Perl | Anchors in Regex

Anchors in Perl Regex do not match any character at all. Instead, they match a particular position as before, after, or between the characters. These are used to check not the string but its positional boundaries.
Following are the respective anchors in Perl Regex:

'^' '$', '\b', '\A', '\Z', '\z', '\G', '\p{....}', '\P{....}', '[:class:]'

^ or \A: It matches the pattern at the beginning of the string.

Syntax: (/^pattern/, /\Apattern/).

Example:

#!/usr/bin/perl
$str = "guardians of the galaxy";

# prints the pattern as it is
# starting with 'guardians'
print "$&\n" if($str =~ /^guardians/);

# prints the pattern 'gua'
print "$&\n" if($str =~ /\Agua/);

# prints nothing because
# the 0th position doesn't start with 'a'
print "$&" if($str =~ /^ans/)

Output:

guardians
gua

$ or \z: It matches the pattern at the end of the string.

Syntax: (/pattern$/, /pattern\z/).

Example:

#!/usr/bin/perl
$str = "guardians of the galaxy";

# prints nothing as it is not
# ending with 'guardians'
print "$&\n" if($str =~ /guardians$/);

# prints the pattern 'y'
print "$&\n" if($str =~ /y\z/);

# prints the pattern as it is
# ending with 'galaxy'
print "$&" if($str =~ /galaxy$/)

Output:

y
galaxy

\b: It matches at the word boundary of the string from \w to \W. In precise, it either gets a match to beginning or end of the string if it is a word or to a word character or a non-word character.

Syntax: (/\bpattern\b/).

Example:

#!/usr/bin/perl
$str = "guardians-of-the-galaxy";

# prints '-galaxy' as it forms
# a word even with '-'.
print "$&\n" if($str =~ /\b-galaxy\b/);

# prints '-guardians' as it forms
# a word even with '-'.
print "$&\n" if($str =~ /\bguardians-\b/);

# prints nothing as it is bounded
# with a character 't'.
print "$&" if($str =~ /\be-galaxy\b/);

# prints 'guardians-of-the-galaxy' as it
# is bounded with the beginning and end.
print "$&" if($str =~ /\bguardians-of-the-galaxy\b/);

Output:

-galaxy
guardians-
guardians-of-the-galaxy

\Z: It matches at the ending of the string or before the newline. ‘\z‘ and ‘\Z‘ both differ from $ in that they are not affected by the /m “multiline” flag, which allows $ to match at the end of any line.

#!/usr/bin/perl

# Prints one due to m//
print "one\n" if ('galaxy' =~ m/galaxy\z/);

# Prints two due to m//
print "two\n" if('galaxy' =~ m/galaxy\Z/);

# Prints three due to /Z
# as it forms a newline
print "three\n" if ("galaxy\n" =~ m/galaxy\Z/);

# Prints four due to m// as
# the line ended \z gets affected
print "four\n" if ("galaxy\n" =~ m/galaxy\n\z/);

# Prints five as it forms a new line
print "five\n" if("galaxy\n" =~ m/galaxy\n\Z/);

# Due to the "" it forms a newline and
# \z doesn't get affected. Prints nothing
print "six" if("galaxy\n" =~ m/galaxy\z/);

Output:

one
two
three
four
five

\G: It matches at the specified position. If a pattern’s length is 5 then it starts from the start of the string till 5 positions, if the pattern is valid then it is forced to check the string from 6th position onwards, moves forward in this fashion till pattern not valid or end of the string.

#!/usr/bin/perl
$str = "galaxy8222as";

# prints until the pattern is valid
print "one: $& " while($str =~ /\G[a-z]{2}/gc);
print "\n";

# prints until the pattern is valid
print "two: $& " while("1122a44" =~ /\G\d\d/gc);
print "\n";

# Take the string as a new value and
# searches from the start to false
print "three: $& " while("galaxy8222as" =~ /\G\w{2}/gc);
print "four: $& " while($str =~ /\G[a-z]{2}/gc);

# Take the false position of the
# above string and searches from there
# Prints if the pattern is valid from that position
# onwards(prints nothing). As it is false
# it stays at the same position as before.
print "\n";
print "five: $& " while($str =~ /\G\w{2}/gc);

Output:

one: ga one: la one: xy 
two: 11 two: 22 
three: ga three: la three: xy three: 82 three: 22 three: as 
five: 82 five: 22 five: as

\p{…} and \P{…}: \p{…} matches Unicode character class like IsLower, IsAlpha, etc. whereas \P{….} is the complement of Unicode character class.

#!/usr/bin/perl

# unicode class is the pattern to match
print "$&" while("guardians!@#%^*123" =~ /\p{isalpha}/gc);
print "\n";

# unicode class is the pattern to match
print "$&" while("guardians!@#%^&*123" =~ /\p{isalnum}/gc);
print "\n";

# here L matches the alphabets where \P is the complement
print "$&" while("guardians!@#%^&*123" =~ /\P{L}/gc);
print "\n";

# here L matches the alphabets where \p is non-complement
print "$&" while("guardians!@#%^&*123" =~ /\p{L}/gc);

Output:

guardians
guardians123
!@#%^&*123
guardians

[:class:]: POSIX Character Classes like digit, lower, ascii, etc.

Syntax: (/[[:class:]]/)

POSIX character classes are as follows:

alpha, alnum, ascii, blank, cntrl, digit, graph, lower, punct, space, upper, xdigit, word

#!/usr/bin/perl

# prints only alphabets
print "$&" while('guardians!@#%^&*123' =~ /[[:alpha:]]/gc);
print "\n";

# prints characters and digits
print "$&" while("guardians!@#%^&*123" =~ /[[:alnum:]]/gc);
print "\n";

# prints only digits
print "$&" while("guardians!@#%^&*123" =~ /[[:digit:]]/gc);
print "\n";

# prints anything except space " ".
print "$&" while("guardians!@#%^& 123\n" =~ /[[:graph:]]/gc);
print "\n";

# prints the 1 as it gets matched to
# space " " or horizontal tab.
print "1" while("guardians!@#%^& 123\n" =~ /[[:blank:]]/gc);
print "\n";

# prints lowercase characters
print "$&" while("Guardians!@#%^& 123\n" =~ /[[:lower:]]/gc);
print "\n";

# prints all ascii characters
print "$&" while("guardians!@#%^& 123\n" =~ /[[:ascii:]]/gc);

Output:

guardians
guardians123
123
guardians!@#%^&123
1
uardians
guardians!@#%^& 123