man 3 printf

PRINTF(3) Linux Programmer's Manual PRINTF(3)






NAME
       printf, fprintf, sprintf, snprintf, vprintf, vfprintf, vsprintf, vsnprintf - formatted output conversion


SYNOPSIS
       #include <stdio.h>


       int printf(const char *format, ...);
       int fprintf(FILE *stream, const char *format, ...);
       int sprintf(char *str, const char *format, ...);
       int snprintf(char *str, size_t size, const char *format, ...);


       #include <stdarg.h>


       int vprintf(const char *format, va_list ap);
       int vfprintf(FILE *stream, const char *format, va_list ap);
       int vsprintf(char *str, const char *format, va_list ap);
       int vsnprintf(char *str, size_t size, const char *format, va_list ap);


   Feature Test Macro Requirements for glibc (see feature_test_macros(7)):


       snprintf(), vsnprintf():
  _BSD_SOURCE || _XOPEN_SOURCE >= 500 || _ISOC99_SOURCE || _POSIX_C_SOURCE >= 200112L;
  or cc -std=c99


DESCRIPTION
       The  functions  in  the printf()  family  produce  output  according  to a format as described below.  The functions printf() and
       vprintf() write output to stdout, the standard output stream; fprintf() and vfprintf() write output to the  given  output  stream;
       sprintf(), snprintf(), vsprintf() and vsnprintf() write to the character string str.


       The functions snprintf() and vsnprintf() write at most size bytes (including the terminating null byte ('\0')) to str.


       The  functions  vprintf(),  vfprintf(), vsprintf(),  vsnprintf()  are equivalent to the functions printf(), fprintf(), sprintf(),
       snprintf(), respectively, except that they are called with a va_list instead of a variable number of arguments. These functions
       do not call the va_end macro.  Because they invoke the va_arg macro, the value of ap is undefined after the call.  See stdarg(3).


       These  eight functions write the output under the control of a format string that specifies how subsequent arguments (or arguments
       accessed via the variable-length argument facilities of stdarg(3)) are converted for output.


       C99 and POSIX.1-2001 specify that the results are undefined if a call to sprintf(), snprintf(), vsprintf(), or  vsnprintf()  would
       cause copying to take place between objects that overlap (e.g., if the target string array and one of the supplied input arguments
       refer to the same buffer).  See NOTES.


   Return value
       Upon successful return, these functions return the number of characters printed (excluding the null byte used  to  end  output  to
       strings).


       The  functions  snprintf() and vsnprintf() do not write more than size bytes (including the terminating null byte ('\0')).  If the
       output was truncated due to this limit then the return value is the number of characters (excluding  the  terminating  null  byte)
       which  would have been written to the final string if enough space had been available.  Thus, a return value of size or more means
       that the output was truncated.  (See also below under NOTES.)


       If an output error is encountered, a negative value is returned.


   Format of the format string
       The format string is a character string, beginning and ending in its initial shift state, if any.   The format string  is  composed
       of  zero or more directives: ordinary characters (not %), which are copied unchanged to the output stream; and conversion specifi‐
       cations, each of which results in fetching zero or more subsequent arguments.  Each conversion specification is introduced by  the
       character  %, and ends with a conversion specifier.  In between there may be (in this order) zero or more flags, an optional mini‐
       mum field width, an optional precision and an optional length modifier.


       The arguments must correspond properly (after type promotion) with the conversion specifier.  By default, the arguments are  used
       in  the order given, where each '*' and each conversion specifier asks for the next argument (and it is an error if insufficiently
       many arguments are given).  One can also specify explicitly which argument is taken, at each place where an argument is required,
       by writing "%m$" instead of '%' and "*m$" instead of '*', where the decimal integer m denotes the position in the argument list of
       the desired argument, indexed starting from 1.  Thus,


  printf("%*d", width, num);


       and


  printf("%2$*1$d", width, num);


       are equivalent. The second style allows repeated references to the same argument.  The C99 standard does not  include  the  style
       using  '$',  which  comes  from the Single UNIX Specification.  If the style using '$' is used, it must be used throughout for all
       conversions taking an argument and all width and precision arguments, but it may be mixed with "%%" formats which do  not  consume
       an  argument.   There may be no gaps in the numbers of arguments specified using '$'; for example, if arguments 1 and 3 are speci‐
       fied, argument 2 must also be specified somewhere in the format string.


       For some numeric conversions a radix character ("decimal point") or thousands' grouping character is used.  The actual character
       used  depends  on  the  LC_NUMERIC part of the locale.  The POSIX locale uses '.' as radix character, and does not have a grouping
       character.  Thus,


      printf("%'.2f", 1234567.89);


       results in "1234567.89" in the POSIX locale, in "1234567,89" in the nl_NL locale, and in "1.234.567,89" in the da_DK locale.


   The flag characters
       The character % is followed by zero or more of the following flags:


        #      The value should be converted to an "alternate form".  For o conversions, the first character of the output string is  made
     zero (by prefixing a 0 if it was not zero already).  For x and X conversions, a nonzero result has the string "0x" (or "0X"
     for X conversions) prepended to it.  For a, A, e, E, f, F, g, and G conversions, the result will always contain  a  decimal
     point,  even  if no digits follow it (normally, a decimal point appears in the results of those conversions only if a digit
     follows).  For g and G conversions, trailing zeros are not removed from the result as they would otherwise be.   For  other
     conversions, the result is undefined.


       0      The  value  should  be  zero  padded.  For d, i, o, u, x, X, a, A, e, E, f, F, g, and G conversions, the converted value is
     padded on the left with zeros rather than blanks.  If the 0 and - flags both appear, the 0 flag is ignored.  If a precision
     is  given  with a numeric conversion (d, i, o, u, x, and X), the 0 flag is ignored.  For other conversions, the behavior is
     undefined.


        -      The converted value is to be left adjusted on the field boundary.  (The default is right justification.) Except for n con‐
     versions,  the converted value is padded on the right with blanks, rather than on the left with blanks or zeros. A - over‐
     rides a 0 if both are given.


       ' '    (a space) A blank should be left before a positive number (or empty string) produced by a signed conversion.


        +      A sign (+ or -) should always be placed before a number produced by a signed conversion. By default a sign  is  used  only
     for negative numbers.  A + overrides a space if both are used.


       The five flag characters above are defined in the C standard.  The SUSv2 specifies one further flag character.


       '      For  decimal conversion (i, d, u, f, F, g, G) the output is to be grouped with thousands' grouping characters if the locale
     information indicates any.  Note that many versions of gcc(1) cannot parse this option and will  issue  a  warning.   SUSv2
     does not include %'F.


       glibc 2.2 adds one further flag character.


       I      For  decimal  integer  conversion  (i,  d, u) the output uses the locale's alternative output digits, if any.  For example,
     since glibc 2.2.3 this will give Arabic-Indic digits in the Persian ("fa_IR") locale.


    The field width
       An optional decimal digit string (with nonzero first digit) specifying a minimum field width.  If the converted value  has  fewer
       characters than the field width, it will be padded with spaces on the left (or right, if the left-adjustment flag has been given).
       Instead of a decimal digit string one may write "*" or "*m$" (for some decimal integer m) to specify that the field width is given
       in  the next argument, or in the m-th argument, respectively, which must be of type int.  A negative field width is taken as a '-'
       flag followed by a positive field width.  In no case does a nonexistent or small field width cause truncation of a field;  if  the
       result of a conversion is wider than the field width, the field is expanded to contain the conversion result.


   The precision
       An  optional  precision,  in the form of a period ('.') followed by an optional decimal digit string.  Instead of a decimal digit
       string one may write "*" or "*m$" (for some decimal integer m) to specify that the precision is given in the next argument, or  in
       the  m-th argument, respectively, which must be of type int.  If the precision is given as just '.', or the precision is negative,
       the precision is taken to be zero.  This gives the minimum number of digits to appear for d, i, o, u, x, and  X conversions,  the
       number  of digits to appear after the radix character for a, A, e, E, f, and F conversions, the maximum number of significant dig‐
       its for g and G conversions, or the maximum number of characters to be printed from a string for s and S conversions.


   The length modifier
       Here, "integer conversion" stands for d, i, o, u, x, or X conversion.


       hh     A following integer conversion corresponds to a signed char or unsigned char argument, or a following n  conversion  corre‐
     sponds to a pointer to a signed char argument.


       h      A  following integer conversion corresponds to a short int or unsigned short int argument, or a following n conversion cor‐
     responds to a pointer to a short int argument.


       l      (ell) A following integer conversion corresponds to a long int or unsigned long int argument, or a following  n  conversion
     corresponds to a pointer to a long int argument, or a following c conversion corresponds to a wint_t argument, or a follow‐
     ing s conversion corresponds to a pointer to wchar_t argument.


       ll     (ell-ell).  A following integer conversion corresponds to a long long int or unsigned long long int argument, or a  follow‐
     ing n conversion corresponds to a pointer to a long long int argument.


       L      A  following  a, A, e, E, f, F, g, or G conversion corresponds to a long double argument.  (C99 allows %LF, but SUSv2 does
     not.)


       q      ("quad". 4.4BSD and Linux libc5 only.  Don't use.)  This is a synonym for ll.


       j      A following integer conversion corresponds to an intmax_t or uintmax_t argument.


       z      A following integer conversion corresponds to a size_t or ssize_t argument.  (Linux libc5 has Z with this  meaning.   Don't
     use it.)


       t      A following integer conversion corresponds to a ptrdiff_t argument.


       The  SUSv2 only knows about the length modifiers h (in hd, hi, ho, hx, hX, hn) and l (in ld, li, lo, lx, lX, ln, lc, ls) and L (in
       Le, LE, Lf, Lg, LG).


   The conversion specifier
       A character that specifies the type of conversion to be applied.  The conversion specifiers and their meanings are:


        d, i   The int argument is converted to signed decimal notation.  The precision, if any, gives the minimum number of  digits  that
     must  appear;  if the converted value requires fewer digits, it is padded on the left with zeros.  The default precision is
     1.  When 0 is printed with an explicit precision 0, the output is empty.


        o, u, x, X
     The unsigned int argument is converted to unsigned octal (o), unsigned decimal (u), or unsigned hexadecimal (x and X) nota‐
     tion.   The  letters  abcdef  are used for x conversions; the letters ABCDEF are used for X conversions. The precision, if
     any, gives the minimum number of digits that must appear; if the converted value requires fewer digits, it is padded on the
     left with zeros. The default precision is 1.  When 0 is printed with an explicit precision 0, the output is empty.


       e, E   The  double  argument  is rounded and converted in the style [-]d.ddde±dd where there is one digit before the decimal-point
     character and the number of digits after it is equal to the precision; if the precision is missing, it is taken  as  6;  if
     the  precision is zero, no decimal-point character appears.  An E conversion uses the letter E (rather than e) to introduce
     the exponent.  The exponent always contains at least two digits; if the value is zero, the exponent is 00.


       f, F   The double argument is rounded and converted to decimal notation in the style [-]ddd.ddd, where the number of digits  after
     the decimal-point character is equal to the precision specification.  If the precision is missing, it is taken as 6; if the
     precision is explicitly zero, no decimal-point character appears.  If a decimal point appears, at least one  digit  appears
     before it.


     (The SUSv2 does not know about F and says that character string representations for infinity and NaN may be made available.
     The C99 standard specifies "[-]inf" or "[-]infinity" for infinity, and a string starting with "nan" for NaN, in the case of
     f conversion, and "[-]INF" or "[-]INFINITY" or "NAN*" in the case of F conversion.)


       g, G   The double argument is converted in style f or e (or F or E for G conversions).  The precision specifies the number of sig‐
     nificant digits. If the precision is missing, 6 digits are given; if the precision is zero, it is treated as 1. Style e
     is  used if the exponent from its conversion is less than -4 or greater than or equal to the precision.  Trailing zeros are
     removed from the fractional part of the result; a decimal point appears only if it is followed by at least one digit.


       a, A   (C99; not in SUSv2) For a conversion, the double argument is converted to hexadecimal notation (using the  letters  abcdef)
     in  the  style  [-]0xh.hhhhp±d;  for  A conversion the prefix 0X, the letters ABCDEF, and the exponent separator P is used.
     There is one hexadecimal digit before the decimal point, and the number of digits after it is equal to the precision.   The
     default  precision suffices for an exact representation of the value if an exact representation in base 2 exists and other‐
     wise is sufficiently large to distinguish values of type double. The digit before the decimal  point  is  unspecified  for
     nonnormalized numbers, and nonzero but otherwise unspecified for normalized numbers.


       c      If no l modifier is present, the int argument is converted to an unsigned char, and the resulting character is written.  If
     an l modifier is present, the wint_t (wide character) argument is converted to a multibyte sequence by a call to the  wcr‐
     tomb(3) function, with a conversion state starting in the initial state, and the resulting multibyte string is written.


       s      If  no  l modifier is present: The const char * argument is expected to be a pointer to an array of character type (pointer
     to a string).  Characters from the array are written up to (but not including) a terminating null byte ('\0'); if a  preci‐
     sion  is specified, no more than the number specified are written.  If a precision is given, no null byte need be present;
     if the precision is not specified, or is greater than the size of the array, the array  must  contain  a terminating  null
     byte.


     If an l modifier is present: The const wchar_t * argument is expected to be a pointer to an array of wide characters.  Wide
     characters from the array are converted to multibyte characters (each by a call to the wcrtomb(3) function, with a  conver‐
     sion  state  starting  in  the  initial state before the first wide character), up to and including a terminating null wide
     character.  The resulting multibyte characters are written up to (but not including) the terminating null byte.  If a  pre‐
     cision  is specified, no more bytes than the number specified are written, but no partial multibyte characters are written.
     Note that the precision determines the number of bytes written, not the number of wide characters or screen positions.  The
     array  must  contain  a  terminating null wide character, unless a precision is given and it is so small that the number of
     bytes written exceeds it before the end of the array is reached.


       C      (Not in C99, but in SUSv2.)  Synonym for lc.  Don't use.


       S      (Not in C99, but in SUSv2.)  Synonym for ls.  Don't use.


        p      The void * pointer argument is printed in hexadecimal (as if by %#x or %#lx).


       n      The number of characters written so far is stored into the integer indicated by the int * (or  variant)  pointer argument.
     No argument is converted.


       m      (Glibc extension.)  Print output of strerror(errno).  No argument is required.


       %      A '%' is written.  No argument is converted.  The complete conversion specification is '%%'.


CONFORMING TO
       The  fprintf(), printf(),  sprintf(), vprintf(), vfprintf(), and vsprintf() functions conform to C89 and C99.  The snprintf() and
       vsnprintf() functions conform to C99.


       Concerning the return value of snprintf(), SUSv2 and C99 contradict each other: when snprintf() is called with size=0  then  SUSv2
       stipulates  an  unspecified return value less than 1, while C99 allows str to be NULL in this case, and gives the return value (as
       always) as the number of characters that would have been written in case the output string has been large enough.


       Linux libc4 knows about the five C standard flags.  It knows about the length modifiers h, l, L, and the conversions c, d,  e,  E,
       f,  F, g, G, i, n, o, p, s, u, x, and X, where F is a synonym for f.  Additionally, it accepts D, O, and U as synonyms for ld, lo,
       and lu. (This is bad, and caused serious bugs later, when support for %D disappeared.) No locale-dependent radix  character,  no
       thousands' separator, no NaN or infinity, no "%m$" and "*m$".


       Linux libc5 knows about the five C standard flags and the ' flag, locale, "%m$" and "*m$".  It knows about the length modifiers h,
       l, L, Z, and q, but accepts L and q both for long double and for long long int (this is a bug). It no longer recognizes F, D,  O,
       and U, but adds the conversion character m, which outputs strerror(errno).


       glibc 2.0 adds conversion characters C and S.


       glibc 2.1 adds length modifiers hh, j, t, and z and conversion characters a and A.


       glibc 2.2 adds the conversion character F with C99 semantics, and the flag character I.


NOTES
       Some programs imprudently rely on code such as the following


  sprintf(buf, "%s some further text", buf);


       to  append  text  to buf.  However, the standards explicitly note that the results are undefined if source and destination buffers
       overlap when calling sprintf(), snprintf(), vsprintf(), and vsnprintf().  Depending on the version of gcc(1) used,  and the  com‐
       piler options employed, calls such as the above will not produce the expected results.


       The  glibc  implementation of the functions snprintf() and vsnprintf() conforms to the C99 standard, that is, behaves as described
       above, since glibc version 2.1. Until glibc 2.0.6 they would return -1 when the output was truncated.


BUGS
       Because sprintf() and vsprintf() assume an arbitrarily long string, callers must be careful not to overflow the actual space; this
       is  often  impossible  to assure.  Note that the length of the strings produced is locale-dependent and difficult to predict.  Use
       snprintf() and vsnprintf() instead (or asprintf(3) and vasprintf(3)).


       Linux libc4.[45] does not have a snprintf(), but provides a libbsd that contains an snprintf() equivalent to sprintf(), that  is,
       one that ignores the size argument.  Thus, the use of snprintf() with early libc4 leads to serious security problems.


       Code  such as printf(foo); often indicates a bug, since foo may contain a % character.  If foo comes from untrusted user input, it
       may contain %n, causing the printf() call to write to memory and creating a security hole.


EXAMPLE
       To print pi to five decimal places:


  #include <math.h>
  #include <stdio.h>
  fprintf(stdout, "pi = %.5f\n", 4 * atan(1.0));


       To print a date and time in the form "Sunday, July 3, 10:02", where weekday and month are pointers to strings:


  #include <stdio.h>
  fprintf(stdout, "%s, %s %d, %.2d:%.2d\n",
  weekday, month, day, hour, min);


       Many countries use the day-month-year order.  Hence, an internationalized version must be able to print the arguments in an  order
       specified by the format:


  #include <stdio.h>
  fprintf(stdout, format,
  weekday, month, day, hour, min);


       where format depends on locale, and may permute the arguments.  With the value:


  "%1$s, %3$d. %2$s, %4$d:%5$.2d\n"


       one might obtain "Sonntag, 3. Juli, 10:02".


       To allocate a sufficiently large string and print into it (code correct for both glibc 2.0 and glibc 2.1):


       #include <stdio.h>
       #include <stdlib.h>
       #include <stdarg.h>


       char *
       make_message(const char *fmt, ...)
       {
  int n;
  int size = 100;     /* Guess we need no more than 100 bytes. */
  char *p, *np;
  va_list ap;


  if ((p = malloc(size)) == NULL)
      return NULL;


  while (1) {


      /* Try to print in the allocated space. */


      va_start(ap, fmt);
      n = vsnprintf(p, size, fmt, ap);
      va_end(ap);


      /* If that worked, return the string. */


      if (n > -1 && n < size)
  return p;


      /* Else try again with more space. */


      if (n > -1)    /* glibc 2.1 */
  size = n+1; /* precisely what is needed */
      else      /* glibc 2.0 */
  size *= 2;  /* twice the old size */


      if ((np = realloc (p, size)) == NULL) {
  free(p);
  return NULL;
      } else {
  p = np;
      }
  }
       }


SEE ALSO
       printf(1), asprintf(3), dprintf(3), scanf(3), setlocale(3), wcrtomb(3), wprintf(3), locale(5)


COLOPHON
       This  page  is part of release 3.35 of the Linux man-pages project.  A description of the project, and information about reporting
       bugs, can be found at http://man7.org/linux/man-pages/.






GNU 2011-09-28 PRINTF(3)