OFS, ORS Output field / rec.

separator (def: one space, \n) Operators

awk Quick Ref compiled by v.ledos
rel 1.0 feb-2010 RESTART, Start / Length of string matched by
RLENGTH match function (see below) && || ! Logical operators. Ex: !($2<4 || $3<20)
SUBSEP Subscript separator (default: \034) < <= == != >= > Comparing operators
Usage ~ !~ matched by, not
selector?if-true-exp:if-false-exp
awk [-v var=val] 'program' [file1 file2...] Main built-in functions
awk [-v var=val] -f progfile [file1 file2...]
r: regex ; s,t: strings ; n,p: integers Basic programs
Structure of an awk program int(n), sqrt(n), exp(n), log(n),
# comments sin(n), cos(n) { print NR, $0 } Precede each line by line #
pattern { action } A sequence of rand() Random number between 0 and 1 { $1 = NR; print } Replace first field by line #
pattern { action } pattern-action { $2 = log($2); $3 =”” ; print }
statements close(file or command)
… Replace the 2nd field by its logarithm, zap field 3
getline [var] Read next line from input file,
NF > 0 Print non-empty lines
getline [var] < file from a specific file,
For each file, NF > 0 {print $1, $NF}
command | getline [var] or from a pipe
For each input line, Print first field and last one of non-empty lines
Return 1 (record found), 0 (end of file), -1 (error)
For each pattern, NF > 4 Print records containing more than 4 fields
gsub(r,s) Substitute s for r globally in $0 / string t;
If pattern matches input line, do the action. $NF > 4 Print if last field greater than 4
gsub(r,s,t) return # of subs made
index(s,t) NR%2==0 Print even-numbered lines
Return first position of string t in s, or 0
"pattern" NR==10, NR==20 Print lines 10 to 20
if t is not present
BEGIN : executes “ action” before starting to view the input file /start/, /end / Print lines between patterns
length(s) Return number of characters in s
END : executes “ action” after ending to view the input file match(s,r) /regex/, EOF Print from pattern to end of file
Test whether s contains a substring
Other : regular, numeric or string expression or combination matched by r; return index or 0; sets /regex/ {print $1}
RSTART and RLENGTH Print first field of lines matching regex
"action" is executable code split(s,a) Split s into array a on FS / field $1 ~ /regex/ Print lines where first field matches
if (expression) statement1 else statement2 split(s,a,fs) separaror fs; return # of fields ORS=NR%5?”,”:”\n”
while (expression) statement sprintf(fmt,expr-list) Concatenate every 5 lines of input, using comma separator
for (expr1;expr2;expr3) statement /regex/ {x++}
Return expr-list formatted according to format string fmt Count and print the number
do statement while (expression) sub(r,s) END {print x} of lines matching /regex/
Substitute s for the leftmost longest
break / continue : immediately leave / start next iteration sub(r,s,t) substring of $0 / t matched by r; return # { nc += length($0) + 1; nw += NF }
of innermost enclosing loop of subs made END { print NR, "lines", nw, "words", nc,
exit / exit expression : go immediately to the END substr(s,p) Return substring of s (of length n) "characters" } wc command
action; if within the END action, exit program substr(s,p,n) starting at position p { sum += $1 } Print sum and
tolower(s), toupper(s) Lower and upper cases END { print sum, sum/NR } average
Built-in variables { x[NR] = $0 }
END {for (i = NR; i > 0; i--) print x[i]}
Formatted output Reverse a file
$0 Whole line,
$1, $2 … $NF first, second… last field { a[$1] += $2 }
{ printf (“FORMAT”,value1,value2,value3,…) } END { for (i in a) print (i,":",a[i]) }
ARGC Number of command line arguments
ARGV Array of command line arguments Group by field 1, and sum field 2
%c %s Print as character, as string function pwr(a,b) { return exp(b*log(a)) }
FILENAME Name of current input file
%-8s Print as 8 characters, left aligned NF >= 2 { print pwr($1,$2) }
FS, RS Input field / record separator (def: one space, \n)
%f Print as float number, User defined function
NF Number of fields in current record
%6.2f with 6 digits (4 as integer, 2 as decimal) BEGIN { RS=””; FS=”\n” } Multi-line records.
NR, FNR Number of record read so far / in current file
\n Line feed and carriage return { print “Name: “,$1 Leading and trailing
OFMT Output format for numbers (default: %.6g)
print “Address: “,$2 } newlines are ignored