Expressions in awk

Expressions describe computations used in patterns and actions. In the following table, valid expression operations are given in groups from highest precedence first to lowest precedence last, with equal-precedence operators grouped between horizontal lines. In expression evaluation, where the grammar is formally ambiguous, higher precedence operators are evaluated before lower precedence operators. In this table expr, expr1, expr2, and expr3 represent any expression, while lvalue represents any entity that can be assigned to (that is, on the left side of an assignment operator).

Each expression has either a string value, a numeric value or both. Except as stated for specific contexts, the value of an expression is implicitly converted to the type needed for the context in which it is used. A string value is converted to a numeric value by the equivalent of the following calls:

setlocale(LC_NUMERIC, "");
numeric_value = atof(string_value);

A numeric value that is exactly equal to the value of an integer is converted to a string by the equivalent of a call to the sprintf function with the string %d as the fmt argument and the numeric value being converted as the first and only expr argument. Any other numeric value is converted to a string by the equivalent of a call to the sprintf function with the value of the variable CONVFMT as the fmt argument and the numeric value being converted as the first and only expr argument.

A string value is considered to be a numeric string in the following case:

If a − character is ignored in the above steps, the numeric value of the numeric string is the negation of the numeric value of the recognized NUMBER token. Otherwise the numeric value of the numeric string is the numeric value of the recognized NUMBER token. Whether or not a string is a numeric string is relevant only in contexts where that term is used in this section.

When an expression is used in a Boolean context, if it has a numeric value, a value of zero is treated as false and any other value is treated as true. Otherwise, a string value of the null string is treated as false and any other value is treated as true. A Boolean context is one of the following:

The awk language supplies arrays that are used for storing numbers or strings. Arrays need not be declared. They are initially empty, and their sizes changes dynamically. The subscripts, or element identifiers, are strings, providing a type of associative array capability. An array name followed by a subscript within square brackets can be used as an lvalue and as an expression, as described in the grammar. Unsubscripted array names are used in only the following contexts:

A valid array index consists of one or more comma-separated expressions, similar to the way in which multi-dimensional arrays are indexed in some programming languages. Because awk arrays are really one-dimensional, such a comma-separated list is converted to a single string by concatenating the string values of the separate expressions, each separated from the other by the value of the SUBSEP variable.

Thus, the following two index operations are equivalent:

var[expr1, expr2, ... exprn]
var[expr1 SUBSEP expr2 SUBSEP ... SUBSEP exprn]

A multi-dimensioned index used with the in operator must be put in parentheses. The in operator, which tests for the existence of a particular array element, does not create the element if it does not exist. Any other reference to a non-existent array element automatically creates it.

Variables and Special Variables

Variables can be used in an awk program by referencing them. With the exception of function parameters, they are not explicitly declared. Uninitialized scalar variables and array elements have both a numeric value of zero and a string value of the empty string.

Field variables are designated by a $ followed by a number or numerical expression. The effect of the field number expression evaluating to anything other than a non-negative integer is unspecified. Uninitialized variables or string values need not be converted to numeric values in this context. New field variables are created by assigning a value to them. References to non-existent fields (that is, fields after $NF) produce the null string. However, assigning to a non-existent field (for example, $(NF+2) = 5) increases the value of NF, create any intervening fields with the null string as their values and cause the value of $0 to be recomputed, with the fields being separated by the value of OFS. Each field variable has a string value when created. If the string, with any occurrence of the decimal-point character from the current locale changed to a period character, is considered a numeric string (see Expressions in awk above), the field variable also has the numeric value of the numeric string.

/usr/bin/awk, /usr/xpg4/bin/awk

awk sets the following special variables that are supported by both /usr/bin/awk and /usr/xpg4/bin/awk:

ARGC

ARGV

CONVFMT

ENVIRON

FILENAME

FNR

FS

NF

NR

OFMT

OFS

ORS

RLENGTH

RS

RSTART

SUBSEP

/usr/bin/awk

The following variable is supported for /usr/bin/awk only:

RT

Regular Expressions

The awk utility makes use of the extended regular expression notation (see regex(7)) except that it allows the use of C-language conventions to escape special characters within the EREs, namely \\, \a, \b, \f, \n, \r, \t, \v, and those specified in the following table. These escape sequences are recognized both inside and outside bracket expressions. Note that records need not be separated by newline characters and string constants can contain newline characters, so even the \n sequence is valid in awk EREs. Using a slash character within the regular expression requires escaping as shown in the table below:

A regular expression can be matched against a specific field or string by using one of the two regular expression matching operators, ~ and !~. These operators interpret their right-hand operand as a regular expression and their left-hand operand as a string. If the regular expression matches the string, the ~ expression evaluates to the value 1, and the !~ expression evaluates to the value 0. If the regular expression does not match the string, the ~ expression evaluates to the value 0, and the !~ expression evaluates to the value 1. If the right-hand operand is any expression other than the lexical token ERE, the string value of the expression is interpreted as an extended regular expression, including the escape conventions described above. Notice that these same escape conventions also are applied in the determining the value of a string literal (the lexical token STRING), and is applied a second time when a string literal is used in this context.

When an ERE token appears as an expression in any context other than as the right-hand of the ~ or !~ operator or as one of the built-in function arguments described below, the value of the resulting expression is the equivalent of:

$0 ~ /ere/

The ere argument to the gsub, match, sub functions, and the fs argument to the split function (see String Functions) is interpreted as extended regular expressions. These can be either ERE tokens or arbitrary expressions, and are interpreted in the same manner as the right-hand side of the ~ or !~ operator.

An extended regular expression can be used to separate fields by using the -F ERE option or by assigning a string containing the expression to the built-in variable FS. The default value of the FS variable is a single space character. The following describes FS behavior:

Except in the gsub, match, split, and sub built-in functions, regular expression matching is based on input records. That is, record separator characters (the first character of the value of the variable RS, a newline character by default) cannot be embedded in the expression, and no expression matches the record separator character. If the record separator is not a newline character, newline characters embedded in the expression can be matched. In those four built-in functions, regular expression matching are based on text strings. So, any character (including the newline character and the record separator) can be embedded in the pattern and an appropriate pattern matches any character. However, in all awk regular expression matching, the use of one or more NULL characters in the pattern, input record or text string produces undefined results.

Patterns

A pattern is any valid expression, a range specified by two expressions separated by comma, or one of the two special patterns BEGIN or END.

Special Patterns

The awk utility recognizes two special patterns, BEGIN and END. Each BEGIN pattern is matched once and its associated action executed before the first record of input is read (except possibly by use of the getline function in a prior BEGIN action) and before command line assignment is done. Each END pattern is matched once and its associated action executed after the last record of input has been read. These two patterns have associated actions.

BEGIN and END do not combine with other patterns. Multiple BEGIN and END patterns are allowed. The actions associated with the BEGIN patterns are executed in the order specified in the program, as are the END actions. An END pattern can precede a BEGIN pattern in a program.

If an awk program consists of only actions with the pattern BEGIN, and the BEGIN action contains no getline function, awk exits without reading its input when the last statement in the last BEGIN action is executed. If an awk program consists of only actions with the pattern END or only actions with the patterns BEGIN and END, the input is read before the statements in the END actions are executed.

Expression Patterns

An expression pattern is evaluated as if it were an expression in a Boolean context. If the result is true, the pattern is considered to match, and the associated action (if any) is executed. If the result is false, the action is not executed.

Pattern Ranges

A pattern range consists of two expressions separated by a comma. In this case, the action is performed for all records between a match of the first expression and the following match of the second expression, inclusive. At this point, the pattern range can be repeated starting at input records subsequent to the end of the matched range.

Actions

An action is a sequence of statements. A statement can be one of the following:

if ( expression ) statement [ else statement ]
while ( expression ) statement
do statement while ( expression )
for ( expression ; expression ; expression ) statement
for ( var in array ) statement
delete array[subscript] #delete an array element
delete array #delete all elements within an array
break
continue
{ [ statement ] ... }
expression        # commonly variable = expression
print [ expression-list ] [ >expression ]
printf format [ ,expression-list ] [ >expression ]
next              # skip remaining patterns on this input line
nextfile          # skip remaining patterns on this input file
exit [expr] # skip the rest of the input; exit status is expr
return [expr]

Any single statement can be replaced by a statement list enclosed in braces. The statements are terminated by newline characters or semicolons, and are executed sequentially in the order that they appear.

The next statement causes all further processing of the current input record to be abandoned. The behavior is undefined if a next statement appears or is invoked in a BEGIN or END action.

The nextfile statement is similar to next, but also skips all other records in the current file, and moves on to processing the next input file if available (or exits the program if there are none). (Note that this keyword is not supported by /usr/xpg4/bin/awk.)

The exit statement invokes all END actions in the order in which they occur in the program source and then terminate the program without reading further input. An exit statement inside an END action terminates the program without further execution of END actions. If an expression is specified in an exit statement, its numeric value is the exit status of awk, unless subsequent errors are encountered or a subsequent exit statement with an expression is executed.

Output Statements

Both print and printf statements write to standard output by default. The output is written to the location specified by output_redirection if one is supplied, as follows:

> expression>> expression| expression

In all cases, the expression is evaluated to produce a string that is used as a full pathname to write into (for > or >>) or as a command to be executed (for |). Using the first two forms, if the file of that name is not currently open, it is opened, creating it if necessary and using the first form, truncating the file. The output then is appended to the file. As long as the file remains open, subsequent calls in which expression evaluates to the same string value simply appends output to the file. The file remains open until the close function, which is called with an expression that evaluates to the same string value.

The third form writes output onto a stream piped to the input of a command. The stream is created if no stream is currently open with the value of expression as its command name. The stream created is equivalent to one created by a call to the popen(3C) function with the value of expression as the command argument and a value of w as the mode argument. As long as the stream remains open, subsequent calls in which expression evaluates to the same string value writes output to the existing stream. The stream remains open until the close function is called with an expression that evaluates to the same string value. At that time, the stream is closed as if by a call to the pclose function.

These output statements take a comma-separated list of expression s referred in the grammar by the non-terminal symbols expr_list, print_expr_list or print_expr_list_opt. This list is referred to here as the expression list, and each member is referred to as an expression argument.

The print statement writes the value of each expression argument onto the indicated output stream separated by the current output field separator (see variable OFS above), and terminated by the output record separator (see variable ORS above). All expression arguments is taken as strings, being converted if necessary; with the exception that the printf format in OFMT is used instead of the value in CONVFMT. An empty expression list stands for the whole input record ($0).

The printf statement produces output based on a notation similar to the File Format Notation used to describe file formats in this document Output is produced as specified with the first expression argument as the string format and subsequent expression arguments as the strings arg1 to argn, inclusive, with the following exceptions:

Both print and printf can output at least {LINE_MAX} bytes.

Functions

The awk language has a variety of built-in functions: arithmetic, string, input/output and general.

Arithmetic Functions

The arithmetic functions, except for int, are based on the ISO C standard. The behavior is undefined in cases where the ISO C standard specifies that an error be returned or that the behavior is undefined. Although the grammar permits built-in functions to appear with no arguments or parentheses, unless the argument or parentheses are indicated as optional in the following list (by displaying them within the [ ] brackets), such use is undefined.

atan2(y,x)

cos(x)

sin(x)

exp(x)

log(x)

sqrt(x)

int(x)

rand()

srand([expr])

String Functions

The string functions in the following list shall be supported. Although the grammar permits built-in functions to appear with no arguments or parentheses, unless the argument or parentheses are indicated as optional in the following list (by displaying them within the [ ] brackets), such use is undefined.

gsub(ere,repl[,in])

index(s,t)

length[([v])]

match(s,ere)

split(s,a[,fs])

sprintf(fmt,expr,expr,...)

sub(ere,repl[,in])

substr(s,m[,n])

tolower(s)

toupper(s)

All of the preceding functions that take ERE as a parameter expect a pattern or a string valued expression that is a regular expression as defined below.

Input/Output and General Functions

The input/output and general functions are:

close(expression)

fflush(expression)

expression|getline[var]

getline

getline var

getline [var] < expression

system(expression)

All forms of getline return 1 for successful input, 0 for end of file, and −1 for an error.

Where strings are used as the name of a file or pipeline, the strings must be textually identical. The terminology ``same string value'' implies that ``equivalent strings'', even those that differ only by space characters, represent different files.

User-defined Functions

The awk language also provides user-defined functions. Such functions can be defined as:

function name(args,...) { statements }

A function can be referred to anywhere in an awk program; in particular, its use can precede its definition. The scope of a function is global.

Function arguments can be either scalars or arrays; the behavior is undefined if an array name is passed as an argument that the function uses as a scalar, or if a scalar expression is passed as an argument that the function uses as an array. Function arguments are passed by value if scalar and by reference if array name. Argument names are local to the function; all other variable names are global. The same name is not used as both an argument name and as the name of a function or a special awk variable. The same name must not be used both as a variable name with global scope and as the name of a function. The same name must not be used within the same scope both as a scalar variable and as an array.

The number of parameters in the function definition need not match the number of parameters in the function call. Excess formal parameters can be used as local variables. If fewer arguments are supplied in a function call than are in the function definition, the extra parameters that are used in the function body as scalars are initialized with a string value of the null string and a numeric value of zero, and the extra parameters that are used in the function body as arrays are initialized as empty arrays. If more arguments are supplied in a function call than are in the function definition, the behavior is undefined.

When invoking a function, no white space can be placed between the function name and the opening parenthesis. Function calls can be nested and recursive calls can be made upon functions. Upon return from any nested or recursive function call, the values of all of the calling function's parameters are unchanged, except for array parameters passed by reference. The return statement can be used to return a value. If a return statement appears outside of a function definition, the behavior is undefined.

In the function definition, newline characters are optional before the opening brace and after the closing brace. Function definitions can appear anywhere in the program where a pattern-action pair is allowed.