Morawk
Lecture 15
Michael J Wise
Urk … Morawk
L14 Yet more Awk - 2
General For Loop
• Awk also has a general for loop as we often need to
iterate over other things than just arrays
• Format is the same as C’s.
for(<initial expr> ; <continuation test> ; <update expr>)
<statement or statement block>
• If required, the <initial expression> sets up the loop, e.g.
by initializing a loop variable.
• Immediately after initialization and after each iteration
the <continuation test> is evaluated. If FALSE, the loop
terminates; otherwise computation continues to the
first/next iteration.
L14 Yet more Awk - 3
General For Loop
• At the end of each iteration, just before the <continuation
test>, the <update expression> is evaluated. It generally
updates some loop variable.
• Any of the three parts may be absent, but the structure
including the semicolons must be present
# Sum the numbers found across lines, down
# the file
BEGIN{ sum = 0}
{for(i=1; i<=NF; i++)
sum += $i
}
END{printf(”Total across the file: %d\n”, sum)}
L14 Yet more Awk - 4
Built-in Functions
• Awk has a number of very handy building functions,
including:
length(<string>) #return the length of the string
substr(s, m, n)
• substring of s, starting at m, with length n (or till end, if
n absent)
sprintf(fmt, expr, …)
• Like printf, but returns a string rather than printing.
Only way to concatenate strings within Awk
{cat = ”Sherlock”; weight = 7.9
cat = sprintf(”%s is a %0.1f kg cat”, cat, weight)
L14 Yet more Awk - 5
Built-in Functions
sub(r, t, s) # Substitute t for first regex r in s
gsub(r, t, s) # Substitute t for all regex r in s
• E.g. gsub(” *”, ” ”, $1) # multiple spaces
• These two commands allow you to do in Awk much
of what you can do in Sed
system(cmd)
• Execute Unix command cmd
• Returns the exit status
• Tactic can be to create a Unix command string which
results in a file, and then use getline to reimport
the new data back into Awk, e.g. sort
L14 Yet more Awk - 6
Demo
• my_cut.awk is a small Awk script that reports a
named column from a text file. Note the use of
command line option –v to import value
# report a column specified by command line variable
# e.g., -v COL=3 (default first column)
BEGIN{
if(COL == "")
COL = 1
}
NF >= COL {print $COL}
Example:
% gawk -v COL=3 -f my_cut.awk jab.txt
L14 Yet more Awk - 7
Demo
• Extracting a single column using Awk is straight-
forward (and more flexible than Cut).
• Write a Gawk script, excise_col.awk, which
excises a column and reports whatever is left
• Once again, the identity of the column is imported
by the command line argument, -v COL = …
• Hint: Needs two for loops
L14 Yet more Awk - 8
User Defined Functions
The format for Awk functions is:
function <name>(<parameter list>)
{
<statements>
}
L14 Yet more Awk - 9
User Defined Functions
• For example:
function max(a, b)
{
if(a>=b)
return(a)
return(b)
}
{print max($1, $2)}
L14 Yet more Awk - 10
About User Defined Functions
• The open-bracket for the parameter list must
immediately follow the name of the function (no
spaces between).
• If there are no parameters, the brackets must still
be present.
• The parameter list is a comma-separated list of
names.
– Variables in the parameter list are local to the
function (i.e. distinct from other variables found
in the awk-script).
– All other variables in a function definition are
global!
L14 Yet more Awk - 11
About User Defined Functions
• Scalar values, numbers, strings, etc, are copied to
function parameters (i.e. call-by-value).
• Arrays are passed to function parameters as call-by-
reference, i.e. a reference to the original array is
passed, so alterations to array elements apply to the
original array.
• Recursion is permitted.
return <expression>
• is used to return a value to the function caller
(otherwise the return value is undefined)
L14 Yet more Awk - 12
Why Have User Defined Functions?
• Literate programming (again)
– Think of the poor folk who will have to read your
code and maintain it
• Problem decomposition
• Saves retyping identical text, e.g. multiple calls to
max function
L14 Yet more Awk - 13
Demo
• Awk does not have max function, i.e. given an array,
return value of the largest element. Write a function
to compute max_array.
L14 Yet more Awk - 14
