NetBSD/gnu/usr.bin/awk/awk.1

1570 lines
34 KiB
Groff

.TH MAWK 1 "Jan 22 1992" "Version 1.1" "USER COMMANDS"
.\" strings
.ds ex \fIexpr\fR
.SH NAME
mawk \- pattern scanning and text processing language
.SH SYNOPSIS
.B mawk
[\-\fBW
.IR option ]
[\-\fBF
.IR value ]
[\-\fBv
.IR var=value ]
[\-\|\-] 'program text' [file ...]
.br
.B mawk
[\-\fBW
.IR option ]
[\-\fBF
.IR value ]
[\-\fBv
.IR var=value ]
[\-\fBf
.IR program-file ]
[\-\|\-] [file ...]
.SH DESCRIPTION
.B mawk
is an interpreter for the AWK Programming Language.
The AWK language
is useful for manipulation of data files,
text retrieval and processing,
and for prototyping and experimenting with algorithms.
.B mawk
is a \fInew awk\fR meaning it implements the AWK language as
defined in Aho, Kernighan and Weinberger,
.I "The AWK Programming Language,"
Addison-Wesley Publishing, 1988. (Hereafter referred to as
the AWK book.)
.B mawk
conforms to the Posix 1003.2
(draft 11.2)
definition of the AWK language
which contains a few features not described in the AWK
book, and
.B mawk
provides a small number of extensions.
An AWK program is a sequence of \fIpattern {action}\fR pairs and
function definitions.
Short programs are entered on the command line
usually enclosed in ' ' to avoid shell
interpretation.
Longer programs can be read in from a
file with the \-f option.
Data input is read from the list of files on
the command line or from standard input when the list is empty.
The input is broken into records as determined by the
record separator variable, \fBRS\fR. Initially,
.B RS
= "\\n" and records are synonymous with lines.
Each record is compared against each
.I pattern
and if it matches, the program text for
.I "{action}"
is executed.
.SH OPTIONS
.TP \w'\-\fBv'u+\w'\fIvar=value'u+2n
\-\fBF \fIvalue
sets the field separator, \fBFS\fR, to
.IR value .
.IP "\-\fBf \fIfile"
Program text is read from \fIfile\fR instead of from the
command line. Multiple \-f options are allowed.
.IP "\-\fBv \fIvar=value"
assigns
.I value
to program variable
.IR var .
.IP "\-\|\-"
indicates the unambiguous end of options.
.PP
The above options will be available with any Posix compatible
implementation of AWK, and implementation specific options are
prefaced with \-W.
.B mawk
provides three:
.TP \w'\-\fBv'u+\w'\fIvar=value'u+2n
\-\fBW \fRversion
.B mawk
writes its version and copyright
to stdout and compiled limits to
stderr and exits 0.
.TP
\-\fBW \fRdump
writes an assembler like listing of the internal
representation of the program to stderr.
.TP
\-\fBW \fRsprintf=\fInum
adjusts the size of
.B mawk's
internal sprintf buffer to
.I num
bytes. More than rare use of this option indicates
.B mawk
should be recompiled.
.TP
\-\fBW \fRposix_space
forces
.B mawk
not to consider '\\n' to be space.
.SH "THE AWK LANGUAGE"
.SS "\fB1. Program structure"
An AWK program is a sequence of
.I "pattern {action}"
pairs and user
function definitions.
.PP
A pattern can be:
.nf
.RS
\fBBEGIN
END\fR
expression
expression , expression
.sp
.RE
.fi
One, but not both,
of \fIpattern {action}\fR can be omitted. If
.I {action}
is omitted it is implicitly { print }. If
.I pattern
is omitted, then it is implicitly matched.
.B BEGIN
and
.B END
patterns require an action.
.PP
Statements are terminated by newlines, semi-colons or both.
Groups of statements such as
actions or loop bodies are blocked via { ... } as in C. The
last statement in a block doesn't need a terminator. Blank lines
have no meaning; an empty statement is terminated with a
semi-colon. Long statements
can be continued with a backslash, \\\|. A statement can be broken
without a backslash after a comma, left brace, &&, ||,
.BR do ,
.BR else ,
the right parenthesis of an
.BR if ,
.B while
or
.B for
statement, and the
right parenthesis of a function definition.
A comment starts with # and extends to, but does not include
the end of line.
.PP
The following statements control program flow inside blocks.
.RS
.PP
.B if
( \*(ex )
.I statement
.PP
.B if
( \*(ex )
.I statement
.B else
.I statement
.PP
.B while
( \*(ex )
.I statement
.PP
.B do
.I statement
.B while
( \*(ex )
.PP
.B for
(
\fIopt_expr\fR ;
\fIopt_expr\fR ;
\fIopt_expr\fR
)
.I statement
.PP
.B for
( \fIvar \fBin \fIarray\fR )
.I statement
.PP
.B continue
.PP
.B break
.RE
.\"
.SS "\fB2. Data types, conversion and comparison"
There are two basic data types, numeric and string.
Numeric constants can be integer like \-2,
decimal like 1.08, or in scientific notation like
\-1.1e4 or .28E\-3. All numbers are represented internally and all
computations are done in floating point arithmetic.
So for example, the expression
0.2e2 == 20
is true and true is represented as 1.0.
.PP
String constants are enclosed in double quotes.
.sp
.ce
"This is a string with a newline at the end.\\n"
.sp
Strings can be continued across a line by escaping (\\) the newline.
The following escape sequences are recognized.
.nf
.sp
\\\\ \\
\\" "
\\a alert, ascii 7
\\b backspace, ascii 8
\\t tab, ascii 9
\\n newline, ascii 10
\\v vertical tab, ascii 11
\\f formfeed, ascii 12
\\r carriage return, ascii 13
\\ddd 1, 2 or 3 octal digits for ascii ddd
\\xhh 1 or 2 hex digits for ascii hh
.sp
.fi
If you escape any other character \\c, you get \\c, i.e.,
.B mawk
ignores the escape.
.PP
There are really three basic data types; the third is
.I "number and string"
which has both a numeric value and a string value
at the same time.
User defined variables come into existence when first referenced
and are initialized to
.IR null ,
a number and string value which has numeric value 0 and string value
"".
Non-trivial number and string typed data come from input
and are typically stored in fields. (See section 4).
.PP
The type of an expression is determined by its context and automatic
type conversion occurs if needed. For example, to evaluate the
statements
.nf
.sp
y = x + 2 ; z = x "hello"
.sp
.fi
The value stored in variable y will be typed numeric.
If x is not numeric,
the value taken from x is converted to numeric before it is added to
2 and stored in y. The value stored in variable z will be typed
string, and the value of x will be converted to string if necessary
and concatenated with "hello". (Of course, the value and type
stored in x is not changed by any conversions.)
A string expression is converted to numeric using its longest
numeric prefix as with
.IR atof (3).
A numeric expression is converted to string by replacing
.I expr
with
.BR sprintf(CONVFMT ,
.IR expr ),
unless
.I expr
can be represented on the host machine as an exact integer then
it is converted to \fBsprintf\fR("%d", \*(ex).
.B Sprintf()
is an AWK built-in that duplicates the functionality of
.IR sprintf (3),
and
.B CONVFMT
is a built-in variable used for internal conversion
from number to string and initialized to "%.6g".
Explicit type conversions can be forced,
\*(ex ""
is string and
.IR expr +0
is numeric.
.PP
To evaluate,
\*(ex\d1\u \fBrel-op \*(ex\d2\u,
if both operands are numeric or number and string then the comparison
is numeric; if both operands are string the comparison is string.
If exactly one operand is string and after trimming spaces and
tabs from the front and back the remaining string is entirely
numeric in form, then the string is converted to number and the
comparison is numeric; otherwise, the numeric operand is converted
to string and the comparison is string.
The result of a comparison is numeric, 0 or 1.
.PP
In boolean contexts such as,
\fBif\fR ( \*(ex ) \fIstatement\fR,
a string expression evaluates true if and only if it is not the
empty string "";
numeric values if and only if not numerically zero.
.\"
.SS "\fB3. Regular expressions"
In the AWK language, records, fields and strings are often
tested for matching a
.IR "regular expression" .
Regular expressions are enclosed in slashes, and
.nf
.sp
\*(ex ~ /\fIr\fR/
.sp
.fi
is an AWK expression that evaluates to 1 if \*(ex "matches"
.IR r ,
which means a substring of \*(ex is in the set of strings
defined by
.IR r .
With no match the expression evaluates to 0; replacing
~ with the "not match" operator, !~ , reverses the meaning.
As pattern-action pairs,
.nf
.sp
/\fIr\fR/ { \fIaction\fR } and\
\fB$0\fR ~ /\fIr\fR/ { \fIaction\fR }
.sp
.fi
are the same,
and for each input record that matches
.IR r ,
.I action
is executed.
In fact, /\fIr\fR/ is an AWK expression that is
equivalent to (\fB$0\fR ~ /\fIr\fR/) anywhere except when on the
right side of a match operator or passed as an argument to
a built-in function that expects a regular expression
argument.
.PP
AWK uses extended regular expressions as with
.IR egrep (1).
The regular expression metacharacters, i.e., those with special
meaning in regular expressions are
.nf
.sp
\ ^ $ . [ ] | ( ) * + ?
.sp
.fi
Regular expressions are built up from characters as follows:
.RS
.TP \w'[^c\d1\uc\d2\uc\d3\u...]'u+1n
\fIc\fR
matches any non-metacharacter
.IR c .
.IP "\e\fIc\fR"
matches a character defined by the same escape sequences used
in string constants or the literal
character
.I c
if
\\\fIc\fR
is not an escape sequence.
.IP \.
matches any character (including newline).
.TP
^
matches the front of a string.
.TP
$
matches the back of a string.
.TP
[c\d1\uc\d2\uc\d3\u...]
matches any character in the class
c\d1\uc\d2\uc\d3\u... . An interval of characters is denoted
c\d1\u\-c\d2\u inside a class [...].
.TP
[^c\d1\uc\d2\uc\d3\u...]
matches any character not in the class
c\d1\uc\d2\uc\d3\u...
.RE
.sp
Regular expressions are built up from other regular expressions
as follows:
.RS
.TP
\fIr\fR\d1\u\fIr\fR\d2\u
matches
\fIr\fR\d1\u
followed immediately by
\fIr\fR\d2\u
(concatenation).
.TP
\fIr\fR\d1\u | \fIr\fR\d2\u
matches
\fIr\fR\d1\u or
\fIr\fR\d2\u
(alternation).
.TP
\fIr\fR*
matches \fIr\fR repeated zero or more times.
.TP
\fIr\fR+
matches \fIr\fR repeated one or more times.
.TP
\fIr\fR?
matches \fIr\fR zero or once.
.TP
(\fIr\fR)
matches \fIr\fR, providing grouping.
.RE
.sp
The increasing precedence of operators is alternation,
concatenation and
unary (*, + or ?).
.PP
For example,
.nf
.sp
/^[_a\-zA-Z][_a\-zA\-Z0\-9]*$/ and
/^[\-+]?([0\-9]+\\\|.?|\\\|.[0\-9])[0\-9]*([eE][\-+]?[0\-9]+)?$/
.sp
.fi
are matched by AWK identifiers and AWK numeric constants
respectively. Note that . has to be escaped to be
recognized as a decimal point, and that metacharacters are not
special inside character classes.
.PP
Any expression can be used on the right hand side of the ~ or !~
operators or
passed to a built-in that expects
a regular expression.
If needed, it is converted to string, and then interpreted
as a regular expression. For example,
.nf
.sp
BEGIN { identifier = "[_a\-zA\-Z][_a\-zA\-Z0\-9]*" }
$0 ~ "^" identifier
.sp
.fi
prints all lines that start with an AWK identifier.
.PP
.B mawk
recognizes the empty regular expression, //\|, which matches the
empty string and hence is matched by any string at the front,
back and between every character. For example,
.nf
.sp
echo abc | mawk { gsub(//, "X") ; print }
XaXbXcX
.sp
.fi
.\"
.SS "\fB4. Records and fields"
Records are read in one at a time, and stored in the
.I field
variable
.BR $0 .
The record is split into
.I fields
which are stored in
.BR $1 ,
.BR $2 ", ...,"
.BR $NF .
The built-in variable
.B NF
is set to the number of fields,
and
.B NR
and
.B FNR
are incremented by 1.
Fields above
.B $NF
are set to "".
.PP
Assignment to
.B $0
causes the fields and
.B NF
to be recomputed.
Assignment to
.B NF
or to a field
causes
.B $0
to be reconstructed by
concatenating the
.B $i's
separated by
.BR OFS .
Assignment to a field with index greater than
.BR NF ,
increases
.B NF
and causes
.B $0
to be reconstructed.
.PP
Data input stored in fields
is string, unless the entire field has numeric
form and then the type is number and string.
For example,
.sp
.nf
echo 24 24E |
mawk '{ print($1>100, $1>"100", $2>100, $2>"100") }'
0 0 1 1
.fi
.sp
.B $0
and
.B $2
are string and
.B $1
is number and string. The first
and second comparisons are numeric and the last
two are string. In the second "100" is
converted to 100, and in the third 100 is
converted to "100".
.\"
.SS "\fB5. Expressions and operators"
.PP
The expression syntax is
similar to C. Primary expressions are numeric constants,
string constants, variables, fields, arrays and functions.
The identifier
for a variable, array or function can be a sequence of
letters, digits and underscores, that does
not start with a digit.
Variables are not declared; they exist when first referenced and
are initialized to
.IR null .
.PP
New
expressions are composed with the following operators in
order of increasing precedence.
.PP
.RS
.nf
.vs +2p \" open up a little
\fIassignment\fR = += \-= *= /= %= ^=
\fIconditional\fR ? :
\fIlogical or\fR ||
\fIlogical and\fR &&
\fIarray membership\fR \fBin
\fImatching\fR ~ !~
\fIrelational\fR < > <= >= == !=
\fIconcatenation\fR (no explicit operator)
\fIadd ops\fR + \-
\fImul ops\fR * / %
\fIunary\fR + \-
\fIlogical not\fR !
\fIexponentiation\fR ^
\fIinc and dec\fR ++ \-\|\- (both post and pre)
\fIfield\fR $
.vs
.RE
.PP
.fi
Assignment, conditional and exponentiation associate right to
left; the other operators associate left to right. Any
expression can be parenthesized.
.\"
.SS "\fB6. Arrays"
.ds ae \fIarray\fR[\fIexpr\fR]
Awk provides one-dimensional arrays. Array elements are expressed
as \*(ae.
.I Expr
is internally converted to string type, so, for example,
A[1] and A["1"] are the same element and the actual
index is "1".
Arrays indexed by strings are called associative arrays.
Initially an array is empty; elements exist when first accessed.
An expression,
\fIexpr\fB in\fI array\fR
evaluates to 1 if
\*(ae
exists, else to 0.
.PP
There is a form of the
.B for
statement that loops over each index of an array.
.nf
.sp
\fBfor\fR ( \fIvar\fB in \fIarray \fR) \fIstatement\fR
.sp
.fi
sets
.I var
to each index of
.I array
and executes
.IR statement .
The order that
.I var
transverses the indices of
.I array
is not defined.
.PP
The statement,
.B delete
\*(ae,
causes
\*(ae
not to exist.
.PP
Multidimensional arrays are synthesized with concatenation using
the built-in variable
.BR SUBSEP .
\fIarray\fR[\fIexpr\fR\d1\u,\|\fIexpr\fR\d2\u]
is equivalent to
\fIarray\fR[\fIexpr\fR\d1\u \fBSUBSEP \fIexpr\fR\d2\u].
Testing for a multidimensional element uses a parenthesized index,
such as
.sp
.nf
if ( (i, j) in A ) print A[i, j]
.fi
.sp
.\"
.SS "\fB7. Builtin-variables\fR"
.PP
The following variables are built-in and initialized before program
execution.
.RS
.TP \w'FILENAME'u+2n
.B ARGC
number of command line arguments.
.TP
.B ARGV
array of command line arguments, 0..ARGC-1.
.TP
.B CONVFMT
format for internal conversion of numbers to string,
initially = "%.6g".
.TP
.B ENVIRON
array indexed by environment variables. An environment string,
\fIvar=value\fR is stored as
\fBENVIRON\fR[\fIvar\fR] =
.IR value .
.TP
.B FILENAME
name of the current input file.
.TP
.B FNR
current record number in
.BR FILENAME .
.TP
.B FS
splits records into fields as a regular expression.
.TP
.B NF
number of fields in the current record.
.TP
.B NR
current record number in the total input stream.
.TP
.B OFMT
format for printing numbers; initially = "%.6g".
.TP
.B OFS
inserted between fields on output, initially = " ".
.TP
.B ORS
terminates each record on output, initially = "\\n".
.TP
.B RLENGTH
length set by the last call to the built-in function,
.BR match() .
.TP
.B RS
input record separator, initially = "\\n".
.TP
.B RSTART
index set by the last call to
.BR match() .
.TP
.B SUBSEP
used to build multiple array subscripts, initially = "\\034".
.RE
.\"
.SS "\fB8. Built-in functions"
String functions
.RS
.TP
gsub(\fIr,s,t\fR) gsub(\fIr,s\fR)
Global substitution, every match of regular expression
.I r
in variable
.I t
is replaced by string
.IR s .
The number of replacements is returned.
If
.I t
is omitted,
.B $0
is used. An & in the replacement string
.I s
is replaced by the matched substring of
.IR t .
\\& puts a literal & in the replacement string.
.TP
index(\fIs,t\fR)
If
.I t
is a substring of
.IR s ,
then the position where
.I t
starts is returned, else 0 is returned.
The first character of
.I s
is in position 1.
.TP
length(\fIs\fR) length()
Returns the length of string
.IR s ;
without an argument, returns the length of
.BR $0 .
.TP
match(\fIs,r\fR)
Returns the index of the first longest match of regular expression
.I r
in string
.IR s .
Returns 0 if no match.
As a side effect,
.B RSTART
is set to the return value.
.B RLENGTH
is set to the length of the match or \-1 if no match. If the
empty string is matched,
.B RLENGTH
is set to 0, and 1 is returned if the match is at the front, and
length(\fIs\fR)+1 is returned if the match is at the back.
.TP
split(\fIs,A,r\fR) split(\fIs,A\fR)
String
.I s
is split into fields by regular expression
.I r
and the fields are loaded into array
.IR A .
The number of fields
is returned. See section 11 below for more detail.
If
.I r
is omitted,
.B FS
is used.
.TP
sprintf(\fIformat,expr-list\fR)
Returns a string constructed from
.I expr-list
according to
.IR format .
See the description of printf() below.
.TP
sub(\fIr,s,t\fR) sub(\fIr,s\fR)
Single substitution, same as gsub() except at most one substitution.
.TP
substr(\fIs,i,n\fR) substr(\fIs,i\fR)
Returns the substring of string
.IR s ,
starting at index
.IR i ,
of length
.IR n .
If
.I n
is omitted, the suffix of
.IR s ,
starting at
.I i
is returned.
.TP
tolower(\fIs\fR)
Returns a copy of
.I s
with all upper case characters converted to lower case.
.TP
toupper(\fIs\fR)
Returns a copy of
.I s
with all lower case characters converted to upper case.
.RE
.PP
Arithmetic functions
.RS
.PP
.nf
atan2(\fIy,x\fR) Arctan of \fIy\fR/\fIx\fR between -\(*p and \(*p.
.PP
cos(\fIx\fR) Cosine function, \fIx\fR in radians.
.PP
exp(\fIx\fR) Exponential function.
.PP
int(\fIx\fR) Returns \fIx\fR truncated towards zero.
.PP
log(\fIx\fR) Natural logarithm.
.PP
rand() Returns a random number between zero and one.
.PP
sin(\fIx\fR) Sine function, \fIx\fR in radians.
.PP
sqrt(\fIx\fR) Returns square root of \fIx\fR.
.fi
.TP
srand(\fIexpr\fR) srand()
Seeds the random number generator, using the clock if
.I expr
is omitted, and returns the value of the previous seed.
.B mawk
seeds the random number generator from the clock at startup
so there is no real need to call srand(). Srand(\fIexpr\fR)
is useful for repeating pseudo random sequences.
.RE
.\"
.SS "\fB9. Input and output"
There are two output statements,
.B print
and
.BR printf .
.RS
.TP
print
writes
.B "$0 ORS"
to standard output.
.TP
print \*(ex\d1\u, \*(ex\d2\u, ..., \*(ex\dn\u
writes
\*(ex\d1\u \fBOFS \*(ex\d2\u \fBOFS\fR ... \*(ex\dn\u
.B ORS
to standard output. Numeric expressions are converted to
string with
.BR OFMT .
.TP
printf \fIformat, expr-list\fR
duplicates the printf C library function writing to standard output.
The complete ANSI C format specifications are recognized with
conversions %c, %d, %e, %E, %f, %g, %G,
%i, %o, %s, %u, %x, %X and %%,
and conversion qualifiers h and l.
.RE
.PP
The argument list to print or printf can optionally be enclosed in
parentheses.
Print formats numbers using
.B OFMT
or "%d" for exact integers.
"%c" with a numeric argument prints the corresponding 8 bit
character, with a string argument it prints the first character of
the string.
The output of print and printf can be redirected to a file or
command by appending >
.IR file ,
>>
.I file
or
|
.I command
to the end of the print statement.
Redirection opens
.I file
or
.I command
only once, subsequent redirections append to the already open stream.
By convention,
.B mawk
associates the filename "/dev/stderr" with stderr which allows
print and printf to be redirected to stderr.
.PP
The input function
.B getline
has the following variations.
.RS
.TP
getline
reads into
.BR $0 ,
updates the fields,
.BR NF ,
.B NR
and
.BR FNR .
.TP
getline < \fIfile\fR
reads into
.B $0
from \fIfile\fR,
updates the fields and
.BR NF .
.TP
getline \fIvar
reads the next record into
.IR var ,
updates
.B NR
and
.BR FNR .
.TP
getline \fIvar\fR < \fIfile
reads the next record of
.I file
into
.IR var .
.TP
\fI command\fR | getline
pipes a record from
.I command
into
.B $0
and updates the fields and
.BR NF .
.TP
\fI command\fR | getline \fIvar
pipes a record from
.I command
into
.IR var .
.RE
.PP
Getline returns 0 on end-of-file, \-1 on error, otherwise 1.
.PP
Commands on the end of pipes are executed by /bin/sh.
.PP
The function \fBclose\fR(\*(ex) closes the file or pipe
associated with
.IR expr .
Close returns 0 if
.I expr
is an open file,
the exit status if
.I expr
is a piped command, and -1 otherwise.
Close() is used to reread a file or command, make sure the other
end of an output pipe is finished or conserve file resources.
.PP
The function
\fBsystem\fR(\fIexpr\fR)
uses
/bin/sh
to execute
.I expr
and returns the exit status of the command
.IR expr .
Changes made to the
.B ENVIRON
array are not passed to commands executed with
.B system
or pipes.
.SS \fB10. User defined functions
The syntax for a user defined function is
.nf
.sp
\fBfunction\fR name( \fIargs\fR ) { \fIstatements\fR }
.sp
.fi
The function body can contain a return statement
.nf
.sp
\fBreturn\fI opt_expr\fR
.sp
.fi
A return statement is not required.
Function calls may be nested or recursive.
Functions are passed expressions by value
and arrays by reference.
Extra arguments serve as local variables
and are initialized to
.IR null .
For example, csplit(\fIs,\|A\fR) puts each character of
.I s
into array
.I A
and returns the length of
.IR s .
.nf
.sp
function csplit(s, A, n, i)
{
n = length(s)
for( i = 1 ; i <= n ; i++ ) A[i] = substr(s, i, 1)
return n
}
.sp
.fi
Putting extra space between passed arguments and local
variables is conventional.
Functions can be referenced before they are defined, but the
function name and the '(' of the arguments must touch to
avoid confusion with concatenation.
.\"
.SS "\fB11. Splitting strings, records and files"
Awk programs use the same algorithm to
split strings into arrays with split(), and records into fields
on
.BR FS .
.B mawk
uses essentially the same algorithm to split files into
records on
.BR RS .
.PP
Split(\fIexpr,\|A,\|sep\fR) works as follows:
.RS
.TP
(1)
If
.I sep
is omitted, it is replaced by
.BR FS .
.I Sep
can be an expression or regular expression. If it is an
expression of non-string type, it is converted to string.
.TP
(2)
If
.I sep
= " " (a single space),
then <SPACE> is trimmed from the front and back of
.IR expr ,
and
.I sep
becomes <SPACE>.
.B mawk
defines <SPACE> as the regular expression
/[\ \\t\\n]+/.
Otherwise
.I sep
is treated as a regular expression, except that meta-characters
are ignored for a string of length 1,
e.g.,
split(x, A, "*") and split(x, A, /\\*/) are the same.
.TP
(3)
If \*(ex is not string, it is converted to string.
If \*(ex is then the empty string "", split() returns 0
and
.I A
is unchanged.
Otherwise,
all non-overlapping, non-null and longest matches of
.I sep
in
.IR expr ,
separate
.I expr
into fields which are loaded into
.IR A .
The fields are placed in
A[1], A[2], ..., A[n] and split() returns n, the number
of fields which is the number
of matches plus one.
Data placed in
.I A
that looks numeric is typed number and string.
.RE
.PP
Splitting records into fields works the same except the
pieces are loaded into
.BR $1 ,
\fB$2\fR,...,
.BR $NF .
If
.B $0
is empty,
.B NF
is set to 0 and all
.B $i
to "".
.PP
.B mawk
splits files into records by the same algorithm, but with the
slight difference that
.B RS
is really a terminator instead of a separator.
(\fBORS\fR is really a terminator too).
.RS
.PP
E.g., if
.B FS
= ":+" and
.B $0
= "a::b:" , then
.B NF
= 3 and
.B $1
= "a",
.B $2
= "b" and
.B $3
= "", but
if "a::b:" is the contents of an input file and
.B RS
= ":+", then
there are two records "a" and "b".
.RE
.PP
.B RS
= " " is not special.
.\"
.SS "\fB12. Multi-line records"
Since
.B mawk
interprets
.B RS
as a regular expression, multi-line
records are easy. Setting
.B RS
= "\\n\\n+", makes one or more blank
lines separate records. If
.B FS
= " " (the default), then single
newlines, by the rules for <SPACE> above, become space and
single newlines are field separators.
.RS
.PP
For example, if a file is "a\ b\\nc\\n\\n",
.B RS
= "\\n\\n+" and
.B FS
= "\ ", then there is one record "a\ b\\nc" with three
fields "a", "b" and "c". Changing
.B FS
= "\\n", gives two
fields "a b" and "c"; changing
.B FS
= "", gives one field
identical to the record.
.RE
.PP
If you want lines with spaces or tabs to be considered blank,
set
.B RS
= "\\n([\ \\t]*\\n)+".
For compatibility with other awks, setting
.B RS
= "" has the same
effect as if blank lines are stripped from the
front and back of files and then records are determined as if
.B RS
= "\\n\\n+".
Posix requires that "\\n" always separates records when
.B RS
= "" regardless of the value of
.BR FS .
.B mawk
does not support this convention, because defining
"\\n" as <SPACE> makes it unnecessary.
.\"
.PP
Most of the time when you change
.B RS
for multi-line records, you
will also want to change
.B ORS
to "\\n\\n" so the record spacing is preserved on output.
.\"
.SS "\fB13. Program execution"
This section describes the order of program execution.
First
.B ARGC
is set to the total number of command line arguments passed to
the execution phase of the program.
.B ARGV[0]
is set the name of the AWK interpreter and
\fBARGV[1]\fR ...
.B ARGV[ARGC-1]
holds the remaining command line arguments exclusive of
options and program source.
For example with
.nf
.sp
mawk \-f prog v=1 A t=hello B
.sp
.fi
.B ARGC
= 5 with
.B ARGV[0]
= "mawk",
.B ARGV[1]
= "v=1",
.B ARGV[2]
= "A",
.B ARGV[3]
= "t=hello" and
.B ARGV[4]
= "B".
Next, each
.B BEGIN
block is executed in order.
If the program consists
entirely of
.B BEGIN
blocks, then execution terminates, else
an input stream is opened and execution continues.
If
.B ARGC
equals 1,
the input stream is set to stdin,
else the command line arguments
.BR ARGV[1] " ...
.B ARGV[ARGC-1]
are examined for a file argument.
.PP
The command line arguments divide into three sets:
file arguments, assignment arguments and empty strings "".
An assignment has the form
\fIvar\fR=\fIstring\fR.
When an
.B ARGV[i]
is examined as a possible file argument,
if it is empty it is skipped;
if it is an assignment argument, the assignment to
.I var
takes place and
.B i
skips to the next argument;
else
.B ARGV[i]
is opened for input.
If it fails to open, execution terminates with exit code 1.
If no command line argument is a file argument, then input
comes from stdin.
Getline in a
.B BEGIN
action opens input. "\-" as a file argument denotes stdin.
.PP
Once an input stream is open, each input record is tested
against each
.IR pattern ,
and if it matches, the associated
.I action
is executed.
An expression pattern matches if it is boolean true (see
the end of section 2).
A
.B BEGIN
pattern matches before any input has been read, and
an
.B END
pattern matches after all input has been read.
A range pattern,
\fIexpr\fR1,\|\fIexpr\fR2 ,
matches every record between the match of
.IR expr 1
and the match
.IR expr 2
inclusively.
.PP
When end of file occurs on the input stream, the remaining
command line arguments are examined for a file argument, and
if there is one it is opened, else the
.B END
.I pattern
is considered matched
and all
.B END
.I actions
are executed.
.PP
In the example, the assignment
v=1
takes place after the
.B BEGIN
.I actions
are executed, and
the data placed in
v
is typed number and string.
Input is then read from file A.
On end of file A,
t
is set to the string "hello",
and B is opened for input.
On end of file B, the
.B END
.I actions
are executed.
.PP
Program flow at the
.I pattern
.I {action}
level can be changed with the
.nf
.sp
\fBnext\fR and
\fBexit \fIopt_expr\fR
.sp
.fi
statements.
A
.B next
statement
causes the next input record to be read and pattern testing
to restart with the first
.I "pattern {action}"
pair in the program.
An
.B exit
statement
causes immediate execution of the
.B END
actions or program termination if there are none or
if the
.B exit
occurs in an
.B END
action.
The
.I opt_expr
sets the exit value of the program unless overridden by
a later
.B exit
or subsequent error.
.SH EXAMPLES
.nf
1. emulate cat.
{ print }
2. emulate wc.
{ chars += length($0) + 1 # add one for the \\n
words += NF
}
END{ print NR, words, chars }
3. count the number of unique "real words".
BEGIN { FS = "[^A-Za-z]+" }
{ for(i = 1 ; i <= NF ; i++) word[$i] = "" }
END { delete word[""]
for ( i in word ) cnt++
print cnt
}
.fi
4. sum the second field of
every record based on the first field.
.nf
$1 ~ /credit\||\|gain/ { sum += $2 }
$1 ~ /debit\||\|loss/ { sum \-= $2 }
END { print sum }
5. sort a file, comparing as string
{ line[NR] = $0 "" } # make sure of comparison type
# in case some lines look numeric
END { isort(line, NR)
for(i = 1 ; i <= NR ; i++) print line[i]
}
#insertion sort of A[1..n]
function isort( A, n, i, j, hold)
{
for( i = 2 ; i <= n ; i++)
{
hold = A[j = i]
while ( A[j\-1] > hold )
{ j\-\|\- ; A[j+1] = A[j] }
A[j] = hold
}
# sentinel A[0] = "" will be created if needed
}
.fi
.SH "COMPATIBILITY ISSUES"
The Posix 1003.2(draft 11.2) definition of the AWK language
is AWK as described in the AWK book with a few extensions
that appeared in SystemVR4 nawk. The extensions are:
.sp
.RS
New functions: toupper() and tolower().
New variables: ENVIRON[\|] and CONVFMT.
ANSI C conversion specifications for printf() and sprintf().
New command options: \-v var=value, multiple -f options and
implementation options as arguments to \-W.
.RE
.sp
Posix AWK is oriented to operate on files a line at
a time.
.B RS
can be changed from "\\n" to another single character,
but it
is hard to find any use for this \(em there are no
examples in the AWK book.
By convention, \fBRS\fR = "", makes one or more blank lines
separate records, allowing multi-line records. When
\fBRS\fR = "", "\\n" is always a field separator
regardless of the value in
.BR FS .
.PP
.BR mawk ,
on the other hand,
allows
.B RS
to be a regular expression.
When "\\n" appears in records, it is treated as space, and
.B FS
always determines fields.
.PP
Removing the line at a time paradigm can make some programs
simpler and can
often improve performance. For example,
redoing example 3 from above,
.nf
.sp
BEGIN { RS = "[^A-Za-z]+" }
{ word[ $0 ] = "" }
END { delete word[ "" ]
for( i in word ) cnt++
print cnt
}
.sp
.fi
counts the number of unique words by making each word a record.
On moderate size files,
.B mawk
executes twice as fast, because of the simplified inner loop.
.PP
The following program replaces each comment by a single space in
a C program file,
.nf
.sp
BEGIN {
RS = "/\|\\*([^*]\||\|\\*+[^/*])*\\*+/"
# comment is record separator
ORS = " "
getline hold
}
{ print hold ; hold = $0 }
END { printf "%s" , hold }
.sp
.fi
Buffering one record is needed to avoid terminating the last
record with a space.
.PP
With
.BR mawk ,
the following are all equivalent,
.nf
.sp
x ~ /a\\+b/ x ~ "a\\+b" x ~ "a\\\\+b"
.sp
.fi
The strings get scanned twice, once as string and once as
regular expression. On the string scan,
.B mawk
ignores the escape on non-escape characters while the AWK
book advocates
.I \ec
be recognized as
.I c
which necessitates the double escaping of meta-characters in
strings.
Posix explicitly declines to define the behavior which passively
forces programs that must run under a variety of awks to use
the more portable but less readable, double escape.
.PP
Posix AWK does not recognize "/dev/stderr" or \\x hex escape
sequences in strings. Unlike ANSI C,
.B mawk
limits the number of digits that follows \\x to two.
.PP
Finally, here is how
.B mawk
handles exceptional cases not discussed in the
AWK book or the Posix draft. It is unsafe to assume
consistency across awks and safe to skip to
the next section.
.PP
.RS
substr(s, i, n) returns the characters of s in the intersection
of the closed interval [1, length(s)] and the half-open interval
[i, i+n). When this intersection is empty, the empty string is
returned; so substr("ABC", 1, 0) = "" and
substr("ABC", \-4, 6) = "A".
Every string, including the empty string, matches the empty string
at the
front so, s ~ // and s ~ "", are always 1 as is match(s, //) and
match(s, ""). The last two set
.B RLENGTH
to 0.
index(s, t) is always the same as match(s, t1) where t1 is the
same as t with metacharacters escaped. Hence consistency
with match requires that
index(s, "") always returns 1.
Also the condition, index(s,t) != 0 if and only t is a substring
of s, requires index("","") = 1.
If getline encounters end of file, getline var, leaves var
unchanged. Similarly, on entry to the
.B END
actions,
.BR $0 ,
the fields and
.B NF
have their value unaltered from the last record.
.SH SEE ALSO
.I egrep
(1)
.PP
Aho, Kernighan and Weinberger,
.IR "The AWK Programming Language" ,
Addison-Wesley Publishing, 1988, (the AWK book),
defines the language, opening with a tutorial
and advancing to many interesting programs that delve into
issues of software design and analysis relevant to programming
in any language.
.PP
.IR "The GAWK Manual" ,
The Free Software Foundation, 1991, is a tutorial
and language reference
that does not attempt the depth of the AWK book
and assumes the reader may be a novice programmer.
The section on AWK arrays is excellent. It also
discusses Posix requirements for AWK.
.SH BUGS
.B mawk
cannot handle ascii NUL \\0 in the source or data files. You
can output NUL using printf with %c, and any other 8 bit
character is acceptable input.
.B mawk
implements printf() and sprintf() using the C library functions,
printf and sprintf, so full ANSI compatibility requires an ANSI
C library. In practice this means the h conversion qualifier may
not be available. Also
.B mawk
inherits any bugs or limitations of the library functions.
Implementors of the AWK language have shown a consistent lack
of imagination when naming their programs.
.SH AUTHOR
Mike Brennan (brennan@boeing.com).