Conditionals

condition ? actionA : actionB
# equivalent multi-branch form:
if (cond1) {
    stmt1
} else if (cond2) {
    stmt2
} else {
    stmt3
}

Looping Constructs

While loop:

while (cond)
    statement

Do-while loop:

do
    statement
while (cond)

For loop:

for (init; test; step)
    statement

Filter Lines by Field Value Range

Print entries from /etc/passwd where the third colon-separated field (numeric UID) lies between 50 and 100:

BEGIN { FS=":" }
($3 > 50 && $3 < 100) { print }

Classify Users by UID

File uid_class.awk:

BEGIN { FS=":" }
{
    if ($3 < 50)
        printf "%-20s %-20s %-10d\n", "UID<50", $1, $3
    else if ($3 > 50 && $3 < 100)
        printf "%-20s %-20s %-10d\n", "50<UID<100", $1, $3
    else
        printf "%-20s %-20s %-10d\n", "UID>100", $1, $3
}

Run with:

awk -f uid_class.awk passwd

Compute and Filter Student Averages

Input scores.txt:

Allen    80    90    96    98
Mike    93    98    92    91
Zhang    78    76    87    92
Jerry    86    89    68    92
Han    85    95    75    90
Li    78    88    98    100

Print header and averages:

BEGIN {
    printf "%-20s %-20s %-20s %-20s %-20s %-20s\n",
           "Name","Chinese","English","Math","Physical","Average"
}
{
    total = $2 + $3 + $4 + $5
    mean = total / 4
    printf "%-20s %-20d %-20d %-20d %-20d %-0.2f\n", $1, $2, $3, $4, $5, mean
}

Add condition to show only means above 90:

BEGIN {
    printf "%-20s %-20s %-20s %-20s %-20s %-20s\n",
           "Name","Chinese","English","Math","Physical","Average"
}
{
    total = $2 + $3 + $4 + $5
    mean = total / 4
    if (mean > 90)
        printf "%-20s %-20d %-20d %-20d %-20d %-0.2f\n", $1, $2, $3, $4, $5, mean
}

Summing 1 to 100 Using Different Loops

While version – file sum_while.awk:

BEGIN {
    idx = 1
    total = 0
    while (idx <= 100) {
        total += idx
        idx++
    }
    print total
}

Do-while version – file sum_dowhile.awk:

BEGIN {
    idx = 1
    total = 0
    do {
        total += idx
        idx++
    } while (idx <= 100)
    print total
}

For version – file sum_for.awk:

BEGIN {
    total = 0
    for (idx = 1; idx <= 100; idx++)
        total += idx
    print total
}

String Function Examples

Field length per line – file field_len.awk:

BEGIN { FS=":" }
{
    pos = 1
    while (pos <= NF) {
        sep = (pos == NF) ? "" : ":"
        printf "%d%s", length($pos), sep
        pos++
    }
    print ""
}

Usage:

awk -f field_len.awk passwd

Find substring position:

BEGIN {
    txt = "I have a gream"
    printf "%d\n", index(txt, "ea")
}

Convert to lowercase:

BEGIN {
    txt = "Hadoop is a bigdata Framework"
    print tolower(txt)
}

Convert to uppercase:

BEGIN {
    txt = "Hadoop is a bigdata Framework"
    print toupper(txt)
}

Split string into array:

BEGIN {
    txt = "Hadoop Kafka Spark Storm HDFS YARN Zookeeper"
    n = split(txt, parts, " ")
    for (j = 1; j <= n; j++)
        print parts[j]
}

Locate first digit:

BEGIN {
    txt = "Transaction 2345 Start:Select * from master"
    print match(txt, /[0-9]/)
}

Extract substring:

BEGIN {
    txt = "transaction start"
    print substr(txt, 4, 5)
}

Replace first numeric sequence:

BEGIN {
    txt = "Transaction 243 Start,Event ID:9002"
    reps = sub(/[0-9]+/, "$", txt)
    print reps
    print txt
}

Replace all numeric sequences:

BEGIN {
    txt = "Transaction 243 Start,Event ID:9002"
    reps = gsub(/[0-9]+/, "$", txt)
    print reps
    print txt
}

Array indexing starts at 1:

BEGIN {
    txt = "Hadoop Kafka Spark Storm HDFS YARN Zookeeper"
    split(txt, elems, " ")
    print elems[1]   # first element
    for (k in elems)
        print elems[k]
}