The syntax of a format specification fs is

([flist]) 

where flist is a list of format specifiers of one of the following forms, separated by commas:

[r]fd 
ed 
[r]fs 

where

The comma used to separate the format specifiers in flist can be omitted as follows:

Some descriptors can be repeated, others cannot. The repeatable descriptors are

where

These descriptors are described in the respective section.

The nonrepeatable descriptors are

where

Each field descriptor corresponds to a particular data type I/O list item:

Ow, and Zw are extensions to Fortran 77.

The terms r, c, n, d, m, e, and w must all be unsigned integer constants, and, additionally, r, c, n, e, and w must be nonzero. k is an optionally signed integer constant. Descriptions of these list items are given in the sections that describe the individual field descriptors.

The repeat specifier r can be used only with the I, O, Z, F, E, D, G, L, and A field descriptors and with format groups.

The d is required in the F, E, D, and G field descriptors. Ee is optional in the E and G field descriptors and invalid in the others.

Use of named constants anywhere in a format specification is not allowed.

Table 9-1 contains an alphabetical summary of the field and edit descriptors.

Each of the field descriptors described Table 9-1 is discussed in detail in the following sections.

Variable format expressions provide a means for substituting run-time expressions for the field width and other parameters of the field and edit descriptors of the statement. Any expression can be enclosed in angle brackets (<>) and used as an integer constant would be used in the same situation. This facility is not available for anything other than a compile-time FORMAT statement.

Here is an example that uses a variable format expression:
    program VariableExample
    character*12 greeting
    greeting = 'Good Morning!'
    do 110 I = 1, 12
    write (*,115) (greeting)
 115 format (A<I>)
 110 continue
     end 

In the above example, the field descriptor for greeting has the format Aw where w is a variable width specifier I (initially set to 1) for the iolist item greeting. In twelve successive WRITE operations, I is incremented by 1 to produce the following output:

G
Go
Goo
Good
Good 
Good M
Good Mo
Good Mor
Good Morn
Good Morni
Good Mornin
Good Morning

The following rules apply to variable format expressions:

Because FORMAT allows exact specification of input and output format, it is necessarily complex. Some guidelines to its correct usage are outlined below.

Guidelines that apply specifically to input are

Guidelines that apply specifically to output are

The I, O, Z, F, E, D, and G field descriptors are used for numeric editing. This section also describes the P edit descriptor, which is a scale factor, that alters the effect of F, E, D, and G field descriptors.

Unless otherwise indicated, the following rules apply:

You can optionally specify a field-width value (w, d, and e) for the I, O, Z, L, F, E, D, G, and A field descriptors. If you do not specify a value, the default values shown in Table 9-2 apply. The length of the I/O variable determines the length n for the A field descriptor.

The I field descriptor is used for conversion between an internal integer data item and an external decimal integer. It has the form

Iw[.m] 

where

In an input statement, the I field descriptor reads a field of w characters from the record, interprets it as an integer constant, and assigns the integer value to the corresponding I/O list item. The corresponding I/O list element must be of the INTEGER or LOGICAL data type. The external data must have the form of an integer constant; it must not contain a decimal point or exponent.

A LOGICAL data type is displayed as either the value 0 (false) or 1 (true).

If the first nonblank character of the external field is a minus sign, the field is treated as a negative value. If the first nonblank character is a plus sign, or if no sign appears in the field, the field is treated as a positive value. An all-blank field is treated as a value of zero.

Table 9-3 contains input examples.

In an output statement, the I field descriptor constructs an integer constant representing the value of the corresponding I/O list item and writes it to the right-justified record in an external field w characters long. If the value does not fill the field, leading blanks are inserted; if the value exceeds the field width, the entire field is filled with asterisks. If the value of the list item is negative, the field will have a minus sign as its left most, nonblank character. The term w must therefore be large enough to provide for a minus sign, when necessary. If m is present, the external field consists of at least m digits, with leading zeros, if necessary.

If m is zero, and the internal representation is zero, the external field is filled with blanks.

Table 9-4 contains output examples.

The O field descriptor transfers data values and converts them to octal form. It has the form

Ow[m] 

where

This repeatable descriptor interprets and assigns data in the same way as the I field descriptor, except that the external field represents an octal number constructed with the digits 0 through 7. On input, if BZ is in effect, embedded and trailing blanks in the field are treated as zeros; otherwise, blanks are ignored. On output, S, SP, and SS do not apply.

In an input statement, the field is terminated when a non-octal digit is encountered. Fortran 77 treats embedded and trailing blanks as zeros.

In an input statement, the O field descriptor reads w characters from the record; the input field must have:

A field that is entirely blank is treated as the value zero.

Table 9-5 contains examples of O field input values. BN is assumed in effect, and internal values are expressed in decimal (base 10).

In an output statement, the O field descriptor constructs an octal number representing the unsigned value of the corresponding I/O list element as follows:

Table 9-6 lists examples of O field output.

The Z field descriptor transfers data values and converts them to hexadecimal form. It has the form

Zw[m] 

where

This repeatable descriptor interprets and assigns data in the same way as the I field descriptor, except that the external field represents a hexadecimal number constructed with the digits 0 through 9 and the letters A through F. On output, the output list item is interpreted as an unsigned integer value.

In an input statement, the O field descriptor reads w characters from the record. After embedded and trailing blanks are converted to zeros or ignored, as applicable, the input field must have

A field that is entirely blank is given a value of zero.

Table 9-7 lists examples of Z field input. BN is assumed in effect, and internal values are expressed in decimal (base 10).

Table 9-8 lists examples of Z field output.

The F field descriptor transfers real values. It has the form

Fw.d 

where

The corresponding I/O list element must be of type REAL, DOUBLEPRECISION, or COMPLEX.

In an input statement, the F field descriptor reads a field of w characters from the record and, after appropriate editing of leading, trailing, and embedded blanks, interprets it as an integer or a real constant. It then assigns the real value to the corresponding I/O list element. (Refer to Chapter 2, "Constants and Data Structures," for more information.) If the external field contains an exponent, the letter E can be omitted as long as the value of the exponent is a signed integer. If the first nonblank character of the external field is a minus sign, the field is treated as a negative value. If the first nonblank character is a plus sign, or if no sign appears in the field, the field is treated as a positive value. An all-blank field is given a value of zero.

If the field contains neither a decimal point nor an exponent, it is treated as a real number in which the right most d digits are to the right of the decimal point, with leading zeros assumed if necessary. If the field contains an explicit decimal point, the location of that decimal point overrides the location specified by the value of d in the field descriptor. If the field contains a real exponent, the effect of any associated scale factor kP (see Scale Factor) is suppressed, and the real exponent is used to establish the magnitude of the value in the input field before it is assigned to the list element.

Table 9-9 provides examples of F field input.

In an output statement, the F field descriptor constructs a basic real constant representing the value of the corresponding I/O list element, rounded to d decimal positions, and writes it to the record right-justified in an external field w characters long.

The term w must be large enough to include:

If w is insufficiently large, the entire field width is filled with asterisks. Therefore, w must be > d + 2.

Table 9-10 provides examples of F field output.

The E field descriptor transfers real values in exponential form. It has the form

Ew.d[Ee] 

where

The corresponding I/O list element must be of REAL, DOUBLEPRECISION, or COMPLEX data type.

In an input statement, the E field descriptor interprets and assigns data in exactly the same way as the F field descriptor.

Table 9-11 provides examples of E field input.

In Table 9-11, the E field descriptor treats the D exponent field indicator the same as an E exponent indicator.

In an output statement, the E field descriptor constructs a real constant representing the value of the corresponding I/O list element, rounded to d decimal digits, and writes it to the right-justified record in an external field w characters long. If the value does not fill the field, leading spaces are inserted; if the value exceeds the field width, the entire field is filled with asterisks.

When an E field descriptor is used, data output is transferred in a standard form. This form consists of

The exponent has one of the following forms:

The exponent field-width specification is optional; if it is omitted, the exponent part is as shown above. If the exponent value is too large to be output with the given value e as shown in the third form above, the entire field is filled with asterisks.

The term w must be large enough to include

Given these limitations and assuming a P edit descriptor is in effect, w is \xb3 d + 7, or \xb3 d + e + 5 if e is present.

Table 9-12 provides examples of E field output.

The D field descriptor transfers real values in exponential form. It has the form

Dw.d

where

The corresponding I/O list element must be of REAL, DOUBLEPRECISION, or COMPLEX data type.

In an input statement, the D field descriptor interprets and assigns data in exactly the same way as the F field descriptor.

Table 9-13 provides examples of D field input.

In an output statement, the D field descriptor is the same as the E field descriptor, except the D exponent field indicator replaces the E indicator.

Table 9-14 provides examples of D field output.

A G field descriptor is used for the conversion and editing of real data when the magnitude of the data is unknown. On output, the G field descriptor produces a field as do the F or E field descriptors, depending on the value. On input, the G field descriptor interprets and assigns data in exactly the same way as the F field descriptor. It has the form

Gw.d[Ee] 

where

The corresponding I/O list element must be of REAL, DOUBLEPRECISION, or COMPLEX data type.

In an input statement, the G field descriptor interprets and assigns data in exactly the same way as the F field descriptor.

In an output statement, the G field descriptor constructs a real constant representing the value of the corresponding I/O list element rounded to d decimal digits and writes it to the right-justified record in an external field w characters long. The form in which the value is written is a function of the magnitude of the value m, as described in Table 9-1. In the table, n is 4 if Ee was omitted from the G field descriptor; otherwise n is e + 2.

Table 9-15 illustrates the effect of data magnitude on G format conventions.

The term w must be large enough to include

Given these limitations, w must therefore be \xb3 d + 7 or \xb3 d + e + 5.

Table 9-16 provides examples of G field output.

For comparison, the examples in Table 9-17 use the same values with an equivalent F field descriptor.

The P edit descriptor specifies a scale factor and has the form

kP 

where k is an optionally signed integer constant called the scale factor.

A P edit descriptor can appear anywhere in a format specification but must precede the first field descriptor that is to be associated with it. For example

kPFw.d  kPEw.d  kPD w.d  kPGw.d 

The value of k must not be greater than d + 1, where d is the number of digits in the Ew.d, Dw.d, or Gw.d output fields.

The L edit descriptor is used for logical data. The specified I/O list item must be of type LOGICAL. It has the form

Lw 

where w is a nonzero, unsigned integer constant denoting field width.

For input, the field must consist of optional blanks followed by an optional decimal point followed by a T (for true) or F (for false). The T or F can be followed by additional characters that have no effect. The logical constants .TRUE. and .FALSE. are acceptable input forms.

For output, the field consists of w – 1 blanks followed by a T or an F, for true and false, respectively, according to the value of the internal data. Table 9-20 shows L field examples.

The L edit descriptor can also be used to process integer data items. All nonzero values are displayed as .TRUE. and all zero values as .FALSE..

The A edit descriptor is used for editing character or Hollerith data. It has the form

A[w] 

where w is a nonzero, unsigned integer constant denoting the width, in number of characters, of the external data field. If w is omitted, the size of the I/O list item determines the length w.

The corresponding I/O list item can be any data type. If it is character data type, character data is transmitted. If it is any other data type, Hollerith data is transmitted.

In an input statement, the A edit descriptor reads a field of w characters from the record without interpretation and assigns it to the corresponding I/O list item. The maximum number of characters that can be stored depends on the size of the I/O list item. For character I/O list elements, the size is the length of the character variable, character substring reference, or character array element. For numeric and logical I/O list elements, the size depends on the data type, as shown in Table 9-21.

If w is greater than the maximum number of characters that can be stored in the corresponding I/O list item, only the right most characters of the field are assigned to that element. The left most excess characters are ignored. If w is less than the number of characters that can be stored, w characters are assigned to the list item and left justified, and trailing blanks are added to fill it to its maximum size.

The I, O, Z, F, E, D, G, L, and A field descriptors can be applied to a number of successive I/O list items by preceding the field descriptor with an unsigned integer constant, called the repeat count. For example, 4F5.2 is equivalent to F5.2, F5.2, F5.2, F5.2.

Enclosing a group of field descriptors in parentheses, and preceding the enclosed group with a repeat count, repeats the entire group. Thus, 2(I6,F8.4) is equivalent to I6,F8.4,I6,F8.4.

The H field descriptor is used for output of character literal data. It has the form:

nHxxx… x 

where

In an output statement, the H field descriptor writes the n characters following the letter H from the field descriptor to the record as an external field n characters long. The H field descriptor does not correspond to an output list item.

Table 9-24 lists examples of H edit description output.

An H field descriptor must not be encountered by a READ statement.

A character edit descriptor has one of the following forms:

'X1 X2 … Xn' 
X1 X2 … Xn 

where X1 X2 … Xn are members of the Fortran character set forming a valid character literal. The width of the output field is the number of characters contained in the character literal, excluding the enclosing apostrophes or quotation marks. The character edit descriptor does not correspond to an output list item. Within a character edit descriptor delimited by apostrophes, an apostrophe is represented by two successive apostrophe characters. Within a character edit descriptor delimited by quotation marks, a quotation mark is represented by two successive quotation mark characters.

The Q edit descriptor is used to determine the number of characters remaining to be read from the current input record. It has the form

Q 

When a Q descriptor is encountered during the execution of an input statement, the corresponding input list item must be type integer. Interpretation of the Q edit descriptor causes the input list item to be defined with a value that represents the number of character positions in the formatted record remaining to be read. Therefore, if c is the character position within the current record of the next character to be read and the record consists of len characters, then the item is defined with the value

n = max (len - c + 1, 0) 

If no characters have yet been read, then n = len, the length of the record. If all the characters of the record have been read (c > len), then n is zero.

The Q edit descriptor must not be encountered during the execution of an output statement.

     INTEGER N 
     CHARACTER LINE * 80 
     READ (5, 100) N, LINE (1:N) 
100  FORMAT (Q, A) 

The X edit descriptor specifies a position forward (to the right) of the current position. It is used to skip characters on the external medium for input and output. It has the form

nX 

where n is a nonzero, unsigned integer constant denoting the number of characters to be skipped.

The T edit descriptor specifies an absolute position in an input or output record. It has the form:

Tn 

where n indicates that the next character transferred to or from the record is the nth character of the record.

The TL edit descriptor specifies a position to the left of the current position. It has the form

TLn 

where n indicates that the next character to be transferred from or to the record is the nth character to the left of the current character. The value of n must be greater than or equal to one.

If n is the current character position, then the first character in the record is specified.

The TR edit descriptor specifies a position to the right of the current position. It has the form

TRn 

where n indicates that the next character to be transferred from or to a record is the nth character to the right of the current character. The value of n must be greater than or equal to one.

The BN edit descriptor causes the processor to ignore blank characters in a numeric input field and to right justify the remaining characters, as though the blanks that were ignored were leading blanks. It has the form

BN 

The BN descriptor affects only I, O, Z, F, E, D, and G editing and then only on input fields.

The BZ edit descriptor causes the processor to treat all the embedded and trailing blank characters it encounters within a numeric input field as zeros. It has the form:

BZ 

The BZ descriptor affects only I, O, Z, F, E, D, and G editing and then only on input fields.

The SP edit descriptor specifies that a plus sign be inserted in any character position that normally contains an optional plus sign and whose actual value is \xb3 0. It has the form

SP 

The SP descriptor affects only I, F, E, D, and G editing and then only on output fields.

The SS edit descriptor specifies that a plus sign should not be inserted in any character position that normally contains an optional plus sign. It has the form

SS 

The SS descriptor affects only I, F, E, D, and G editing and then only on output fields.

The S edit descriptor resets the option of inserting plus characters (+) in numeric output fields to the processor default. It has the form

S 

The S descriptor counters the action of either the SP or the SS descriptor by restoring to the processor the discretion of producing plus characters (+) on an optional basis. The default is to SS processing; the optional plus sign is not inserted when S is in effect.

The S descriptor affects only I, F, E, D, and G editing and then only on output fields.

The colon character (:) in a format specification terminates format control if no more items are in the I/O list. The colon descriptor has no effect if I/O list items remain.

The $ edit descriptor suppresses the terminal line-mark character at the end of the current output record. It has the form

$ 

The $ descriptor is nonrepeatable and is ignored when encountered during input operations.

       print 100, 'enter a number:'
100    format (1x, a, $)
       read *, x 

A formatted record can contain a prescribed carriage-control character as the first character of the record. The carriage-control character determines vertical spacing in printing when the CARRIAGECONTROL keyword of the OPEN statement is set to FORTRAN (as described in ``OPEN'' of Chapter 8, "Input/Output Statements.") Table 9-28 lists the carriage-control characters.

The carriage-control character is not printed, and the remaining characters, if any, are printed on one line beginning at the left margin. If there are no characters in the record, the vertical spacing is one line and no characters will be printed in that line.

A slash (/) placed in a format specification terminates input or output for the current record and initiates a new record. For example

    WRITE (6,40) K,L,M,N,O,P 
40  FORMAT (3I6.6/I6,2F8.4) 

is equivalent to

    WRITE (6,40) K,L,M 
40  FORMAT (3I6.6) 
    WRITE (6,50) N,O,P 
50  FORMAT (I6,2F8.4) 

On input from a sequential-access file, the current portion of the remaining record is skipped, a new record is read, and the current position is set to the first character of the record. n slashes in succession cause n – 1 records to be skipped.

On output to a file connected for sequential access, a new record is created and becomes the last and current record of the file. Also, n slashes in succession cause n – 1 blank lines to be generated.

Through the use of two or more successive slashes in a format specification, entire records can be skipped for input and records containing no characters can be generated for output. If the file is an internal file, or a file connected for direct access, skipped records are filled with blank characters on output.

n slashes at the beginning or end of a format specification result in n skipped or blank records. On input and output from a direct-access file, the record number is increased by one and the file is positioned at the beginning of the record that has that record number. This record becomes the current record.

The characters in one or more list-directed records form a sequence of values and value separators. Each value is either a constant, or a null value or has one of the following forms:

r*c
r*

where

The r* form is equivalent to r successive null values. Neither form can contain embedded blanks, except where permitted within the constant c.

Data values can be separated with one of the following value separators:

The end of a record has the effect of a blank, except when it appears within a character constant. Two or more consecutive blanks are treated as a single blank, unless they occur within a character constant.

There are three differences between the input forms acceptable to format specifiers for a data type and those used for list-directed formatting. A data value must have the same type as the list item to which it corresponds. Blanks are not interpreted as zeros. Embedded blanks are only allowed in constants of character or complex type.

Rules governing input forms of list items for list-directed formatting are

The form of the values produced is the same as that required for input, except as noted below: