emails: automatically fix typos in email addresses.
Try to automatically fix various kind of typos and common mistakes in email addresses when a user creates a new account. It's common for users to signup with addresses like `name@gmai.com`, which leads to bounces when we try to send the welcome email.
This commit is contained in:
130
app/logical/danbooru/email_address.rb
Normal file
130
app/logical/danbooru/email_address.rb
Normal file
@@ -0,0 +1,130 @@
|
||||
# frozen_string_literal: true
|
||||
|
||||
# A utility class that represents an email address. A wrapper around Mail::Address
|
||||
# that adds extra utility methods for normalizing and validating email addresses.
|
||||
#
|
||||
# @see https://www.rubydoc.info/gems/mail/Mail/Address
|
||||
# @see app/logical/email_address_type.rb
|
||||
# @see config/initializers/types.rb
|
||||
module Danbooru
|
||||
class EmailAddress
|
||||
class Error < StandardError; end
|
||||
|
||||
# https://www.regular-expressions.info/email.html
|
||||
EMAIL_REGEX = /\A[a-z0-9._%+-]+@(?:[a-z0-9][a-z0-9-]{0,61}\.)+[a-z]{2,}\z/i
|
||||
|
||||
# @return [String] The original email address as a string.
|
||||
attr_reader :address
|
||||
|
||||
# @return [Mail::Address] The parsed email address.
|
||||
attr_reader :parsed_address
|
||||
|
||||
delegate :local, to: :parsed_address
|
||||
alias_method :name, :local
|
||||
alias_method :to_s, :address
|
||||
|
||||
# Parse a string into an email address, or raise an exception if the string is not a syntactically valid address.
|
||||
#
|
||||
# @param string [String, Danbooru::EmailAddress]
|
||||
def initialize(string)
|
||||
raise Error, "#{string} is not a valid email address" if !string.match?(EMAIL_REGEX)
|
||||
|
||||
@address = string.to_s
|
||||
@parsed_address = Mail::Address.new(parsed_address)
|
||||
end
|
||||
|
||||
# Parse a string into an email address, or return nil if the string is not a syntactically valid email address.
|
||||
#
|
||||
# @param url [String, Danbooru::EmailAddress]
|
||||
# @return [Danbooru::EmailAddress]
|
||||
def self.parse(address)
|
||||
new(address)
|
||||
rescue Error
|
||||
nil
|
||||
end
|
||||
|
||||
# Parse a string into an email address while attempting to fix common typos and mistakes, or return
|
||||
# nil if the string can't be normalized into a valid email address.
|
||||
#
|
||||
# @param address [String]
|
||||
# @return [Danbooru::EmailAddress]
|
||||
def self.normalize(address)
|
||||
address = address.gsub(/[[:space:]]+/, " ").strip
|
||||
|
||||
address = address.gsub(/[\\\/]$/, '') # @qq.com\ -> @qq.com, @web.de/ -> @web.de
|
||||
#address = address.gsub(/,/, ".") # foo,bar@gmail.com -> foo.bar@gmail.com | @gmail,com -> @gmail.com
|
||||
address = address.gsub(/^https?:\/\/(www\.)?/i, "") # https://xxx@gmail.com -> xxx@gmail.com
|
||||
address = address.gsub(/^mailto:/i, "") # mailto:foo@gmail.com -> foo@gmail.com
|
||||
address = address.gsub(/.* <(.*)>$/, '\1') # foo <bar@gmail.com> -> bar@gmail.com
|
||||
address = address.gsub(/@\./, "@") # @.gmail.com -> @gmail.com
|
||||
address = address.gsub(/@com$/i, ".com") # @gmail@com -> @gmail.com
|
||||
address = address.gsub(/\.co,$/i, '.com') # @gmail.co, -> @gmail.com
|
||||
address = address.gsub(/\.com.$/i, '.com') # @gmail.com, -> @gmail.com
|
||||
address = address.gsub(/\.con$/i, '.com') # @gmail.con -> @gmail.com
|
||||
address = address.gsub(/\.\.com$/i, '.com') # @gmail..com -> @gmail.com
|
||||
|
||||
# @gmail -> @gmail.com
|
||||
address = address.gsub(/@gmai$/i, "@gmail.com")
|
||||
address = address.gsub(/@gmail$/i, "@gmail.com")
|
||||
address = address.gsub(/@yahoo$/i, "@yahoo.com")
|
||||
address = address.gsub(/@hotmai$/i, "@hotmail.com")
|
||||
address = address.gsub(/@hotmail$/i, "@hotmail.com")
|
||||
address = address.gsub(/@hot[^m]ail$/i, "@hotmail.com")
|
||||
address = address.gsub(/@live$/i, "@live.com")
|
||||
|
||||
address = address.gsub(/@.gmail\.com$/i, "@gmail.com") # @-gmail.com -> @gmail.com
|
||||
address = address.gsub(/@g.ail\.com$/i, "@gmail.com") # @g,ail.com -> @gmail.com
|
||||
address = address.gsub(/@gmail\.co.$/i, "@gmail.com") # @gmail.co, -> @gmail.com
|
||||
address = address.gsub(/@gamil\.com$/i, "@gmail.com") # @gamil.com -> @gmail.com
|
||||
address = address.gsub(/@gnail\.com$/i, "@gmail.com") # @gnail.com -> @gmail.com
|
||||
address = address.gsub(/@gmail\.co$/i, "@gmail.com") # @gmail.co -> @gmail.com
|
||||
address = address.gsub(/@gmai.\.com$/i, "@gmail.com") # @gmai;.com -> @gmail.com
|
||||
address = address.gsub(/@gmai\.com$/i, "@gmail.com") # @gmai.com -> @gmail.com
|
||||
address = address.gsub(/@gmai\.co$/i, "@gmail.com") # @gmai.co -> @gmail.com
|
||||
address = address.gsub(/@hotmai\.com$/i, "@hotmail.com") # @hotmai.com -> @hotmail.com
|
||||
address = address.gsub(/@hot.ail\.com$/i, "@hotmail.com") # @hot.ail.com -> @hotmail.com
|
||||
address = address.gsub(/@hot.mail\.com$/i, "@hotmail.com") # @hot,mail.com -> @hotmail.com
|
||||
address = address.gsub(/@hanm.ail\.net$/i, "@hanmail.net") # @hanmiail.net -> @hanmail.net
|
||||
|
||||
address = address.gsub(/@(gmail|yahoo|hotmail|outlook|live).com$/i, '@\1.com') # @gmail,com -> @gmail.com
|
||||
address = address.gsub(/@(gmail|yahoo|hotmail|outlook|live)com$/i, '@\1.com') # @gmailcom -> @gmail.com
|
||||
|
||||
address = address.gsub(/@([a-z]+)\.com@\1\.com$/i, '@\1.com') # @gmail.com@gmail.com -> @gmail.com
|
||||
address = address.gsub(/@([a-z]+)@\1\.com$/i, '@\1.com') # @gmail@gmail.com -> @gmail.com
|
||||
|
||||
address = address.gsub(/(@.*)$/) { $1.downcase } # @Gmail.com -> @gmail.com
|
||||
|
||||
parse(address)
|
||||
end
|
||||
|
||||
# @return [Danbooru::EmailAddress] The email address, normalized to fix typos.
|
||||
def normalized_address
|
||||
Danbooru::EmailAddress.normalize(address)
|
||||
end
|
||||
|
||||
# @return [PublicSuffix::Domain] The domain part of the email address.
|
||||
def domain
|
||||
@domain ||= PublicSuffix.parse(parsed_address.domain)
|
||||
rescue PublicSuffix::DomainNotAllowed
|
||||
nil
|
||||
end
|
||||
|
||||
def as_json
|
||||
to_s
|
||||
end
|
||||
|
||||
def inspect
|
||||
"#<Danbooru::EmailAddress #{to_s}>"
|
||||
end
|
||||
|
||||
def ==(other)
|
||||
self.class == other.class && to_s == other.to_s
|
||||
end
|
||||
|
||||
def hash
|
||||
to_s.hash
|
||||
end
|
||||
|
||||
alias_method :eql?, :==
|
||||
end
|
||||
end
|
||||
29
app/logical/email_address_type.rb
Normal file
29
app/logical/email_address_type.rb
Normal file
@@ -0,0 +1,29 @@
|
||||
# frozen_string_literal: true
|
||||
|
||||
# Define a custom email address type that allows models to declare attributes of type Danbooru::EmailAddress.
|
||||
#
|
||||
# @see app/logical/danbooru/email_address.rb
|
||||
# @see config/initializers/types.rb
|
||||
# @see https://www.bigbinary.com/blog/rails-5-attributes-api
|
||||
# @see https://api.rubyonrails.org/classes/ActiveModel/Type/Value.html
|
||||
class EmailAddressType < ActiveRecord::Type::Value
|
||||
# Cast a String (or nil) value from the database to a Danbooru::EmailAddress object.
|
||||
#
|
||||
# @param value [String] the email address from the database
|
||||
# @return [Danbooru::EmailAddress]
|
||||
def cast(value)
|
||||
return nil if value.blank?
|
||||
super(Danbooru::EmailAddress.new(value))
|
||||
rescue Danbooru::EmailAddress::Error
|
||||
nil
|
||||
end
|
||||
|
||||
# Serialize a Danbooru::EmailAddress to a String for the database.
|
||||
#
|
||||
# @param value [Danbooru::EmailAddress] the email address object
|
||||
# @return [String]
|
||||
def serialize(value)
|
||||
return value.to_s if value.is_a?(Danbooru::EmailAddress)
|
||||
super value
|
||||
end
|
||||
end
|
||||
@@ -3,8 +3,11 @@
|
||||
class EmailAddress < ApplicationRecord
|
||||
belongs_to :user, inverse_of: :email_address
|
||||
|
||||
validates :address, presence: true, confirmation: true, format: { with: EmailValidator::EMAIL_REGEX }
|
||||
validates :normalized_address, uniqueness: true
|
||||
attribute :address
|
||||
attribute :normalized_address
|
||||
|
||||
validates :address, presence: true, format: { message: "is invalid", with: EmailValidator::EMAIL_REGEX }
|
||||
validates :normalized_address, presence: true, uniqueness: true
|
||||
validates :user_id, uniqueness: true
|
||||
validate :validate_deliverable, on: :deliverable
|
||||
|
||||
@@ -17,6 +20,7 @@ class EmailAddress < ApplicationRecord
|
||||
end
|
||||
|
||||
def address=(value)
|
||||
value = Danbooru::EmailAddress.normalize(value)&.to_s || value
|
||||
self.normalized_address = EmailValidator.normalize(value) || address
|
||||
super
|
||||
end
|
||||
|
||||
@@ -1,3 +1,4 @@
|
||||
Rails.application.reloader.to_prepare do
|
||||
ActiveRecord::Type.register(:ip_address, IpAddressType)
|
||||
ActiveRecord::Type.register(:email_address, EmailAddressType)
|
||||
end
|
||||
|
||||
@@ -69,6 +69,7 @@ en:
|
||||
uploader: "You"
|
||||
uploader_id: "You"
|
||||
user/email_address:
|
||||
address: "Email address"
|
||||
normalized_address: "Email address"
|
||||
user_feedback:
|
||||
creator: "You"
|
||||
|
||||
@@ -12,29 +12,32 @@ class EmailAddressTest < ActiveSupport::TestCase
|
||||
should allow_value("foo+bar@gmail.com").for(:address)
|
||||
should allow_value("foo@foo.bar.com").for(:address)
|
||||
|
||||
should_not allow_value("foo@gmail.com ").for(:address)
|
||||
should_not allow_value(" foo@gmail.com").for(:address)
|
||||
should_not allow_value("foo@-gmail.com").for(:address)
|
||||
should_not allow_value("foo@.gmail.com").for(:address)
|
||||
should_not allow_value("foo@gmail").for(:address)
|
||||
should_not allow_value("foo@gmail.").for(:address)
|
||||
should_not allow_value("foo@gmail,com").for(:address)
|
||||
should_not allow_value("foo@gmail.com.").for(:address)
|
||||
should_not allow_value("foo@gmail.co,").for(:address)
|
||||
should_not allow_value("foo@example").for(:address)
|
||||
should_not allow_value("fooqq@.com").for(:address)
|
||||
should_not allow_value("foo@gmail..com").for(:address)
|
||||
should_not allow_value("foo@gmailcom").for(:address)
|
||||
should_not allow_value("mailto:foo@gmail.com").for(:address)
|
||||
should_not allow_value('foo"bar"@gmail.com').for(:address)
|
||||
should_not allow_value('foo<bar>@gmail.com').for(:address)
|
||||
should_not allow_value("foo@gmail.com@gmail.com").for(:address)
|
||||
should_not allow_value("foo@g,ail.com").for(:address)
|
||||
should_not allow_value("foo@gmai;.com").for(:address)
|
||||
should_not allow_value("foo@gmail@com").for(:address)
|
||||
should_not allow_value("foo@gmail.c").for(:address)
|
||||
should_not allow_value("foo@foo.-bar.com").for(:address)
|
||||
should_not allow_value("foo@127.0.0.1").for(:address)
|
||||
should_not allow_value("foo@localhost").for(:address)
|
||||
end
|
||||
|
||||
should "fix typos" do
|
||||
assert_equal("foo@gmail.com", EmailAddress.new(address: "foo@gmail.com ").address.to_s)
|
||||
assert_equal("foo@gmail.com", EmailAddress.new(address: " foo@gmail.com").address.to_s)
|
||||
assert_equal("foo@gmail.com", EmailAddress.new(address: "foo@gmail.com\n").address.to_s)
|
||||
assert_equal("foo@gmail.com", EmailAddress.new(address: "foo@-gmail.com").address.to_s)
|
||||
assert_equal("foo@gmail.com", EmailAddress.new(address: "foo@.gmail.com").address.to_s)
|
||||
assert_equal("foo@gmail.com", EmailAddress.new(address: "foo@gmail").address.to_s)
|
||||
assert_equal("foo@gmail.com", EmailAddress.new(address: "foo@gmail,com").address.to_s)
|
||||
assert_equal("foo@gmail.com", EmailAddress.new(address: "foo@gmail.com.").address.to_s)
|
||||
assert_equal("foo@gmail.com", EmailAddress.new(address: "foo@gmail.co,").address.to_s)
|
||||
assert_equal("foo@gmail.com", EmailAddress.new(address: "foo@gmail..com").address.to_s)
|
||||
assert_equal("foo@gmail.com", EmailAddress.new(address: "foo@gmailcom").address.to_s)
|
||||
assert_equal("foo@gmail.com", EmailAddress.new(address: "mailto:foo@gmail.com").address.to_s)
|
||||
assert_equal("foo@gmail.com", EmailAddress.new(address: "foo@gmail.com@gmail.com").address.to_s)
|
||||
assert_equal("foo@gmail.com", EmailAddress.new(address: "foo@g,ail.com").address.to_s)
|
||||
assert_equal("foo@gmail.com", EmailAddress.new(address: "foo@gmai;.com").address.to_s)
|
||||
assert_equal("foo@gmail.com", EmailAddress.new(address: "foo@gmail@com").address.to_s)
|
||||
end
|
||||
end
|
||||
end
|
||||
|
||||
Reference in New Issue
Block a user