分类: C/C++
2010-06-13 10:38:31
[Last modified : 2006-11-30]
: Diagnostics
: Character Class Tests
: Error Codes Reported by (Some) Library Functions
: Implementation-defined Floating-Point Limits
: Implementation-defined Limits
: Locale-specific Information
: Mathematical Functions
: Non-local Jumps
: Signals
: Variable Argument Lists
: Definitions of General Use
: Input and Output
: Utility functions
: String functions
: Time and Date functionsvoid assert(int expression);
NDEBUG
is defined where
is included.) If expression equals zero, message printed on
stderr
and abort
called to terminate execution. Source filename and line number in
message are
from preprocessor macros __FILE__
and __LINE__
.[Contents]
int isalnum(int c);
isalpha(c)
or isdigit(c)
int isalpha(int c);
isupper(c)
or islower(c)
int iscntrl(int c);
0x00
(NUL
) to 0x1F
(US
), and 0x7F
(DEL
)int isdigit(int c);
int isgraph(int c);
int islower(int c);
int isprint(int c);
0x20
(' '
) to 0x7E
('~'
)int ispunct(int c);
int isspace(int c);
int isupper(int c);
int isxdigit(int c);
int tolower(int c);
int toupper(int c);
[Contents]
Notes:
[Contents]
FLT_ROUNDS
Where the prefix "FLT
" pertains to type float
,
"DBL
" to type double
, and "LDBL
"
to type long double
:
FLT_DIG
DBL_DIG
LDBL_DIG
FLT_EPSILON
DBL_EPSILON
LDBL_EPSILON
1.0 + x !=
1.0
FLT_MANT_DIG
DBL_MANT_DIG
LDBL_MANT_DIG
FLT_RADIX
,
in mantissaFLT_MAX
DBL_MAX
LDBL_MAX
FLT_MAX_EXP
DBL_MAX_EXP
LDBL_MAX_EXP
FLT_RADIX
can be raised and remain representableFLT_MIN
DBL_MIN
LDBL_MIN
FLT_MIN_EXP
DBL_MIN_EXP
LDBL_MIN_EXP
FLT_RADIX
can be raised and remain representable[Contents]
CHAR_BIT
char
CHAR_MAX
char
CHAR_MIN
char
SCHAR_MAX
signed char
SCHAR_MIN
signed char
UCHAR_MAX
unsigned char
SHRT_MAX
short
SHRT_MIN
short
USHRT_MAX
unsigned short
INT_MAX
int
INT_MIN
int
UINT_MAX
unsigned int
long
long
unsigned long
[Contents]
char* ;
char* grouping;
char* thousands_sep;
char* currency_symbol;
char* int_curr_symbol;
char* mon_decimal_point;
char* mon_grouping;
char* mon_thousands_sep;
char* negative_sign;
char* positive_sign;
char frac_digits;
char int_frac_digits;
char n_cs_precedes;
1
) or follows (0
)
negative monetary valueschar n_sep_by_space;
1
) or is not (0
)
separated by space from negative monetary values;
0
1
2
3
4
char p_cs_precedes;
1
) or follows (0
)
positive monetary valueschar p_sep_by_space;
1
) or is not (0
)
separated by space from non-negative monetary valueschar p_sign_posn;
n_sign_posn
decimal_point
.struct
lconv* localeconv(void);
char* setlocale(int ,
const char* locale);
category
and locale
. Returns string describing new
locale or null on error. (Implementations are permitted to define
values of category
additional to those describe here.)LC_ALL
category
argument for all categoriesLC_NUMERIC
category
for numeric formatting informationLC_MONETARY
category
for monetary formatting informationLC_COLLATE
category
for information affecting collating functionsLC_CTYPE
category
for information affecting character
class tests functionsLC_TIME
category
for information affecting time conversions functionsNULL
[Contents]
On domain error, implementation-defined value returned and errno
set to EDOM
.
On range error, errno
set to ERANGE
and return value is HUGE_VAL
with correct sign for overflow, or zero for underflow. Angles are in
radians.
double exp(double x);
x
double log(double x);
x
double log10(double x);
x
double pow(double x, double y);
x
raised to power y
double sqrt(double x);
x
double ceil(double x);
x
double floor(double x);
x
double fabs(double x);
x
double ldexp(double x, int n);
x
times 2 to the power n
double frexp(double x, int* exp);
x
non-zero, returns value, with absolute
value in interval [1/2, 1), and assigns to *exp
integer such that product of return value and 2 raised to the power *exp
equals x
; if x
zero,
both return value and *exp
are zerodouble modf(double x, double* ip);
*ip
integral part of x
, both with same sign as x
double fmod(double x, double y);
y
non-zero, floating-point remainder of x/y
,
with same sign as x
; if y
zero, result is implementation-defineddouble sin(double x);
x
double cos(double x);
x
double tan(double x);
x
double asin(double x);
x
double acos(double x);
x
double atan(double x);
x
double atan2(double y, double x);
y/x
double sinh(double x);
x
double cosh(double x);
x
double tanh(double x);
x
[Contents]
int setjmp(jmp_buf
env);
env
and returns
zero. Subsequent call to longjmp
with same env returns non-zero.void longjmp(jmp_buf
env, int val);
setjmp
with specified env
. Execution resumes as a
second return from setjmp
,
with returned value val
if specified value
non-zero, or 1 otherwise.[Contents]
SIGFPE
SIGILL
SIGINT
SIGSEGV
SIGTERM
signal
return value indicating error
void (*(int sig, void (*handler)(int)))(int);
handler
is SIG_DFL
,
implementation-defined default behaviour will be used; if SIG_IGN
,
signal will be ignored; otherwise function pointed to by handler
will be invoked with argument sig. In the last case,
handling is restored to default behaviour before handler
is called. If handler
returns, execution
resumes where signal occurred. signal
returns the previous handler or SIG_ERR
on error. Initial
state is implementation-defined. Implementations may may define
signals additional to those listed here.
sig
. Returns zero on success.[Contents]
void (va_list
ap, lastarg);
ap
.
lastarg
is the last named parameter of the
function.type (va_list
ap, type);
type
) and value of
the next unnamed argument.void (va_list
ap);
[Contents]
NULL
offsetof(stype, m)
m
from start of
structure type stype
.ptrdiff_t
size_t
sizeof
operator.[Contents]
setbuf
.
FILENAME_MAX
FOPEN_MAX
tmpnam
.
origin
argument to fseek
specifying current file position.
origin
argument to fseek
specifying end of file.
origin
argument to fseek
specifying beginning of file.
tmpnam
.
mode
argument to setvbuf
specifying full buffering.
mode
argument to setvbuf
specifying line buffering.
mode
argument to setvbuf
specifying no buffering.
sizeof
operator.FILE*
(const char* filename, const char* mode);
filename
and returns a
stream, or NULL
on failure. mode
may be one of the following
for text files:"r"
"w"
"a"
"r+"
"w+"
"a+"
b
included (after the
first character), for binary files.FILE*
freopen(const char* filename, const char* mode, FILE*
stream);
stream
, then
opens file filename
with specified mode and
associates it with stream
. Returns stream
or NULL
on error.int (FILE*
stream);
stream
and returns zero on
success or EOF
on error. Effect undefined for input stream. fflush(NULL)
flushes all output streams.int (FILE*
stream);
EOF
on error, zero otherwise.int (const char* filename);
int (const char* oldname,
const char* newname);
oldname
to newname
.
Returns non-zero on failure.FILE*
();
"wb+"
) which will be
removed when closed or on normal program termination. Returns stream or
NULL
on failure.char* (char s[L_tmpnam]);
s
(if s
non-null) and returns unique name for a temporary file. Unique name is
returned for each of the first TMP_MAX
invocations.int (FILE*
stream, char* buf, int mode, size_t
size);
stream
. mode
is _IOFBF
for full buffering, _IOLBF
for line buffering, _IONBF
for no buffering. Non-null buf
specifies
buffer of size size
to be used; otherwise, a
buffer is allocated. Returns non-zero on error. Call must be before
any other operation on stream.void (FILE*
stream, char* buf);
buf
,
turns off buffering, otherwise equivalent to (void)setvbuf(stream, buf, _IOFBF, BUFSIZ)
.int (FILE*
stream, const char* format, ...);
format
) and
writes output to stream stream
. Number of
characters written, or negative value on error, is returned. Conversion
specifications consist of:%
-
+
0
#
o
, first digit
will be zero, for [xX
], prefix 0x
or 0X
to non-zero value, for [eEfgG
], always decimal point, for [gG
]
trailing zeros not removed.*
, value
taken from next argument (which must be int
)..
(separating width from precision):s
,
maximum characters to be printed from the string, for [eEf
],
digits after decimal point, for [gG
], significant digits,
for an integer, minimum number of digits to be printed. If specified as
*
, value taken from next argument (which must be int
).h
short
or unsigned short
l
long
or unsigned long
L
long double
d,i
int
argument, printed in signed decimal notationo
int
argument, printed in unsigned octal notationx,X
int
argument, printed in unsigned hexadecimal notationu
int
argument, printed in unsigned decimal notationc
int
argument, printed as single characters
char*
argumentf
double
argument, printed with format [-
]mmm.
ddde,E
double
argument, printed with format [-
]m.
dddddd(e
|E
)(+
|-
)xxg,G
double
argumentp
void*
argument, printed as pointern
int*
argument : the number of characters written to
this point is written into argument%
int (const char* format,
...);
printf(f,
...)
is equivalent to
fprintf(stdout,
f, ...)
int (char* s, const
char* format, ...);
s
, which must
be large enough to hold the output, rather than to a stream.
Output is NUL
-terminated. Returns length (excluding the
terminating NUL
).int (FILE*
stream, const char* format, va_list
arg);
fprintf
with variable argument list replaced by arg
,
which must have been initialised by the va_start
macro (and may have been used in calls to va_arg
).int (const char* format,
va_list
arg);
printf
with variable argument list replaced by arg
,
which must have been initialised by the va_start
macro (and may have been used in calls to va_arg
).int vsprintf(char* s, const char* format,
va_list
arg);
sprintf
with variable argument list replaced by arg
,
which must have been initialised by the va_start
macro (and may have been used in calls to va_arg
).int (FILE*
stream, const char* format, ...);
stream
according to format format
. The function
returns when format
is fully processed. Returns
number of items converted and assigned, or EOF
if end-of-file or error occurs before any conversion. Each of the
arguments following format
must be a
pointer. Format string may contain:%
*
"h
short
rather than int
l
long
rather than int
,
or double
rather than float
L
long double
rather than float
d
int*
parameter requiredi
int*
parameter required; decimal, octal or hexo
int*
parameter requiredu
unsigned int*
parameter
requiredx
int*
parameter requiredc
char*
parameter required; white-space is
not skipped, and NUL
-termination is not performeds
char*
parameter required;
string is NUL
-terminatede,f,g
float*
parameter requiredp
void*
parameter requiredn
int*
parameter required[...]
char*
parameter required; string is NUL
-terminated[^...]
char*
parameter required; string is NUL
-terminated%
%
; no assignmentint (const char* format,
...);
scanf(f,
...)
is equivalent to fscanf(stdin,
f, ...)
int sscanf(char* s, const char* format,
...);
fscanf
,
but input read from string s
.int (FILE*
stream);
stream
,
or EOF
on end-of-file or error.char* (char* s, int n,
FILE*
stream);
stream
to s
, stopping when n
-1
characters copied, newline copied, end-of-file reached or error occurs.
If no error, s
is NUL
-terminated.
Returns NULL
on end-of-file or error, s
otherwise.(int c, FILE*
stream);
c
, to stream stream
.
Returns c
, or EOF
on error.char* (const char* s, FILE*
stream);
s
, to (output) stream stream. Returns
non-negative on success or EOF
on error.int (FILE*
stream);
fgetc
except that it may be a macro.int (void);
getc(stdin)
.char* (char* s);
stdin
into s
until newline encountered, end-of-file
reached, or error occurs. Does not copy newline. NUL
-terminates
s
. Returns s
, or NULL
on end-of-file or error. Should not be used because of the
potential for buffer overflow.int (int c, FILE*
stream);
fputc
except that it may be a macro.int (int c);
putchar(c)
is equivalent to putc(c,
stdout)
.int (const char* s);
s
(excluding terminating NUL
)
and a newline to stdout
.
Returns non-negative on success, EOF
on error.int (int c, FILE*
stream);
c
(which must not be EOF
),
onto (input) stream stream
such that it will be
returned by the next read. Only one character of pushback is
guaranteed (for each stream). Returns c
, or EOF
on error.size_t
(void* ptr, size_t
size, size_t
nobj, FILE*
stream);
nobj
objects of size size
from stream stream
into ptr
and returns number of objects read. (feof
and ferror
can be used to check status.)size_t (const void* ptr,
size_t
size, size_t
nobj, FILE*
stream);
stream
, nobj
objects of size size
from array ptr
.
Returns number of objects written.int (FILE*
stream, long offset, int origin);
stream
and
clears end-of-file indicator. For a binary stream, file position is set
to offset
bytes from the position indicated by origin
:
beginning of file for SEEK_SET
,
current position for SEEK_CUR
,
or end of file for SEEK_END
.
Behaviour is similar for a text stream, but offset
must be zero or, for SEEK_SET
only, a value returned by ftell
.
Returns non-zero on error.long (FILE*
stream);
stream
,
or -1
on error.void (FILE*
stream);
fseek(stream,
0L, SEEK_SET); clearerr(stream)
.int (FILE*
stream, fpos_t*
ptr);
stream
in *ptr
. Returns non-zero on error.int (FILE*
stream, const fpos_t*
ptr);
*ptr
.
Returns non-zero on error.void (FILE*
stream);
stream
.int (FILE*
stream);
stream
.int (FILE*
stream);
stream
.void (const char* s);
s
(if non-null) and strerror(errno)
to standard error as would:fprintf(stderr, "%s: %s\n", (s != NULL ? s : ""), strerror(errno))
[Contents]
status
argument to exit
indicating failure.
status
argument to exit
indicating success.
rand()
.
div()
.
Structure having members:int quot;
int rem;
ldiv()
.
Structure having members:long quot;
long rem;
sizeof
operator.int abs(int n);
long labs(long n);
n
.div_t
(int num, int denom);
ldiv_t
(long num, long denom);
num/denom
.double atof(const char* s);
strtod(s,
(char**)NULL)
except that errno
is not necessarily set on conversion error.int atoi(const char* s);
(int)strtol(s,
(char**)NULL,
10)
except that errno
is not necessarily set on conversion error.long atol(const char* s);
strtol(s,
(char**)NULL,
10)
except that errno
is not necessarily set on conversion error.double (const char* s,
char** endp);
s
to type double
. If endp
non-null,
stores pointer to unconverted suffix in *endp
.
On overflow, sets errno
to ERANGE
and returns HUGE_VAL
with the appropriate sign; on underflow, sets errno
to ERANGE
and returns zero; otherwise returns converted value.long (const char* s,
char** endp, int base);
s
to type long
. If endp
non-nu
ll, stores pointer to unconverted suffix in *endp
.
If base
between 2 and 36, that base used for
conversion; if zero, leading (after any sign) 0X
or 0x
implies hexadecimal, leading 0
(after any sign) implies
octal, otherwise decimal assumed. Leading 0X
or 0x
permitted for base hexadecimal. On overflow, sets errno
to ERANGE
and returns LONG_MAX
or LONG_MIN
(as appropriate for sign); otherwise returns converted value.unsigned long (const char* s,
char** endp, int base);
strtol
except result is unsigned long
and value on overflow is ULONG_MAX
.void* calloc(size_t
nobj, size_t
size);
nobj
objects each of size size
,
or NULL
on error.void* malloc(size_t
size);
NULL
on error.void* realloc(void* p, size_t size);
size
,
initialised, to minimum of old and new sizes, to existing contents of p
(if non-null), or NULL
on error. On success, old object deallocated, otherwise unchanged.void free(void* p);
p
non-null, deallocates space to which it
points.void ();
raise(SIGABRT)
.void (int status);
atexit
are called (in reverse order to that in which installed), open files
are flushed, open streams are closed and control is returned to
environment. status
is returned to environment
in implementation-dependent manner. Zero or EXIT_SUCCESS
indicates successful termination and EXIT_FAILURE
indicates unsuccessful termination. Implementations may define other
values.int (void (*fcm)(void));
fcn
to be called when program
terminates normally (or when main
returns). Returns
non-zero on failure.int system(const char* s);
s
is not NULL
,
passes s
to environment for execution, and
returns status reported by command processor; if s
is NULL
,
non-zero returned if environment has a command processor.char* getenv(const char* name);
name
from implementation's environment, or NULL
if no such string exists.void* bsearch(const void* key, const void* base,
size_t
n, size_t
size, int (*cmp)(const void* keyval, const void*
datum));
base
(of n
objects each of size size
) for item matching key
according to comparison function cmp
. cmp
must return negative value if first argument is less than second, zero
if equal and positive if greater. Items of base
are assumed to be in ascending order (according to cmp
).
Returns a pointer to an item matching key
, or NULL
if none found.void qsort(void* base, size_t
n, size_t
size, int (*cmp)(const void*, const void*));
base
(of
n
objects each of size size
)
according to comparison function cmp
. cmp
must return negative value if first argument is less than second, zero
if equal and positive if greater.int (void);
0
to RAND_MAX
.void srand(unsigned int seed);
seed
as seed for new sequence of
pseudo-random numbers. Initial seed is 1
.[Contents]
sizeof
operator.char* strcpy(char* s, const char* ct);
ct
to s
including terminating NUL
and returns s
.char* strncpy(char* s, const char* ct, size_t
n);
n
characters of ct
to s
. Pads with NUL
characters if
ct
is of length less than n
.
Note that this may leave s
without NUL
-termination.
Return s
.char* strcat(char* s, const char* ct);
ct
to s
and return s
.char* strncat(char* s, const char* ct, size_t
n);
n
characters of ct
to s
. NUL
-terminates s
and return it.int strcmp(const char* cs, const char* ct);
cs
with ct
,
returning negative value if cs<ct
,
zero if cs==ct
, positive value if cs>ct
.int strncmp(const char* cs, const char* ct,
size_t
n);
n
characters of
cs
and ct
, returning
negative value if cs<ct
, zero if cs==ct
,
positive value if cs>ct
.int strcoll(const char* cs, const char* ct);
cs
with ct
according to locale, returning negative value if cs<ct
,
zero if cs==ct
, positive value if cs>ct
.char* strchr(const char* cs, int c);
c
in cs
,
or NULL
if not found.char* strrchr(const char* cs, int c);
c
in cs
,
or NULL
if not found.size_t
strspn(const char* cs, const char* ct);
cs
which
consists of characters which are in ct
.size_t
strcspn(const char* cs, const char* ct);
cs
which
consists of characters which are not in ct
.char* strpbrk(const char* cs, const char* ct);
cs
of
any character of ct
, or NULL
if none is found.char* strstr(const char* cs, const char* ct);
ct
within cs
, or NULL
if none is found.size_t
strlen(const char* cs);
cs
.
n
.char* strtok(char* s, const char* t);
s
for next token delimited by any
character from ct
. Non-NULL
s
indicates the first call of a sequence. If a
token is found, it is NUL
-terminated and returned,
otherwise NULL
is returned. ct
need not be identical for each
call in a sequence.size_t
strxfrm(char* s, const char* ct, size_t
n);
s
no more than n
characters (including terminating NUL
) of a string
produced from ct
according to a locale-specific
transformation. Returns length of entire transformed string.void* (void* s, const
void* ct, size_t
n);
n
characters from ct
to s
and returns s
. s
may be corrupted if objects overlap.void* memmove(void* s, const void* ct, size_t
n);
n
characters from ct
to s
and returns s
. s
will not be corrupted if objects overlap.int memcmp(const void* cs, const void* ct,
size_t
n);
n
characters of
cs
and ct
, returning
negative value if cs<ct
, zero if cs==ct
,
positive value if cs>ct
.void* memchr(const void* cs, int c, size_t
n);
c
in
first n
characters of cs
,
or NULL
if not found.void* memset(void* s, int c, size_t
n);
n
characters of s
by c
and returns s
.[Contents]
CLOCKS_PER_SEC
clock_t
units per second.
int tm_sec;
int tm_min;
int tm_hour;
int tm_mday;
int tm_mon;
int tm_year;
int tm_wday;
int tm_yday;
int tm_isdst;
clock_t
clock(void);
-1
if
not available.time_t
time(time_t*
tp);
-1
if not available.
If tp
is non-NULL
,
return value is also assigned to *tp
.double difftime(time_t
time2, time_t
time1);
time2
and time1
.time_t
mktime(struct
tm* tp);
*tp
to fall
withing normal ranges. Returns the corresponding calendar time, or -1
if it cannot be represented.char* (const struct
tm* tp);
Sun Jan 3 13:08:42 1988\n\0
char* ctime(const time_t*
tp);
tp
converted to local time. Equivalent to:asctime(localtime(tp))
struct
tm* gmtime(const time_t*
tp);
*tp
converted to
Coordinated Universal Time, or NULL
if not available.struct
tm* (const time_t*
tp);
*tp
converted into
local
time.size_t strftime(char* s, size_t smax,
const char* fmt, const struct
tm* tp);
*tp
into s
according to fmt
. Places no more than smax
characters into s
, and returns number of
characters produced (excluding terminating NUL
), or 0
if greater than smax
. Formatting conversions (%c
)
are:A
a
B
b
c
d
01
-31
]H
00
-23
]I
01
-12
]j
001
-366
]M
00
-59
]m
01
-12
]p
AM
" or "PM
"S
00
-61
]U
00
-53
]W
00
-53
]w
0
) [0
-6
]X
x
Y
y
00
-99
]Z
%
%
Local time may differ from calendar time because of time zone.