TopolearnPatterns let one awk command replace a grep plus a cut, and END blocks turn a filter into a report.
Each awk rule is pattern { action }. The pattern decides which records the action runs on:
- a comparison:
$3 > 100,$1 == "alice",NF > 0; - a regex against the whole line:
/ERROR/, or against a field:$2 ~ /^api/,$2 !~ /test/; - combinations with
&&,||,!; - a range
/start/,/stop/, like in sed.
If the action is omitted, the default action prints the line, so awk '$3 > 100' behaves like a field-aware grep. If the pattern is omitted, the action runs for every line.
Two special patterns run outside the input loop: BEGIN { ... } runs once before any input is read (set FS, OFS, print a header) and END { ... } runs once after the last line (print totals). In END, NR holds the total number of records.
Variables need no declaration: they start as an empty string, which is also 0 in numeric context. That is why { sum += $2 } END { print sum } just works.
Comparisons are numeric when both sides look like numbers and string-based otherwise. Be careful with fields like 007 versus "7" written as a string constant: $1 == "7" compares as strings.
Use printf for aligned output: printf "%-10s %5.1f\n", $1, $2.
Example
printf 'name dept salary\nana eng 120\nbo ops 80\ncy eng 95\n' > staff.txt
awk 'NR > 1 && $3 > 90' staff.txt # ana eng 120 / cy eng 95
awk '$2 ~ /^e/ { print $1 }' staff.txt # ana / cy
awk 'BEGIN { print "Report" }
NR > 1 { total += $3; n++ }
END { printf "%d people, avg %.1f\n", n, total / n }' staff.txt
# Report
# 3 people, avg 98.3
awk 'END { print NR " lines" }' staff.txt # 4 linesRemember
- pattern { action }: no action prints the line, no pattern matches all lines
- ~ and !~ match a field against a regex
- BEGIN runs before input, END after it; uninitialised variables are 0 or empty
Watch out for
- Forgetting to skip the header line (NR > 1) and having it counted or summed.