Grep examples: Perl-Compatible Regular Expressions

All examples run under grep version 3 on Linux (ubuntu)

Match regex

Return the lines containing the given regex in the given file(s)

Example: lines that contain "foobar" followed by a digit:

$ grep -P 'foo\d+' myfile.txt

Matched 2 lines that contain the pattern

Match whole line

Return the lines that fully match the given regex

Use --line-regexp modifier

Example: lines that fully match "foobar" followed by a digit:

$ grep --line-regexp -P 'foo\d+' myfile.txt

Only the line that fully matches the regex is returned

Not matching regex

Return lines not containing the given regex.

Use -v modifier

Example: lines that do not contain "foo" followed by digits

$ grep -v -P 'foo\d+' myfile.txt 

Match raw string

To match a fixed string (i.e. don't interpret any characters as special characters or modifiers)

Use -F modifier

$ grep -F '()' myfile-special-chars.txt

OR operator

A simple | between patterns:

Example: lines starting with "foo" OR ending in "bar":

$ grep -P '^foo|bar$' myfile.txt

Simple OR behaviour

Repeat N times

Use {N}

Example: lines containing exactly 2 digits:

$ grep -P '\d{2}' myfile.txt

Only lines with exactly 2 digits are a match

Character classes

Assuming perl-compatible regex (i.e. using the -P modifier)

Escape characters

To escape any of the special characters .?*+{|()[\^$ use a backslash \:

Example: match a literal '(' character

$ grep -P '\(' myfile.txt

Matched a literal (

Search 2 files

Use -- at the end of the pattern and pass the file names:

$ grep -P 'foo.+' -- file1.txt file2.txt

Use -- after the pattern to match multiple files

Type comparison

grep: invalid option -- P

PCRE is not available on BSD grep (such as the MacOS distribution).

Use extended regexes (-E) instead: Grep examples: Extended Regular Expressions