Built-in string variables

awk also provides a number of built-in string variables:

FILENAME

Contains the name of the current input file. For example, when running programs against the hobbies file, the value of FILENAME would be hobbies (if that is the file you are using). If the input is coming from the awk standard input, the value is -.

FS

Is the field separator string, giving the character that is used to separate fields in the current file. The default value for FS is "" (a single blank), which as a special case matches both blank and tab. However, if the command line contains an –F option specifying a different field separator, FS is a string containing the given separator character. A program may also assign values to FS to indicate new field separator characters. For example, you could create a data file with a first line that provides the character used to separate fields in the records in the rest of the file. An awk program could then contain the instruction:

FNR == 1 { FS = $0 }

This says that the field separator string FS should be assigned the contents of the first record in the current data file. The character in this line is then taken to be the field separator for the rest of the file (unless FS changes value again). Any FS value of more than one character is used as a regular expression. For details, see the Input topic of the awk command description in z/OS UNIX System Services Command Reference.

RS

Is the input record separator. Just as FS indicates the character that separates fields within records, RS indicates the character that separates one record from another. By default, RS contains a newline character, which means that input records are separated by newlines. However, you can assign a different character to RS; for example, with:

RS = ";"

input records are separated by semicolons. This lets you have several records on a single line, or a single record that extends over several lines. Records are separated by a semicolon, not a <newline> character. As an important special case:

RS = ""

separates records by empty lines.

OFS

Gives the output field separator string. When you use the print action to print several values, as in:

{ print A, B, C }

awk prints the output field separator string between each of the values. By default, OFS contains a single blank character, which is why output values are separated by a single blank. However, if you make the assignment:

OFS = " : "

the output values are separated by the given string. You can also use OFS to reconstruct the $0 field during field assignment.

ORS

Gives the output record separator. When you use the print action to print records, awk prints the output record separator at the end of each record. By default, ORS is the newline character, which is why print prints a new output line each time it is called. However, you can use a different separator string by assigning the string to ORS.

OFMT

Is the default output format for numbers when they are displayed by print. This is a format string like the one used by printf. By default, it is %.6g, indicating that numbers are to be displayed with a maximum of six digits after the decimal point. By changing OFMT, you can obtain more or less displayed precision.

CONVFMT

Is the default format which awk uses when converting numbers into strings internally. This differs from the OFMT variable, which is used only when displaying numbers. The internal conversion of a number to a string occurs when you perform concatenation, indexing, and some comparison operations. awk converts floating-point numbers (numbers that are not integers) to strings as if you had specified the operation:

sprintf(CONVFMT, number ...)

By default, the value of CONVFMT is %.6g.

Note: CONVFMT is a POSIX extension not found in traditional implementations of awk.