Pattern matching rules

• Welcome to linuxzoo
  • Essential Linux
    • File Manipulation
      • Pattern matching rules

You can use special characters to cause a pattern to match various strings, in much the same way you use wildcard characters in file names to match a variety of filenames.

• . - The period matches any single character. 'new.' matches 'news', 'newt', 'new1'.
• [] - Match any of the single characters in the brackets in a string. '[wt]alk' matches 'walk' or 'talk', anywhere the four letter sequence occurs in the file.
• * - Will match all characters. If you use 'm*a', you will match every line with a 'm' appearing before a 'a'. Using a '.*' would be more specific for example: 'Wh.*ed' would match 'When the internet was designed'.
• ^ - Will only match a string at the beginning of a line. For example, '^split'.
• $ - Will only match a string at the end of a line. For example, 'routing$'.

Examples of grep usage:

% grep '.outing' demofile
% grep 'RFC-97[34]' demofile
% grep -i 'rfc-97[34]' demofile       (-i mean 'case-insensitive')
% grep 'Tw.*on' demofile
% grep 'inf.*' demofile
% grep '^network' demofile
% grep 'routing$' demofile

Common options for the grep command:

• -c : Only COUNT the matching lines
• -n : Include line numbers in the output
• -v : Produce all lines EXCEPT those matching
• -i : ignore case when matching strings

You should also note that you can exclude characters with a combination of the caret ^ and square brackets []. For example:

% grep 'J[aei]' names
James
Jane
Jess
Jim

% grep 'J[^aei]' names
Jock
John
Jordan

Linuxzoo created by Gordon Russell.
@ Copyright 2004-2024 Edinburgh Napier University