Formatted Input:Scan Formats, Scan Functions, Scan Conversion Specifiers

<< Indefinite Integrals of Logarithmic Functions:Combinations of Logarithms and Polynomials, The Logarithm

Formatted Output:Print Formats, Print Functions, Print Conversion Specifiers >>

Chapter 7. Formatted Input

Scan Formats (page 25) · Scan Functions (page 25) · Scan Conversion Specifiers

(page 26)

Several library functions help you convert data values from text sequences that are

generally readable by people to encoded internal representations. You provide a

format string (page 31) as the value of the format argument to each of these

functions, hence the term formatted input. The functions fall into two categories:

The byte scan functions (declared in <stdio.h>) convert sequences of type char to

internal representations, and help you scan such sequences that you read: fscanf,

scanf, and sscanf. For these functions, a format string is a multibyte string that

begins and ends in the initial shift state (page 12).

The wide scan functions (declared in <wchar.h> and hence added with

Amendment 1) convert sequences of type wchar_t, to internal representations, and

help you scan such sequences that you read: fwscanf, wscanf and swscanf. For

these functions, a format string is a wide-character string. In the descriptions that

follow, a wide character wc from a format string or a stream is compared to a

specific (byte) character c as if by evaluating the expression wctob(wc) == c.

Scan Formats

A format string has the same general syntax (page 31) for the scan functions as for

the print functions (page 31): zero or more conversion specifications (page 31),

interspersed with literal text and white space (page 31). For the scan functions,

however, a conversion specification is one of the scan conversion specifications

(page 25) described below.

A scan function scans the format string once from beginning to end to determine

what conversions to perform. Every scan function accepts a varying number of

arguments, either directly or under control of an argument of type va_list. Some

scan conversion specifications in the format string use the next argument in the

list. A scan function uses each successive argument no more than once. Trailing

arguments can be left unused.

In the description that follows, the integer conversions (page 32) and floating-point

conversions (page 32) are the same as for the print functions (page 31).

Scan Functions

For the scan functions, literal text in a format string must match the next characters

to scan in the input text. White space in a format string must match the longest

possible sequence of the next zero or more white-space characters in the input.

Except for the scan conversion specifier (page 26) %n (which consumes no input),

each scan conversion specification determines a pattern that one or more of the

next characters in the input must match. And except for the scan conversion

specifiers c, n, and [, every match begins by skipping any white space characters in

the input.

A scan function returns when:

v it reaches the terminating null in the format string

v it cannot obtain additional input characters to scan (input failure)

v a conversion fails (matching failure)

A scan function returns EOF if an input failure occurs before any conversion.

Otherwise it returns the number of converted values stored. If one or more

characters form a valid prefix but the conversion fails, the valid prefix is consumed

before the scan function returns. Thus:

scanf("%i", &i)

consumes 0X from field 0XZ

scanf("%f", &f)

consumes 3.2E from field 3.2EZ

A scan conversion specification typically converts the matched input characters to

a corresponding encoded value. The next argument value must be the address of

an object. The conversion converts the encoded representation (as necessary) and

stores its value in the object. A scan conversion specification has the format shown

in the diagram.

Following the percent character (%) in the format string, you can write an asterisk

(*) to indicate that the conversion should not store the converted value in an

object.

Following any *, you can write a nonzero field width that specifies the maximum

number of input characters to match for the conversion (not counting any white

space that the pattern can first skip).

Scan Conversion Specifiers

Following any field width (page 26), you must write a one-character scan

conversion specifier, either a one-character code or a scan set (page 28), possibly

preceded by a one-character qualifier. Each combination determines the type

required of the next argument (if any) and how the scan functions interpret the

text sequence and converts it to an encoded value. The integer (page 32) and

floating-point conversions (page 32) also determine what base to assume for the

text representation. (The base is the base argument to the functions strtol and

strtoul.) The following table lists all defined combinations and their properties.

Conversion

Argument

Conversion

Specifier

Type

Function Base

char x[]

%lc

wchar_t x[]

int *x

strtol

%hd

short *x

strtol

%ld

long *x

strtol

float *x

strtod

%le

double *x

strtod

%Le

long double *x

strtod

float *x

strtod

%lE

double *x

strtod

%LE

long double *x

strtod

float *x

strtod

Standard C++ Library

%lf

double *x

strtod

%Lf

long double *x

strtod

float *x

strtod

%lg

double *x

strtod

%Lg

long double *x

strtod

float *x

strtod

%lG

double *x

strtod

%LG

long double *x

strtod

int *x

strtol

%hi

short *x

strtol

%li

long *x

strtol

int *x

%hn

short *x

%ln

long *x

unsigned int *x

strtoul

%ho

unsigned short *x

strtoul

%lo

unsigned long *x

strtoul

void **x

char x[]

%ls

wchar_t x[]

unsigned int *x

strtoul

%hu

unsigned short *x

strtoul

%lu

unsigned long *x

strtoul

unsigned int *x

strtoul

%hx

unsigned short *x

strtoul

%lx

unsigned long *x

strtoul

unsigned int *x

strtoul

%hX

unsigned short *x

strtoul

%lX

unsigned long *x

strtoul

%[...]

char x[]

%l[...]

wchar_t x[]

none

The scan conversion specifier (or scan set (page 28)) determines any behavior not

summarized in this table. In the following descriptions, examples follow each of

the scan conversion specifiers. In each example, the function sscanf matches the

bold characters.

You write %c to store the matched input characters in an array object. If you

specify no field width w, then w has the value one. The match does not skip

leading white space (page 31). Any sequence of w characters matches the

conversion pattern.

sscanf("129E-2", "%c", &c)

stores '1'

sscanf("129E-2", "%2c", &c[0])

stores '1', '2'

For a wide stream (page 18), conversion occurs as if by repeatedly calling wcrtomb,

beginning in the initial conversion state (page 12).

swscanf(L"129E-2", L"%c", &c)

stores '1'

You write %lc to store the matched input characters in an array object, with

elements of type wchar_t. If you specify no field width w, then w has the value

one. The match does not skip leading white space (page 31). Any sequence of w

characters matches the conversion pattern. For a byte stream (page 18), conversion

occurs as if by repeatedly calling mbrtowc, beginning in the initial conversion state

(page 12).

sscanf("129E-2", "%lc", &c)

stores L'1'

sscanf("129E-2", "%2lc", &c)

stores L'1', L'2'

swscanf(L"129E-2", L"%lc", &c)

stores L'1'

You write %d, %i, %o, %u, %x, or %X to convert the matched input characters as

a signed integer and store the result in an integer object.

Chapter 7. Formatted Input

sscanf("129E-2", "%o%d%x", &i, &j, &k) stores 10, 9, 14

You write %e, %E, %f, %g, or %G to convert the matched input characters as a

signed fraction, with an optional exponent, and store the result in a floating-point

object.

sscanf("129E-2", "%e", &f)

stores 1.29

You write %n to store the number of characters matched (up to this point in the

format) in an integer object. The match does not skip leading white space and does

not match any input characters.

sscanf("129E-2", "12%n", &i)

stores 2

You write %p to convert the matched input characters as an external representation

of a pointer to void and store the result in an object of type pointer to void. The input

characters must match the form generated by the %p print conversion specification

(page 32).

sscanf("129E-2", "%p", &p)

stores, e.g. 0x129E

You write %s to store the matched input characters in an array object, followed by

a terminating null character. If you do not specify a field width w, then w has a

large value. Any sequence of up to w non white-space characters matches the

conversion pattern.

sscanf("129E-2", "%s", &s[0])

stores "129E-2"

For a wide stream (page 18), conversion occurs as if by repeatedly calling wcrtomb

beginning in the initial conversion state (page 12).

swscanf(L"129E-2", L"%s", &s[0])

stores "129E-2"

You write %ls to store the matched input characters in an array object, with

elements of type wchar_t, followed by a terminating null wide character. If you do

not specify a field width w, then w has a large value. Any sequence of up to w non

white-space characters matches the conversion pattern. For a byte stream (page 18),

conversion occurs as if by repeatedly calling mbrtowc, beginning in the initial

conversion state.

sscanf("129E-2", "%ls", &s[0])

stores L"129E-2"

swscanf(L"129E-2", L"%ls", &s[0])

stores L"129E-2"

You write %[ to store the matched input characters in an array object, followed by

a terminating null character. If you do not specify a field width w, then w has a

large value. The match does not skip leading white space. A sequence of up to w

characters matches the conversion pattern in the scan set that follows. To complete

the scan set, you follow the left bracket ([) in the conversion specification with a

sequence of zero or more match characters, terminated by a right bracket (]).

If you do not write a caret (^) immediately after the [, then each input character

must match one of the match characters. Otherwise, each input character must not

match any of the match characters, which begin with the character following the ^.

If you write a ] immediately after the [ or [^, then the ] is the first match

character, not the terminating ]. If you write a minus (-) as other than the first or

last match character, an implementation can give it special meaning. It usually

indicates a range of characters, in conjunction with the characters immediately

preceding or following, as in 0-9 for all the digits.) You cannot specify a null

match character.

sscanf("129E-2", "[54321]", &s[0])

stores "12"

Standard C++ Library

For a wide stream (page 18), conversion occurs as if by repeatedly calling wcrtomb,

beginning in the initial conversion state.

swscanf(L"129E-2", L"[54321]", &s[0])

stores "12"

You write %l[ to store the matched input characters in an array object, with

elements of type wchar_t, followed by a terminating null wide character. If you do

not specify a field width w, then w has a large value. The match does not skip

leading white space. A sequence of up to w characters matches the conversion

pattern in the scan set (page 28) that follows.

For a byte stream (page 18), conversion occurs as if by repeatedly calling mbrtowc,

beginning in the initial conversion state.

sscanf("129E-2", "l[54321]", &s[0])

stores L"12"

swscanf(L"129E-2", L"l[54321]", &s[0]) stores L"12"

You write %% to match the percent character (%). The function does not store a

value.

sscanf("%

0XA", "%% %i")

stores 10

Chapter 7. Formatted Input

Standard C++ Library

Table of Contents: