Glossary

Anchor — a pattern element that matches a position rather than characters: ^ (start of line) and $ (end of line). Chapter 2.

Associative array — awk's table type: an array indexed by strings, used to keep a count or total per category: count[$3]++. Chapter 7.

Back-reference — a reuse of text a pattern captured: \1 in a sed replacement stands for the first capture group. Chapter 5.

Basic regular expression (BRE) — the default pattern dialect of grep and sed, in which + ? | ( ) are literal unless escaped with a backslash. The extended dialect (ERE, grep's -E, sed's -E) makes them operators. Chapter 3.

Capture group — a part of a pattern wrapped so its matched text is remembered: \( … \) in the basic dialect, ( … ) in the extended. Chapter 5.

Delimiter — the character separating the parts of a sed substitution. A slash by habit, but whatever character follows the s — s|a|b| uses bars, which keeps paths readable. Chapter 4.

Exit code — the number a command leaves in $? reporting how it went. grep's are 0 (lines selected), 1 (none selected), 2 (an error). Chapter 3.

Field — one piece of a record after splitting: $1, $2, … in awk. $0 is the whole record and $NF the last field. Chapter 6.

Field separator (FS) — the character or pattern awk splits records on: whitespace by default, set with -F or in BEGIN. OFS is its output twin, used when printing. Chapters 6–7.

Fixed string — a pattern treated as plain text with no pattern characters at all: grep's -F. The right tool for text paved with dots. Chapter 2.

Global flag — the g after a sed substitution, meaning replace every match on the line instead of the first. Chapter 4.

Hold space — sed's spare storage for parking a line and retrieving it later. Named in chapter 5, deliberately not taught: the programs that need it read better in awk.

In-place editing — sed's -i, which writes the edited stream back over the input file. Travels with a backup suffix (-i.bak) in this guide. Chapter 5.

Line — the unit grep and sed work in: characters up to a newline. A file whose lines end in an invisible carriage return as well (Windows endings) is chapter 10's trap.

Pattern — a description of text to match: a plain word, or a regular expression. Also, in awk, the condition half of a pattern–action rule.

Pipeline — commands joined by |, each stage's output feeding the next stage's input. Chapter 8.

Record — awk's unit of input: one line, in everything this guide does. NR counts records overall, FNR within the current file, FILENAME names it. Chapters 6–7.

Regular expression — the pattern language: classes like [0-9], repetition, alternation, anchors. Borrowed lightly in chapter 3; taught fully by the sibling guide Regular Expressions for Mere Mortals.

Stream editor — sed's full name and job description: an editor that works on a stream of lines as they pass, printing the result. Chapter 4.

Whole-word match — grep's -w: the match must be bounded by non-word characters, so login doesn't match inside nologin. Chapter 2.