ruby-url-fail.rb

url = 'http://example.com/ ?api_token=a-token&username=USERNAME' # => "http://example.com/ ?api_token=a-token&username=USERNAME"
url =~ URI.regexp(%w[http https]) # => 0
URI.parse(url) # => URI::InvalidURIError: bad URI(is not URI?): http://example.com/ ?api_token=a-token&username=USERNAME

url = 'http://example.com/ ?api_token=a-token&username=USERNAME' # => "http://example.com/ ?api_token=a-token&username=USERNAME"
url.match(URI.regexp(%w[http https])).to_a.first == url # => false

url = 'http://example.com/?api_token=a-token&username=USERNAME' # => "http://example.com/?api_token=a-token&username=USERNAME"
url.match(URI.regexp(%w[http https])).to_a.first == url # => true

url = 'zztop://example.com/?api_token=a-token&username=USERNAME' # => "zztop://example.com/?api_token=a-token&username=USERNAME"
url.match(URI.regexp(%w[http https])).to_a.first == url # => false
```ruby

Ah, so basically URI.regexp seems to match all URIs in a given string and not just one. That explains something. Thanks for helping me see that!

I guess the proper solution would be to normalize the URI like xxx posted over at the HTTParty repo (jnunemaker/httparty#154) and then use the normalized URI which has spaces substituted with + or %20.

However, in my app I've just done it like this for now:

def validate_webhook_url
  return if webhook_url.blank?

  url = URI.parse(webhook_url) rescue nil

  errors.add(:webhook_url, :invalid) if url.blank? || url.scheme !~ /\Ahttps?\z/
end