How to escape two forward slashes in Ruby Regex [closed] - ruby

Question

Closed. This question needs details or clarity. It is not currently accepting answers.
Want to improve this question? Add details and clarify the problem by editing this post.
Closed 8 years ago.
Improve this question
I'm trying to make a regex which finds the domain name using Ruby, so I tried this:
(?<=.*/).(?=.*/)
On Rubular I always see this error message: Forward slashes must be escaped.
How do I solve this?

Answer 1

When you use the // regex literal, you need to escape / using a backslash as \/. When you want literal / in your regex, it is usually simpler to avoid using the // literal. For example, use %r literal with any delimiters that would not cause conflict.
%r{/}
By the way, Ruby onigmo regex engine does not allow look behind with variable length, so your regex will return an error anyway.

Answer 2

Don't reinvent wheels, especially ones that work:
require 'uri'
URI.split('http://user:passwd#www.example.com:81/path/to/index.html?foo=bar#baz')
# => ["http",
# "user:passwd",
# "www.example.com",
# "81",
# nil,
# "/path/to/index.html",
# nil,
# "foo=bar",
# "baz"]
Or:
require 'addressable/uri'
uri = Addressable::URI.parse('http://user:passwd#www.example.com:81/path/to/index.html?foo=bar#baz')
uri.authority # => "user:passwd#www.example.com:81"
uri.fragment # => "baz"
uri.host # => "www.example.com"
uri.password # => "passwd"
uri.path # => "/path/to/index.html"
uri.port # => 81
uri.query # => "foo=bar"
uri.query_values # => {"foo"=>"bar"}
uri.scheme # => "http"
uri.to_hash # => {:scheme=>"http", :user=>"user", :password=>"passwd", :host=>"www.example.com", :port=>81, :path=>"/path/to/index.html", :query=>"foo=bar", :fragment=>"baz"}
uri.user # => "user"
Between the two, Addressable::URI is more full-featured and follows the specs very closely. Ruby's built-in URI is good for lighter-weight lifting.
Root around in their code and you'll find the regular expressions used to tear apart a URL; You'll also see that they aren't trivial because URLs can be quite "interesting", where "interesting" means you'll scream and pull out your hair. See the URI RFC for more information. See "Parsing a URI Reference with a Regular Expression" in that document for a suggested pattern.
...I do a exercise from codewars and I not allowed to use require
First, if so, why are you asking for help on how to write this? You are supposed to figure these things out yourself.
That said, try what has already been created. This uses the pattern in the RFC:
URI_REGEX = %r!^(([^:/?#]+):)?(//([^/?#]*))?([^?#]*)(\?([^#]*))?(#(.*))?!
uri_captures = 'http://user:passwd#www.example.com:81/path/to/index.html?foo=bar#baz'.match(URI_REGEX).captures # !> assigned but unused variable - port
# => ["http:",
# "http",
# "//user:passwd#www.example.com:81",
# "user:passwd#www.example.com:81",
# "/path/to/index.html",
# "?foo=bar",
# "foo=bar",
# "#baz",
# "baz"]
user, passwd, host, port = uri_captures[3].split(/[:#]/)
host # => "www.example.com"
For further convenience, here's a simple pattern to provide named captures:
URI_REGEX = %r!^((?<scheme>[^:/?#]+):)?(//(?<authority>[^/?#]*))?(?<path>[^?#]*)(\?(?<query>[^#]*))?(?<fragment>#(.*))?!
uri_captures = 'http://user:passwd#www.example.com:81/path/to/index.html?foo=bar#baz'.match(URI_REGEX)
authority_captures = uri_captures['authority'].match(/(?<user>[^:]+)?:?(?<passwd>[^#]+)?#?(?<host>.+)(:(?<port>\d+)?)/)
authority_captures['host']
# => "www.example.com"