Mail Build Status


Mail is an internet library for Ruby that is designed to handle email generation, parsing and sending in a simple, rubyesque manner.

The purpose of this library is to provide a single point of access to handle all email functions, including sending and receiving email. All network type actions are done through proxy methods to Net::SMTP, Net::POP3 etc.

Built from my experience with TMail, it is designed to be a pure ruby implementation that makes generating, sending and parsing email a no brainer.

It is also designed from the ground up to work with the more modern versions of Ruby. This is because Ruby > 1.9 handles text encodings much more wonderfully than Ruby 1.8.x and so these features have been taken full advantage of in this library allowing Mail to handle a lot more messages more cleanly than TMail. Mail does run on Ruby 1.8.x… it’s just not as fun to code.

Finally, Mail has been designed with a very simple object oriented system that really opens up the email messages you are parsing, if you know what you are doing, you can fiddle with every last bit of your email directly.

You can contribute to this library

Yes, you! Mail is used in countless apps by people around the world. It is, like all open source software, a labour of love borne from our free time. If you would like to say thanks, please dig in and contribute alongside us! Triage and fix GitHub issues, improve our documentation, add new features—up to you! Thank you for pitching in.



Mail supports Ruby 1.8.7+, including JRuby and Rubinius.

Every Mail commit is tested by Travis on all supported Ruby versions.


If you want to discuss mail with like minded individuals, please subscribe to the Google Group.

Current Capabilities of Mail

  • RFC5322 Support, Reading and Writing
  • RFC6532 Support, reading UTF-8 headers
  • RFC2045-2049 Support for multipart email
  • Support for creating multipart alternate email
  • Support for reading multipart/report email & getting details from such
  • Wrappers for File, Net/POP3, Net/SMTP
  • Auto-encoding of non-US-ASCII bodies and header fields

Mail is RFC5322 and RFC6532 compliant now, that is, it can parse US-ASCII and UTF-8 email and generate US-ASCII email. There are a few obsoleted email syntax that it will have problems with, but it also is quite robust, meaning, if it finds something it doesn’t understand it will not crash, instead, it will skip the problem and keep parsing. In the case of a header it doesn’t understand, it will initialise the header as an optional unstructured field and continue parsing.

This means Mail won’t (ever) crunch your data (I think).

You can also create MIME emails. There are helper methods for making a multipart/alternate email for text/plain and text/html (the most common pair) and you can manually create any other type of MIME email.


Next TODO:

  • Improve MIME support for character sets in headers, currently works, mostly, needs refinement.

Testing Policy

Basically… we do BDD on Mail. No method gets written in Mail without a corresponding or covering spec. We expect as a minimum 100% coverage measured by RCov. While this is not perfect by any measure, it is pretty good. Additionally, all functional tests from TMail are to be passing before the gem gets released.

It also means you can be sure Mail will behave correctly.

API Policy

No API removals within a single point release. All removals to be deprecated with warnings for at least one MINOR point release before removal.

Also, all private or protected methods to be declared as such - though this is still I/P.


Installation is fairly simple, I host mail on rubygems, so you can just do:

# gem install mail


If you didn’t know, handling encodings in Emails is not as straight forward as you would hope.

I have tried to simplify it some:

  1. All objects that can render into an email, have an #encoded method. Encoded will return the object as a complete string ready to send in the mail system, that is, it will include the header field and value and CRLF at the end and wrapped as needed.

  2. All objects that can render into an email, have a #decoded method. Decoded will return the object’s “value” only as a string. This means it will not include the header fields (like ‘To:’ or ‘Subject:’).

  3. By default, calling #to_s on a container object will call its encoded method, while #to_s on a field object will call its decoded method. So calling #to_s on a Mail object will return the mail, all encoded ready to send, while calling #to_s on the From field or the body will return the decoded value of the object. The header object of Mail is considered a container. If you are in doubt, call #encoded, or #decoded explicitly, this is safer if you are not sure.

  4. Structured fields that have parameter values that can be encoded (e.g. Content-Type) will provide decoded parameter values when you call the parameter names as methods against the object.

  5. Structured fields that have parameter values that can be encoded (e.g. Content-Type) will provide encoded parameter values when you call the parameter names through the object.parameters['']</code> method call.


Please do! Contributing is easy in Mail. Please read the document for more info.


All major mail functions should be able to happen from the Mail module. So, you should be able to just require 'mail' to get started.

mail is pretty well documented in its Ruby code. You can look it up e.g. at

Making an email

```ruby mail = do from ‘[email protected]’ to ‘[email protected]’ subject ‘This is a test email’ body‘body.txt’) end

mail.to_s #=> “From: [email protected]\r\nTo: [email protected]… ```

Making an email, have it your way:

```ruby mail = do body‘body.txt’) end

mail[‘from’] = ‘[email protected]’ mail[:to] = ‘[email protected]’ mail.subject = ‘This is a test email’

mail.header[‘X-Custom-Header’] = ‘custom value’

mail.to_s #=> “From: [email protected]\r\nTo: [email protected]… ```

Don’t Worry About Message IDs:

```ruby mail = do to ‘[email protected]’ body ‘Some simple body’ end

mail.to_s =~ /Message-ID: <[\d\w_][email protected]+.mail/ #=> 27 ```

Mail will automatically add a Message-ID field if it is missing and give it a unique, random Message-ID along the lines of:

<[email protected]>

Or do worry about Message-IDs:

```ruby mail = do to ‘[email protected]’ message_id ‘[email protected]’ body ‘Some simple body’ end

mail.to_s =~ /Message-ID: [email protected]/ #=> 27 ```

Mail will take the message_id you assign to it trusting that you know what you are doing.

Sending an email:

Mail defaults to sending via SMTP to local host port 25. If you have a sendmail or postfix daemon running on this port, sending email is as easy as:

ruby Mail.deliver do from '[email protected]' to '[email protected]' subject 'Here is the image you wanted' body'body.txt') add_file '/full/path/to/somefile.png' end


```ruby mail = do from ‘[email protected]’ to ‘[email protected]’ subject ‘Here is the image you wanted’ body‘body.txt’) add_file :filename => ‘somefile.png’, :content =>‘/somefile.png’) end

mail.deliver! ```

Sending via sendmail can be done like so:

```ruby mail = do from ‘[email protected]’ to ‘[email protected]’ subject ‘Here is the image you wanted’ body‘body.txt’) add_file :filename => ‘somefile.png’, :content =>‘/somefile.png’) end

mail.delivery_method :sendmail

mail.deliver ```

Sending via smtp (for example to mailcatcher) ```ruby

Mail.defaults do delivery_method :smtp, address: “localhost”, port: 1025 end ```

Exim requires its own delivery manager, and can be used like so:

```ruby mail.delivery_method :exim, :location => “/usr/bin/exim”

mail.deliver ```

Mail may be “delivered” to a logfile, too, for development and testing:

```ruby # Delivers by logging the encoded message to $stdout mail.delivery_method :logger

Delivers to an existing logger at :debug severity

mail.delivery_method :logger, logger: other_logger, severity: :debug ```

Getting Emails from a POP or IMAP Server:

You can configure Mail to receive email using retriever_method within Mail.defaults:

```ruby # e.g. POP3 Mail.defaults do retriever_method :pop3, :address => “”, :port => 995, :user_name => ‘', :password => '', :enable_ssl => true end


Mail.defaults do retriever_method :imap, :address => “”, :port => 993, :user_name => ‘', :password => '', :enable_ssl => true end ```

You can access incoming email in a number of ways.

The most recent email:

ruby Mail.all #=> Returns an array of all emails Mail.first #=> Returns the first unread email Mail.last #=> Returns the last unread email

The first 10 emails sorted by date in ascending order:

ruby emails = Mail.find(:what => :first, :count => 10, :order => :asc) emails.length #=> 10

Or even all emails:

ruby emails = Mail.all emails.length #=> LOTS!

Reading an Email

```ruby mail =‘/path/to/message.eml’)

mail.envelope_from #=> ‘[email protected]’ mail.from.addresses #=> [‘[email protected]’, ‘[email protected]’] mail.sender.address #=> ‘[email protected]’ #=> ‘[email protected]’ #=> ‘[email protected]’ mail.subject #=> “This is the subject” #=> ‘21 Nov 1997 09:55:06 -0600’ mail.message_id #=> ‘[email protected]’ mail.decoded #=> ‘This is the body of the email… ```

Many more methods available.

Reading a Multipart Email

```ruby mail =‘multipart_email’)

mail.multipart? #=> true #=> 2 mail.body.preamble #=> “Text before the first part” mail.body.epilogue #=> “Text after the last part” { |p| p.content_type } #=> [‘text/plain’, ‘application/pdf’] { |p| p.class } #=> [Mail::Message, Mail::Message][0].content_type_parameters #=> => ‘ISO-8859-1’[1].content_type_parameters #=> => ‘my.pdf’ ```

Mail generates a tree of parts. Each message has many or no parts. Each part is another message which can have many or no parts.

A message will only have parts if it is a multipart/mixed or multipart/related content type and has a boundary defined.

Testing and Extracting Attachments

ruby mail.attachments.each do | attachment | # Attachments is an AttachmentsList object containing a # number of Part objects if (attachment.content_type.start_with?('image/')) # extracting images for example... filename = attachment.filename begin + filename, "w+b", 0644) {|f| f.write attachment.decoded} rescue => e puts "Unable to save data for #{filename} because #{e.message}" end end end ### Writing and Sending a Multipart/Alternative (HTML and Text) Email

Mail makes some basic assumptions and makes doing the common thing as simple as possible…. (asking a lot from a mail library)

```ruby mail = Mail.deliver do to ‘[email protected]’ from ‘Mikel Lindsaar [email protected]’ subject ‘First multipart email sent with Mail’

text_part do body ‘This is plain text’ end

html_part do content_type ‘text/html; charset=UTF-8’ body ‘<h1>This is HTML</h1>’ end end ```

Mail then delivers the email at the end of the block and returns the resulting Mail::Message object, which you can then inspect if you so desire…

``` puts mail.to_s #=>

To: [email protected] From: Mikel Lindsaar [email protected] Subject: First multipart email sent with Mail Content-Type: multipart/alternative; boundary=–==_mimepart_4a914f0c911be_6f0f1ab8026659 Message-ID: [email protected] Date: Mon, 24 Aug 2009 00:15:46 +1000 Mime-Version: 1.0 Content-Transfer-Encoding: 7bit

—-==_mimepart_4a914f0c911be_6f0f1ab8026659 Content-ID: [email protected] Date: Mon, 24 Aug 2009 00:15:46 +1000 Mime-Version: 1.0 Content-Type: text/plain Content-Transfer-Encoding: 7bit

This is plain text —-==_mimepart_4a914f0c911be_6f0f1ab8026659 Content-Type: text/html; charset=UTF-8 Content-ID: [email protected] Date: Mon, 24 Aug 2009 00:15:46 +1000 Mime-Version: 1.0 Content-Transfer-Encoding: 7bit

This is HTML

—-==_mimepart_4a914f0c911be_6f0f1ab8026659– ```

Mail inserts the content transfer encoding, the mime version, the content-IDs and handles the content-type and boundary.

Mail assumes that if your text in the body is only us-ascii, that your transfer encoding is 7bit and it is text/plain. You can override this by explicitly declaring it.

Making Multipart/Alternate, Without a Block

You don’t have to use a block with the text and html part included, you can just do it declaratively. However, you need to add Mail::Parts to an email, not Mail::Messages.

```ruby mail = do to ‘[email protected]’ from ‘Mikel Lindsaar [email protected]’ subject ‘First multipart email sent with Mail’ end

text_part = do body ‘This is plain text’ end

html_part = do content_type ‘text/html; charset=UTF-8’ body ‘<h1>This is HTML</h1>’ end

mail.text_part = text_part mail.html_part = html_part ```

Results in the same email as done using the block form

Getting Error Reports from an Email:

```ruby @mail =‘/path/to/bounce_message.eml’)

@mail.bounced? #=> true @mail.final_recipient #=> rfc822;[email protected] @mail.action #=> failed @mail.error_status #=> 5.5.0 @mail.diagnostic_code #=> smtp;550 Requested action not taken: mailbox unavailable @mail.retryable? #=> false ```

Attaching and Detaching Files

You can just read the file off an absolute path, Mail will try to guess the mime_type and will encode the file in Base64 for you.

ruby @mail = @mail.add_file("/path/to/file.jpg") #=> true #=> 'base64' @mail.attachments.first.mime_type #=> 'image/jpg' @mail.attachments.first.filename #=> 'file.jpg' @mail.attachments.first.decoded =='/path/to/file.jpg') #=> true

Or You can pass in file_data and give it a filename, again, mail will try and guess the mime_type for you.

ruby @mail = @mail.attachments['myfile.pdf'] ='path/to/myfile.pdf') #=> true @mail.attachments.first.mime_type #=> 'application/pdf' @mail.attachments.first.decoded =='path/to/myfile.pdf') #=> true

You can also override the guessed MIME media type if you really know better than mail (this should be rarely needed)

ruby @mail = @mail.attachments['myfile.pdf'] = { :mime_type => 'application/x-pdf', :content =>'path/to/myfile.pdf') } #=> 'application/x-pdf'

Of course… Mail will round trip an attachment as well

```ruby @mail = do to ‘[email protected]’ from ‘Mikel Lindsaar [email protected]’ subject ‘First multipart email sent with Mail’

text_part do body ‘Here is the attachment you wanted’ end

html_part do content_type ‘text/html; charset=UTF-8’ body ‘<h1>Funky Title</h1><p>Here is the attachment you wanted</p>’ end

add_file ‘/path/to/myfile.pdf’ end

@round_tripped_mail =

@round_tripped_mail.attachments.length #=> 1 @round_tripped_mail.attachments.first.filename #=> ‘myfile.pdf’ ``` See “Testing and extracting attachments” above for more details.

Using Mail with Testing or Spec’ing Libraries

If mail is part of your system, you’ll need a way to test it without actually sending emails, the TestMailer can do this for you.

ruby require 'mail' => true Mail.defaults do delivery_method :test end => #<Mail::Configuration:0x19345a8 @delivery_method=Mail::TestMailer> Mail::TestMailer.deliveries => [] Mail.deliver do to '[email protected]' from '[email protected]' subject 'testing' body 'hello' end => #<Mail::Message:0x19284ec ... Mail::TestMailer.deliveries.length => 1 Mail::TestMailer.deliveries.first => #<Mail::Message:0x19284ec ... Mail::TestMailer.deliveries.clear => []

There is also a set of RSpec matchers stolen/inspired by Shoulda’s ActionMailer matchers (you’ll want to set delivery_method as above too):

```ruby Mail.defaults do delivery_method :test # in practice you’d do this in spec_helper.rb end

describe “sending an email” do include Mail::Matchers

before(:each) do Mail::TestMailer.deliveries.clear

Mail.deliver do
  to ['[email protected]', '[email protected]']
  from '[email protected]'
  subject 'testing'
  body 'hello'
end   end

it { have_sent_email } # passes if any email at all was sent

it { have_sent_email.from(‘[email protected]’) } it {‘[email protected]’) }

# can specify a list of recipients… it {[‘[email protected]’, ‘[email protected]’]) }

# …or chain recipients together it {‘[email protected]’).to(‘[email protected]’) }

it { have_sent_email.with_subject(‘testing’) }

it { have_sent_email.with_body(‘hello’) }

# Can match subject or body with a regex # (or anything that responds_to? :match)

it { have_sent_email.matching_subject(/test(ing)?/) } it { have_sent_email.matching_body(/h(a|e)llo/) }

# Can chain together modifiers # Note that apart from recipients, repeating a modifier overwrites old value.

it { have_sent_email.from(‘[email protected]’).to(‘[email protected]’).matching_body(/hell/)

# test for attachments

# … by specific attachment it { have_sent_email.with_attachments(my_attachment) }

# … or any attachment it { have_sent_email.with_attachments(any_attachment) }

# … or attachment with filename it { have_sent_email.with_attachments(an_attachment_with_filename(‘file.txt’)) }

# … or attachment with mime_type it { have_sent_email.with_attachments(an_attachment_with_mime_type(‘application/pdf’)) }

# … by array of attachments it { have_sent_email.with_attachments([my_attachment1, my_attachment2]) } #note that order is important

#… by presence it { have_sent_email.with_any_attachments }

#… or by absence it { have_sent_email.with_no_attachments }

end ```

Excerpts from TREC Spam Corpus 2005

The spec fixture files in spec/fixtures/emails/from_trec_2005 are from the 2005 TREC Public Spam Corpus. They remain copyrighted under the terms of that project and license agreement. They are used in this project to verify and describe the development of this email parser implementation.

They are used as allowed by ‘Permitted Uses, Clause 3’:

"Small excerpts of the information may be displayed to others
 or published in a scientific or technical context, solely for
 the purpose of describing the research and development and
 related issues."



(The MIT License)

Copyright (c) 2009-2016 Mikel Lindsaar

Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the ‘Software’), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.