MAN=		pkg_printf.3

.\" Copyright (c) 1990, 1991, 1993

.\"     The Regents of the University of California.  All rights reserved.

.\" Copyright (c) 2013 Matthew Seaman <matthew@FreeBSD.org>

.\"

.\" This code is derived from software contributed to Berkeley by

.\" Chris Torek and the American National Standards Committee X3,

.\" on Information Processing Systems.

.\"

.\" Redistribution and use in source and binary forms, with or without

.\" modification, are permitted provided that the following conditions

.\" are met:

.\" 1. Redistributions of source code must retain the above copyright

.\"    notice, this list of conditions and the following disclaimer.

.\" 2. Redistributions in binary form must reproduce the above copyright

.\"    notice, this list of conditions and the following disclaimer in the

.\"    documentation and/or other materials provided with the distribution.

.\" 4. Neither the name of the University nor the names of its contributors

.\"    may be used to endorse or promote products derived from this software

.\"    without specific prior written permission.

.\"

.\" THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND

.\" ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE

.\" IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE

.\" ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE

.\" FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL

.\" DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS

.\" OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)

.\" HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT

.\" LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY

.\" OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF

.\" SUCH DAMAGE.

.\"

.Dd January 5, 2013

.Dt PKG_PRINTF 3

.Os

.Sh NAME

.Nm pkg_printf , pkg_fprintf , pkg_dprintf , pkg_snprintf , pkg_asprintf ,

.Nm pkg_sbuf_printf , pkg_sbuf_vprintf

.Nd formatted output of package data

.Sh LIBRARY

.Lb libpkg

.Sh SYNOPSIS

.In pkg.h

.Ft int

.Fn pkg_printf "const char * restrict format" ...

.Ft int

.Fn pkg_fprintf "FILE * restrict stream" "const char * restrict format" ...

.Ft int

.Fn pkg_dprintf "int fd" "const char * restrict format" ...

.Ft int

.Fn pkg_snprintf "char * restrict str" "size_t size" "const char * restrict format" ...

.Ft int

.Fn pkg_asprintf "char **ret" "const char * restrict format" ...

.Ft struct sbuf *

.Fn pkg_sbuf_printf "struct sbuf * restrict sbuf" "const char * restrict format" ...

.Ft struct sbuf *

.Fn pkg_sbuf_vprintf "struct sbuf * restrict sbuf" "const char * restrict format" "va_list ap"

.Sh DESCRIPTION

The

.Fn pkg_printf

family of functions produces output of package data according to a

.Fa format

as described below, analogously to the similarly named

.Xr printf 3

family of functions.

The

.Fn pkg_printf

function

writes output to

.Dv stdout ,

the standard output stream;

.Fn pkg_fprintf

writes output to the given output

.Fa stream ;

.Fn pkg_dprintf

writes output to the given file descriptor;

.Fn pkg_snprintf ,

writes to the character string

.Fa str ;

.Fn pkg_asprintf

dynamically allocates a new string with

.Xr malloc 3 ;

.Fn pkg_sbuf_printf

and

.Fn pkg_sbuf_vprint

write to the given sbuf structure.

.Pp

These functions write the output under the control of a

.Fa format

string that specifies how subsequent arguments

(or arguments accessed via the variable-length argument facilities of

.Xr stdarg 3 )

are converted for output.

.Pp

These functions return the number of characters printed

(not including the trailing

.Ql \e0

used to end output to strings) or a negative value if an output error occurs,

except for

.Fn pkg_snprintf

which returns the number of characters that would have been printed if the

.Fa size

were unlimited

(again, not including the final

.Ql \e0 ) 

and the two functions

.Fn pkg_sbuf_printf

and

.Fn pkg_sbuf_vprintf

which return the given sbuf pointer, or

.Dv NULL

in the case of errors. 

.Pp

The

.Fn pkg_asprintf

function

sets

.Fa *ret

to be a pointer to a buffer sufficiently large to hold the formatted string.

This pointer should be passed to

.Xr free 3

to release the allocated storage when it is no longer needed.

If sufficient space cannot be allocated,

.Fn pkg_asprintf

will return \-1 and set

.Fa ret

to be a

.Dv NULL

pointer.

.Pp

The

.Fn pkg_snprintf

function

will write at most

.Fa size Ns \-1

of the characters printed into the output string

(the

.Fa size Ns 'th

character then gets the terminating

.Ql \e0 ) ;

if the return value is greater than or equal to the

.Fa size

argument, the string was too short

and some of the printed characters were discarded.

The output is always null-terminated.

.Pp

The format string is composed of zero or more directives:

ordinary

.\" multibyte

characters (not

.Cm % ) ,

which are copied unchanged to the output stream;

and conversion specifications, each of which results

in fetching zero or more subsequent arguments.

Each conversion specification is introduced by

the

.Cm %

character.

The arguments must correspond properly (after type promotion)

with the conversion specifier.

After the

.Cm % ,

the following appear in sequence:

.Bl -bullet

.It

Zero or more of the following flags:

.Bl -tag -width ".So \  Sc (space)"

.It Sq Cm ?

The value should be converted to the

.Dq first alternate form .

For integer valued conversions

.Cm ( I s t )

this is a

.Dq humanized

form as a floating point value scaled to the range 0 \- 1000

followed by a SI power-of-10 scale factor.

.It Sq Cm #

The value should be converted to the

.Dq second alternate form .

For integer valued conversions

.Cm ( I s t )

this is a

.Dq humanized

form as a floating point value scaled to the range 0 \- 1024

followed by a SI power-of-2 scale factor.

For

.Cm c , d , i , n , p , s ,

and

.Cm u

conversions, this option has no effect.

For

.Cm o

conversions, the precision of the number is increased to force the first

character of the output string to a zero.

For

.Cm x

and

.Cm X

conversions, a non-zero result has the string

.Ql 0x

(Or

.Ql 0X

for

.Cm X

conversions) prepended to it.

For

.Cm a , A , e , E , f , F , g ,

and

.Cm 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

.Cm g

and

.Cm G

conversions, trailing zeros are not removed from the result as they

would otherwise be.

.It So Cm 0 Sc (zero)

Zero padding.

For all conversions except

.Cm n ,

the converted value is padded on the left with zeros rather than blanks.

If a precision is given with a numeric conversion

.Cm ( d , i , o , u , i , x ,

and

.Cm X ) ,

the

.Cm 0

flag is ignored.

.It Sq Cm \-

A negative field width flag;

the converted value is to be left adjusted on the field boundary.

Except for

.Cm n

conversions, the converted value is padded on the right with blanks,

rather than on the left with blanks or zeros.

A

.Cm \-

overrides a

.Cm 0

if both are given.

.It So "\ " Sc (space)

A blank should be left before a positive number

produced by a signed conversion

.Cm ( a , A , d , e , E , f , F , g , G ,

or

.Cm i ) .

.It Sq Cm +

A sign must always be placed before a

number produced by a signed conversion.

A

.Cm +

overrides a space if both are used.

.It Sq Cm '

Decimal conversions

.Cm ( d , u ,

or

.Cm i )

or the integral portion of a floating point conversion

.Cm ( f

or

.Cm F )

should be grouped and separated by thousands using

the non-monetary separator returned by

.Xr localeconv 3 .

.El

.It

An optional decimal digit string 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) to fill out

the field width.

.It

An optional precision, in the form of a period

.Cm \&.

followed by an

optional digit string.

If the digit string is omitted, the precision is taken as zero.

This gives the minimum number of digits to appear for

.Cm d , i , o , u , x ,

and

.Cm X

conversions, the number of digits to appear after the decimal-point for

.Cm a , A , e , E , f ,

and

.Cm F

conversions, the maximum number of significant digits for

.Cm g

and

.Cm G

conversions, or the maximum number of characters to be printed from a

string for

.Cm s

conversions.

.It

An optional length modifier, that specifies the size of the argument.

The following length modifiers are valid for the

.Cm d , i , n , o , u , x ,

or

.Cm X

conversion:

.Bl -column ".Cm q Em (deprecated)" ".Vt signed char" ".Vt unsigned long long" ".Vt long long *"

.It Sy Modifier Ta Cm d , i Ta Cm o , u , x , X Ta Cm n

.It Cm hh Ta Vt "signed char" Ta Vt "unsigned char" Ta Vt "signed char *"

.It Cm h Ta Vt short Ta Vt "unsigned short" Ta Vt "short *"

.It Cm l No (ell) Ta Vt long Ta Vt "unsigned long" Ta Vt "long *"

.It Cm ll No (ell ell) Ta Vt "long long" Ta Vt "unsigned long long" Ta Vt "long long *"

.It Cm j Ta Vt intmax_t Ta Vt uintmax_t Ta Vt "intmax_t *"

.It Cm t Ta Vt ptrdiff_t Ta (see note) Ta Vt "ptrdiff_t *"

.It Cm z Ta (see note) Ta Vt size_t Ta (see note)

.It Cm q Em (deprecated) Ta Vt quad_t Ta Vt u_quad_t Ta Vt "quad_t *"

.El

.Pp

Note:

the

.Cm t

modifier, when applied to a

.Cm o , u , x ,

or

.Cm X

conversion, indicates that the argument is of an unsigned type

equivalent in size to a

.Vt ptrdiff_t .

The

.Cm z

modifier, when applied to a

.Cm d

or

.Cm i

conversion, indicates that the argument is of a signed type equivalent in

size to a

.Vt size_t .

Similarly, when applied to an

.Cm n

conversion, it indicates that the argument is a pointer to a signed type

equivalent in size to a

.Vt size_t .

.Pp

The following length modifier is valid for the

.Cm a , A , e , E , f , F , g ,

or

.Cm G

conversion:

.Bl -column ".Sy Modifier" ".Cm a , A , e , E , f , F , g , G"

.It Sy Modifier Ta Cm a , A , e , E , f , F , g , G

.It Cm l No (ell) Ta Vt double

(ignored, same behavior as without it)

.It Cm L Ta Vt "long double"

.El

.Pp

The following length modifier is valid for the

.Cm c

or

.Cm s

conversion:

.Bl -column ".Sy Modifier" ".Vt wint_t" ".Vt wchar_t *"

.It Sy Modifier Ta Cm c Ta Cm s

.It Cm l No (ell) Ta Vt wint_t Ta Vt "wchar_t *"

.El

.It

A character that specifies the type of conversion to be applied.

.El

.Pp

A field width or precision, or both, may be indicated by

an asterisk

.Ql *

or an asterisk followed by one or more decimal digits and a

.Ql $

instead of a

digit string.

In this case, an

.Vt int

argument supplies the field width or precision.

A negative field width is treated as a left adjustment flag followed by a

positive field width; a negative precision is treated as though it were

missing.

If a single format directive mixes positional

.Pq Li nn$

and non-positional arguments, the results are undefined.

.Pp

The conversion specifiers and their meanings are:

.Bl -tag -width ".Cm diouxX"

.It Cm diouxX

The

.Vt int

(or appropriate variant) argument is converted to signed decimal

.Cm ( d

and

.Cm i ) ,

unsigned octal

.Pq Cm o ,

unsigned decimal

.Pq Cm u ,

or unsigned hexadecimal

.Cm ( x

and

.Cm X )

notation.

The letters

.Dq Li abcdef

are used for

.Cm x

conversions; the letters

.Dq Li ABCDEF

are used for

.Cm 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.

.It Cm DOU

The

.Vt "long int"

argument is converted to signed decimal, unsigned octal, or unsigned

decimal, as if the format had been

.Cm ld , lo ,

or

.Cm lu

respectively.

These conversion characters are deprecated, and will eventually disappear.

.It Cm eE

The

.Vt double

argument is rounded and converted in the style

.Sm off

.Oo \- Oc Ar d Li \&. Ar ddd Li e \(+- Ar dd

.Sm on

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

.Cm E

conversion uses the letter

.Ql E

(rather than

.Ql e )

to introduce the exponent.

The exponent always contains at least two digits; if the value is zero,

the exponent is 00.

.Pp

For

.Cm a , A , e , E , f , F , g ,

and

.Cm G

conversions, positive and negative infinity are represented as

.Li inf

and

.Li -inf

respectively when using the lowercase conversion character, and

.Li INF

and

.Li -INF

respectively when using the uppercase conversion character.

Similarly, NaN is represented as

.Li nan

when using the lowercase conversion, and

.Li NAN

when using the uppercase conversion.

.It Cm fF

The

.Vt double

argument is rounded and converted to decimal notation in the style

.Sm off

.Oo \- Oc Ar ddd Li \&. Ar ddd ,

.Sm on

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.

.It Cm gG

The

.Vt double

argument is converted in style

.Cm f

or

.Cm e

(or

.Cm F

or

.Cm E

for

.Cm G

conversions).

The precision specifies the number of significant digits.

If the precision is missing, 6 digits are given; if the precision is zero,

it is treated as 1.

Style

.Cm 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.

.It Cm aA

The

.Vt double

argument is rounded and converted to hexadecimal notation in the style

.Sm off

.Oo \- Oc Li 0x Ar h Li \&. Ar hhhp Oo \(+- Oc Ar d ,

.Sm on

where the number of digits after the hexadecimal-point character

is equal to the precision specification.

If the precision is missing, it is taken as enough to represent

the floating-point number exactly, and no rounding occurs.

If the precision is zero, no hexadecimal-point character appears.

The

.Cm p

is a literal character

.Ql p ,

and the exponent consists of a positive or negative sign

followed by a decimal number representing an exponent of 2.

The

.Cm A

conversion uses the prefix

.Dq Li 0X

(rather than

.Dq Li 0x ) ,

the letters

.Dq Li ABCDEF

(rather than

.Dq Li abcdef )

to represent the hex digits, and the letter

.Ql P

(rather than

.Ql p )

to separate the mantissa and exponent.

.Pp

Note that there may be multiple valid ways to represent floating-point

numbers in this hexadecimal format.

For example,

.Li 0x1.92p+1 , 0x3.24p+0 , 0x6.48p-1 ,

and

.Li 0xc.9p-2

are all equivalent.

.Fx 8.0

and later always prints finite non-zero numbers using

.Ql 1

as the digit before the hexadecimal point.

Zeroes are always represented with a mantissa of 0 (preceded by a

.Ql -

if appropriate) and an exponent of

.Li +0 .

.It Cm C

Treated as

.Cm c

with the

.Cm l

(ell) modifier.

.It Cm c

The

.Vt int

argument is converted to an

.Vt "unsigned char" ,

and the resulting character is written.

.Pp

If the

.Cm l

(ell) modifier is used, the

.Vt wint_t

argument shall be converted to a

.Vt wchar_t ,

and the (potentially multi-byte) sequence representing the

single wide character is written, including any shift sequences.

If a shift sequence is used, the shift state is also restored

to the original state after the character.

.It Cm S

Treated as

.Cm s

with the

.Cm l

(ell) modifier.

.It Cm s

The

.Vt "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

.Dv NUL

character;

if a precision is specified, no more than the number specified are

written.

If a precision is given, no null character

need be present; if the precision is not specified, or is greater than

the size of the array, the array must contain a terminating

.Dv NUL

character.

.Pp

If the

.Cm l

(ell) modifier is used, the

.Vt "wchar_t *"

argument is expected to be a pointer to an array of wide characters

(pointer to a wide string).

For each wide character in the string, the (potentially multi-byte)

sequence representing the

wide character is written, including any shift sequences.

If any shift sequence is used, the shift state is also restored

to the original state after the string.

Wide characters from the array are written up to (but not including)

a terminating wide

.Dv NUL

character;

if a precision is specified, no more than the number of bytes specified are

written (including shift sequences).

Partial characters are never written.

If a precision is given, no null character

need be present; if the precision is not specified, or is greater than

the number of bytes required to render the multibyte representation of

the string, the array must contain a terminating wide

.Dv NUL

character.

.It Cm p

The

.Vt "void *"

pointer argument is printed in hexadecimal (as if by

.Ql %#x

or

.Ql %#lx ) .

.It Cm n

The number of characters written so far is stored into the

integer indicated by the

.Vt "int *"

(or variant) pointer argument.

No argument is converted.

.It Cm %

A

.Ql %

is written.

No argument is converted.

The complete conversion specification

is

.Ql %% .

.El

.Pp

The decimal point

character is defined in the program's locale (category

.Dv LC_NUMERIC ) .

.Pp

In no case does a non-existent or small field width cause truncation of

a numeric field; if the result of a conversion is wider than the field

width, the

field is expanded to contain the conversion result.

.Sh EXAMPLES

To print a date and time in the form

.Dq Li "Sunday, July 3, 10:02" ,

where

.Fa weekday

and

.Fa month

are pointers to strings:

.Bd -literal -offset indent

#include <stdio.h>

fprintf(stdout, "%s, %s %d, %.2d:%.2d\en",

	weekday, month, day, hour, min);

.Ed

.Pp

To print \*(Pi

to five decimal places:

.Bd -literal -offset indent

#include <math.h>

#include <stdio.h>

fprintf(stdout, "pi = %.5f\en", 4 * atan(1.0));

.Ed

.Pp

To allocate a 128 byte string and print into it:

.Bd -literal -offset indent

#include <stdio.h>

#include <stdlib.h>

#include <stdarg.h>

char *newfmt(const char *fmt, ...)

{

	char *p;

	va_list ap;

	if ((p = malloc(128)) == NULL)

		return (NULL);

	va_start(ap, fmt);

	(void) vsnprintf(p, 128, fmt, ap);

	va_end(ap);

	return (p);

}

.Ed

.Sh COMPATIBILITY

Many application writers used the name

.Va dprintf

before the

.Fn dprintf

function was introduced in

.St -p1003.1 ,

so a prototype is not provided by default in order to avoid

compatibility problems.

Applications that wish to use the

.Fn dprintf

function described herein should either request a strict

.St -p1003.1-2008

environment by defining the macro

.Dv _POSIX_C_SOURCE

to the value 200809 or greater, or by defining the macro

.Dv _WITH_DPRINTF ,

prior to the inclusion of

.In stdio.h .

For compatibility with GNU libc, defining either

.Dv _BSD_SOURCE

or

.Dv _GNU_SOURCE

prior to the inclusion of

.In stdio.h

will also make

.Fn dprintf

available.

.Pp

The conversion formats

.Cm \&%D , \&%O ,

and

.Cm \&%U

are not standard and

are provided only for backward compatibility.

The effect of padding the

.Cm %p

format with zeros (either by the

.Cm 0

flag or by specifying a precision), and the benign effect (i.e., none)

of the

.Cm #

flag on

.Cm %n

and

.Cm %p

conversions, as well as other

nonsensical combinations such as

.Cm %Ld ,

are not standard; such combinations

should be avoided.

.Sh ERRORS

In addition to the errors documented for the

.Xr write 2

system call, the

.Fn printf

family of functions may fail if:

.Bl -tag -width Er

.It Bq Er EILSEQ

An invalid wide character code was encountered.

.It Bq Er ENOMEM

Insufficient storage space is available.

.El

.Sh SEE ALSO

.Xr printf 1 ,

.Xr fmtcheck 3 ,

.Xr scanf 3 ,

.Xr setlocale 3 ,

.Xr wprintf 3

.Sh STANDARDS

Subject to the caveats noted in the

.Sx BUGS

section below, the

.Fn fprintf ,

.Fn printf ,

.Fn sprintf ,

.Fn vprintf ,

.Fn vfprintf ,

and

.Fn vsprintf

functions

conform to

.St -ansiC

and

.St -isoC-99 .

With the same reservation, the

.Fn snprintf

and

.Fn vsnprintf

functions conform to

.St -isoC-99 ,

while

.Fn dprintf

and

.Fn vdprintf

conform to

.St -p1003.1-2008 .

.Sh HISTORY

The functions

.Fn asprintf

and

.Fn vasprintf

first appeared in the

.Tn GNU C

library.

These were implemented by

.An Peter Wemm Aq peter@FreeBSD.org

in

.Fx 2.2 ,

but were later replaced with a different implementation

from

.Ox 2.3

by

.An Todd C. Miller Aq Todd.Miller@courtesan.com .

The

.Fn dprintf

and

.Fn vdprintf

functions were added in

.Fx 8.0 .

.Sh BUGS

The

.Nm

family of functions do not correctly handle multibyte characters in the

.Fa format

argument.

.Sh SECURITY CONSIDERATIONS

The

.Fn sprintf

and

.Fn vsprintf

functions are easily misused in a manner which enables malicious users

to arbitrarily change a running program's functionality through

a buffer overflow attack.

Because

.Fn sprintf

and

.Fn vsprintf

assume an infinitely long string,

callers must be careful not to overflow the actual space;

this is often hard to assure.

For safety, programmers should use the

.Fn snprintf

interface instead.

For example:

.Bd -literal

void

foo(const char *arbitrary_string, const char *and_another)

{

	char onstack[8];

#ifdef BAD

	/*

	 * This first sprintf is bad behavior.  Do not use sprintf!

	 */

	sprintf(onstack, "%s, %s", arbitrary_string, and_another);

#else

	/*

	 * The following two lines demonstrate better use of

	 * snprintf().

	 */

	snprintf(onstack, sizeof(onstack), "%s, %s", arbitrary_string,

	    and_another);

#endif

}

.Ed

.Pp

The

.Fn printf

and

.Fn sprintf

family of functions are also easily misused in a manner

allowing malicious users to arbitrarily change a running program's

functionality by either causing the program

to print potentially sensitive data

.Dq "left on the stack" ,

or causing it to generate a memory fault or bus error

by dereferencing an invalid pointer.

.Pp

.Cm %n

can be used to write arbitrary data to potentially carefully-selected

addresses.

Programmers are therefore strongly advised to never pass untrusted strings

as the

.Fa format

argument, as an attacker can put format specifiers in the string

to mangle your stack,

leading to a possible security hole.

This holds true even if the string was built using a function like

.Fn snprintf ,

as the resulting string may still contain user-supplied conversion specifiers

for later interpolation by

.Fn printf .

.Pp

Always use the proper secure idiom:

.Pp

.Dl "snprintf(buffer, sizeof(buffer), \*q%s\*q, string);"