CODEWITHSUNDEEP
regex
badu
badu

Reputation: 1079

use of colon symbol in regular expression

I am new to regex. I am studying it in regularexperssion.com. The question is that I need to know what is the use of a colon (:) in regular expressions.

For example:

$pattern = '/^(([\w]+:)?\/\/)?(([\d\w]|%[a-fA-f\d]{2,2})+(:([\d\w]|%[a-fA-f\d]{2,2})+)?@)?([\d\w][-\d\w]{0,253}[\d\w]\.)+[\w]{2,4}(:[\d]+)?(\/([-+_~.\d\w]|%[a-fA-f\d]{2,2})*)*(\?(&?([-+_~.\d\w]|%[a-fA-f\d]{2,2})=?)*)?(#([-+_~.\d\w]|%[a-fA-f\d]{2,2})*)?$/';

which matches:

$url1  = "http://www.somewebsite.com";
$url2  = "https://www.somewebsite.com";
$url3  = "https://somewebsite.com";
$url4  = "www.somewebsite.com";
$url5  = "somewebsite.com";

Yeah, any help would be greatly appreciated.

Upvotes: 57

Views: 119872

Answers (4)

Explosion Pills
Explosion Pills

Reputation: 191779

I've decided to go you one better and explain the entire regex:

^                 # anchor to start of line
(                 # start grouping
 (                # start grouping
  [\w]+           # at least one of 0-9a-zA-Z_
  :               # a literal colon
 )                # end grouping
 ?                # this grouping is optional
 \/\/             # two literal slashes
)                 # end capture
?                 # this grouping is optional
(
 (
  [\d\w]          # exactly one of 0-9a-zA-Z_
                  # having \d is redundant
  |               # alternation
  %               # literal % sign
  [a-fA-f\d]{2,2} # exactly 2 hexadecimal digits
                  # should probably be A-F
                  # using {2} would have sufficed
 )+               # at least one of these groups
 (                # start grouping
  :               # literal colon
  (
   [\d\w]
   |
   %
   [a-fA-f\d]{2,2}
  )+
 )?               # Same grouping, but it is optional
                  # and there can be only one
 @                # literal @ sign
)?                # this group is optional
(
 [\d\w]           # same as [\w], explained above
 [-\d\w]{0,253}   # includes a dash (-) as a valid character
                  # between 0 and 253 of these characters
 [\d\w]           # end with \w.  They want at most 255
                  # total and - cannot be at the start
                  # or end
 \.               # literal period
)+                # at least one of these groups
[\w]{2,4}         # two to four \w characters
(
 :                # literal colon
 [\d]+            # at least one digit
)?
(
 \/               # literal slash
 (
  [-+_~.\d\w]    # one of these characters
  |              # *or*
  %              # % with two hex digit combo
  [a-fA-f\d]{2,2}
 )*              # zero or more of these groups
)*               # zero or more of these groups
(
 \?              # literal question mark
 (
  &?         # literal &amp or & (semicolon optional)
  (
   [-+_~.\d\w]
   |
   %
   [a-fA-f\d]{2,2}
  )
  =?             # optional literal =
 )*              # zero or more of this group
)?               # this group is optional
(
 #               # literal #
 (
  [-+_~.\d\w]
  |
  %
  [a-fA-f\d]{2,2}
 )*
)?
$                # anchor to end of line

It's important to understand what the metacharacters/sequences are. Some sequences are not meta when used in certain contexts (especially a character class). I've cataloged them for you:

meta with no context

  • ^ -- zero width start of line
  • () -- grouping/capture
  • ? -- zero or one of the preceding sequence
  • + -- one or more of the preceding sequence
  • * -- zero or more of the preceding sequence
  • [] -- character class
  • \w -- alphanumeric characters and _. Opposite of \W
  • | -- alternation
  • {} -- length assertion
  • $ -- zero width end of line

This excludes :, @, and % from having any special/meta meaning in the raw context.

meta inside character class

] ends the character class. - creates a range of characters unless it is at the start or the end of the character class or escaped with a backslash.

grouping assertions

A (? combination starts a grouping assertion. For example, (?: means group but do not capture. This means that in the regex /(?:a)/, it will match the string "a", but a is not captured for use in replacement or match groups as it would be from /(a)/.

? can also be used for lookahead/lookbehind assertions with ?=, ?!, ?<=, ?<!. (? followed by any sequence except what I mentioned in this section is just a literal ?.

Upvotes: 49

Igor Chubin
Igor Chubin

Reputation: 64603

Colon : is simply colon. It means nothing, except special cases like, for example, clustering without capturing (also known as a non-capturing group):

(?:pattern)

Also it can be used in character classes, for example:

[[:upper:]]

However, in your case colon is just a colon.

Special characters used in your regex:

In character class [-+_~.\d\w]:

  • - means -
  • + means +
  • _ means _
  • ~ means ~
  • . means .
  • \d means any digit
  • \w means any word character

These symbols have this meaning because they are used in a symbol class []. Without symbol class + and . have special meaning.

Other elements:

  • =? means = that can occur 0 or 1 times; in other words = that can occur or not, optional =.

Upvotes: 86

Kai Sassnowski
Kai Sassnowski

Reputation: 280

A colon has no special meaning in Regular Expressions, it just matches a literal colon.

[\w]+:

This just means any word character 1 or more times followed by a literal colon The brackets are actually not needed here. Square brackets are used to define a group of characters to match. So

[abcd]

means a single character of a, b, c, d

Upvotes: 1

zessx
zessx

Reputation: 68810

There is no special use for colon : in your case :

(([\w]+:)?\/\/)? will match http://, https://, ftp://...

You can find one special use for colon : every capturing group starting by (?: won't appear in the results.
Example, with "foobarbaz" in input :

  • /foo((bar)(baz))/ => { [1] => 'barbaz', [2] => 'bar', [3] => 'baz' }
  • /foo(?:(bar)(baz))/ => { [1] => 'bar', [2] => 'baz' }

Upvotes: 7

Related Questions