu\n", (unsigned long)sizeof(int));
Section 2. What's wrong with...?
2.1: main(int argc, char *argv[])
{
...
if(argv[i] == "-v")
...
A: Applied to pointers, the == operator compares only
whether the pointers are equal. To compare whether the
strings are equal, you'll have to call strcmp.
2.2: a ^= b ^= a ^= b
(What is the expression trying to do?)
A: The expression is undefined because it modifies the
variable a twice between sequence points. What it's
trying to do is swap the variables a and b using a
hoary old assembly programmer's trick.
2.3: char* p1, p2;
A: p2 will be declared as type char, *not* pointer-to-char.
2.4: char c;
while((c = getchar()) != EOF)
...
A: The variable used to contain getchar's return value must
be declared as int if EOF is to be reliably detected.
2.5: while(c = getchar() != EOF)
...
A: Parentheses are missing; the code will call getchar,
compare the result to EOF, assign the result *of the
comparison* to c, and take another trip around the loop
if the condition was true (i.e. if the character read was
not EOF). (The net result will be that the input will be
read as if it were a string of nothing but the character
'\001'. The loop would still halt properly on EOF,
however.)
2.6: int i, a[10];
for(i = 0; i <= 10; i++)
a[i] = 0;
A: The loop assigns to the nonexistent eleventh value of the
array, a[10], because it uses a loop continuation
condition of <= 10 instead of < 10 or <= 9.
2.7: #include
...
#define TRUE 1
#define FALSE 0
...
if(isalpha(c) == TRUE)
...
A: Since *any* nonzero value is considered "true" in C; it's
rarely if ever a good idea to compare explicitly against
a single TRUE value (or FALSE, for that matter). In
particular, the
tend to return nonzero values other than 1, so the test
as written is likely to fail even for alphabetic
characters. The correct test is simply
if(isalpha(c))
2.8: printf("%d\n", sizeof(int));
A: The sizeof operator returns type size_t, which is an
unsigned integral type *not* necessarily the same size as
an int. The correct code is either
printf("sizeof(int) = %u\n", (unsigned int)sizeof(int));
or
printf("sizeof(int) = %lu\n", (unsigned long int)sizeof(int));
2.9: p = realloc(p, newsize);
if(p == NULL)
{
fprintf(stderr, "out of memory\n");
return;
}
A: If realloc returns null, and assuming p used to point to
some malloc'ed memory, the memory remains allocated,
although having overwritten p, there may well be no way
to use or free the memory. The code is a potential
memory leak.
2.10: /* p points to a block of memory obtained from malloc; */
/* p2 points somewhere within that block */
newp = realloc(p, newsize);
if(newp != NULL && newp != p)
{
int offset = newp - p;
p2 += offset;
}
A: Pointer subtraction is well-defined only for pointers
into the same block of memory. If realloc moves the
block of memory while changing its size, which is the
very case the code is trying to test for, then the
subtraction newp - p is invalid, because p and newp
do not point into the same block of memory. (The
subtraction could overflow or otherwise produce
nonsensical results, especially on segmented
architectures. Strictly speaking, *any* use of
the old value of p after realloc moves the block
is invalid, even comparing it to newp.)
The correct way to relocate p2 within the possibly-moved
block, correcting *all* the problems in the original
(including a subtle one not mentioned), is:
ptrdiff_t offset = p2 - p;
newp = realloc(p, newsize);
if(newp != NULL)
{
p = newp;
p2 = p + offset;
}
2.11: int a[10], b[10];
...
a = b;
A: You can't assign arrays.
2.12: int i = 0;
char *p = i;
if(p == 0)
... A: Assigning an integer 0 to a pointer does not reliably
result in a null pointer. (You must use a *constant* 0.)
No comments:
Post a Comment