mime-types for Ruby

Continuous IntegrationTest Coverage


code https://github.com/mime-types/ruby-mime-types/
bugs https://github.com/mime-types/ruby-mime-types/issues
rdoc http://rdoc.info/gems/mime-types/

The mime-types library provides a library and registry for information about MIME content type definitions. It can be used to determine defined filename extensions for MIME types, or to use filename extensions to look up the likely MIME type definitions.

Version 3.0 is a major release that requires Ruby 2.0 compatibility and removes deprecated functions. The columnar registry format introduced in 2.6 has been made the primary format; the registry data has been extracted from this library and put into mime-types-data. Additionally, mime-types is now licensed exclusively under the MIT licence and there is a code of conduct in effect. There are a number of other smaller changes described in the History file.

About MIME Media Types

MIME content types are used in MIME-compliant communications, as in e-mail or HTTP traffic, to indicate the type of content which is transmitted. The mime-types library provides the ability for detailed information about MIME entities (provided as an enumerable collection of MIME::Type objects) to be determined and used. There are many types defined by RFCs and vendors, so the list is long but by definition incomplete; don’t hesitate to add additional type definitions. MIME type definitions found in mime-types are from RFCs, W3C recommendations, the IANA Media Types registry, and user contributions. It conforms to RFCs 2045 and 2231.

mime-types 1.x End of Life

mime-types 1.x is no longer supported as of 2015-10-27.

mime-types 2.x End of Life

mime-types 2.x is supported as 2.99.x receiving quarterly updates of the IANA registry and security updates for two years. It will reach full end of life on 2017-11-21.

mime-types 3.x

Users are encouraged to upgrade to mime-types 3.x as soon as is practical. mime-types 3.x requires Ruby 2.0 compatibility and a simpler licensing scheme.


MIME types are used in MIME entities, as in email or HTTP traffic. It is useful at times to have information available about MIME types (or, inversely, about files). A MIME::Type stores the known information about one MIME type.

require 'mime/types'

plaintext = MIME::Types['text/plain'] # => [ text/plain ]
text = plaintext.first
puts text.media_type            # => 'text'
puts text.sub_type              # => 'plain'

puts text.extensions.join(' ')  # => 'txt asc c cc h hh cpp hpp dat hlp'
puts text.preferred_extension   # => 'txt'
puts text.friendly              # => 'Text Document'
puts text.i18n_key              # => 'text.plain'

puts text.encoding              # => quoted-printable
puts text.default_encoding      # => quoted-printable
puts text.binary?               # => false
puts text.ascii?                # => true
puts text.obsolete?             # => false
puts text.registered?           # => true
puts text.complete?             # => true

puts text                       # => 'text/plain'

puts text == 'text/plain'       # => true
puts 'text/plain' == text       # => true
puts text == 'text/x-plain'     # => false
puts 'text/x-plain' == text     # => false

puts MIME::Type.simplified('x-appl/x-zip') # => 'x-appl/x-zip'
puts MIME::Type.i18n_key('x-appl/x-zip') # => 'x-appl.x-zip'

puts text.like?('text/x-plain') # => true
puts text.like?(MIME::Type.new('x-text/x-plain')) # => true

puts text.xrefs.inspect # => { "rfc" => [ "rfc2046", "rfc3676", "rfc5147" ] }
puts text.xref_urls # => [ "http://www.iana.org/go/rfc2046",
                    #      "http://www.iana.org/go/rfc3676",
                    #      "http://www.iana.org/go/rfc5147" ]

xtext = MIME::Type.new('x-text/x-plain')
puts xtext.media_type # => 'text'
puts xtext.raw_media_type # => 'x-text'
puts xtext.sub_type # => 'plain'
puts xtext.raw_sub_type # => 'x-plain'
puts xtext.complete? # => false

puts MIME::Types.any? { |type| type.content_type == 'text/plain' } # => true
puts MIME::Types.all?(&:registered?) # => false

# Various string representations of MIME types
qcelp = MIME::Types['audio/QCELP'].first # => audio/QCELP
puts qcelp.content_type         # => 'audio/QCELP'
puts qcelp.simplified           # => 'audio/qcelp'

xwingz = MIME::Types['application/x-Wingz'].first # => application/x-Wingz
puts xwingz.content_type        # => 'application/x-Wingz'
puts xwingz.simplified          # => 'application/x-wingz'

Columnar Store

mime-types uses as its primary registry storage format a columnar storage format reducing the default memory footprint. This is done by selectively loading the data on a per-attribute basis. When the registry is first loaded from the columnar store, only the canonical MIME content type and known extensions and the MIME type will be connected to its loading registry. When other data about the type is required (including preferred_extension, obsolete?, and registered?) that data is loaded from its own column file for all types in the registry.

The load of any column data is performed with a Mutex to ensure that types are updated safely in a multithreaded environment. Benchmarks show that while columnar data loading is slower than the JSON store, it cuts the memory use by a third over the JSON store.

If you prefer to load all the data at once, this can be specified in your application Gemfile as:

gem 'mime-types', require: 'mime/types/full'

Projects that do not use Bundler should require the same:

require 'mime/types/full'

Libraries that use mime-types are discouraged from choosing the JSON store.

For applications and clients that used mime-types 2.6 when the columnar store was introduced, the require used previously will still work through at least version 4 and possibly beyond; it is effectively an empty operation. You are recommended to change your Gemfile as soon as is practical.

require 'mime/types/columnar'

Note that MIME::Type::Columnar and MIME::Types::Columnar are considered private variant implementations of MIME::Type and MIME::Types and the specific implementation should not be relied upon by consumers of the mime-types library. Instead, depend on the public implementations (MIME::Type and MIME::Types) only.

Cached Storage

mime-types supports a cache of MIME types using Marshal.dump. The cache is invalidated for each version of the mime-types-data gem so that data version 3.2015.1201 will not be reused with data version 3.2016.0101. If the environment variable RUBY_MIME_TYPES_CACHE is set to a cache file, mime-types will attempt to load the MIME type registry from the cache file. If it cannot, it will load the types normally and then saves the registry to the cache file.

The caching works with both full stores and columnar stores. Only the data that has been loaded prior to saving the cache will be stored.

mime-types Modified Semantic Versioning

The mime-types library has one version number, but this single version number tracks both API changes and registry data changes; this is not wholly compatible with all aspects of Semantic Versioning; removing a MIME type from the registry could be considered a breaking change under some interpretations of semantic versioning (as lookups for that particular type would no longer work by default).

mime-types uses a modified semantic versioning scheme. Given the version MAJOR.MINOR:

  1. If an incompatible API (code) change is made, the MAJOR version will be incremented, MINOR will be set to zero, and PATCH will be reset to the implied zero.

  2. If an API (code) feature is added that does not break compatibilty, the MINOR version will be incremented and PATCH will be reset to the implied zero.

  3. If there is a bugfix to a feature added in the most recent MAJOR.MINOR release, the implied PATCH value will be incremented resulting in MAJOR.MINOR.PATCH.

In practical terms, there will be fewer releases of mime-types focussing on features because of the existence of the mime-types-data gem, and if features are marked deprecated in the course of mime-types 3.x, they will not be removed until mime-types 4.x or possibly later.

Contributor Code of Conduct

As contributors and maintainers of this project, and in the interest of fostering an open and welcoming community, we pledge to respect all people who contribute through reporting issues, posting feature requests, updating documentation, submitting pull requests or patches, and other activities.

We are committed to making participation in this project a harassment-free experience for everyone, regardless of level of experience, gender, gender identity and expression, sexual orientation, disability, personal appearance, body size, race, ethnicity, age, religion, or nationality.

Examples of unacceptable behavior by participants include:

  • The use of sexualized language or imagery
  • Personal attacks
  • Trolling or insulting/derogatory comments
  • Public or private harassment
  • Publishing other’s private information, such as physical or electronic addresses, without explicit permission
  • Other unethical or unprofessional conduct.

Project maintainers have the right and responsibility to remove, edit, or reject comments, commits, code, wiki edits, issues, and other contributions that are not aligned to this Code of Conduct. By adopting this Code of Conduct, project maintainers commit themselves to fairly and consistently applying these principles to every aspect of managing this project. Project maintainers who do not follow or enforce the Code of Conduct may be permanently removed from the project team.

This code of conduct applies both within project spaces and in public spaces when an individual is representing the project or its community.

Instances of abusive, harassing, or otherwise unacceptable behavior may be reported by opening an issue or contacting one or more of the project maintainers.

This Code of Conduct is adapted from the Contributor Covenant, version 1.2.0, available at contributor-covenant.org/version/1/2/0/.


I value any contribution to mime-types you can provide: a bug report, a feature request, or code contributions.

As mime-types is a mature codebase, there are a few guidelines:

  • Code changes will not be accepted without tests. The test suite is written with Minitest.
  • Match my coding style.
  • Use a thoughtfully-named topic branch that contains your change. Rebase your commits into logical chunks as necessary.
  • Use quality commit messages.
  • Do not change the version number; when your patch is accepted and a release is made, the version will be updated at that point.
  • Submit a GitHub pull request with your changes.
  • New or changed behaviours require new or updated documentation.

Adding or Modifying MIME Types

The mime-types registry is no longer contained in mime-types, but in {mime-types-data}[https://github.com/mime-types/mime-types-data]. Please see that project for contributions there.

Test Dependencies

mime-types uses Ryan Davis’s Hoe to manage the release process, and it adds a number of rake tasks. You will mostly be interested in:

$ rake

which runs the tests the same way that:

$ rake test
$ rake travis

will do.

To assist with the installation of the development dependencies for mime-types, I have provided the simplest possible Gemfile pointing to the (generated) mime-types.gemspec file. This will permit you to do:

$ bundle install

to get the development dependencies. If you aleady have hoe installed, you can accomplish the same thing with:

$ rake newb

This task will install any missing dependencies, run the tests/specs, and generate the RDoc.

You can run tests with code coverage analysis by running:

$ rake test:coverage


mime-types offers several benchmark tasks to measure different measures of performance.

There is a repeated load test, measuring how long it takes to start and load mime-types with its full registry. By default, it runs fifty loops and uses the built-in benchmark library.

$ rake benchmark:load

There are two allocation tracing benchmarks (for normal and columnar loads). These can only be run on Ruby 2.1 or better and requires the allocation_tracer gem (not installed by default).

$ rake benchmark:allocations
$ rake benchmark:allocations:columnar

There are two loaded object count benchmarks (for normal and columnar loads). These use ObjectSpace.count_objects.

$ rake benchmark:objects
$ rake benchmark:objects:columnar


Here’s the most direct way to get your work merged into the project:

  • Fork the project.
  • Clone down your fork (git clone git://github.com/<username>/ruby-mime-types.git).
  • Create a topic branch to contain your change (git checkout -b my_awesome_feature).
  • Hack away, add tests. Not necessarily in that order.
  • Make sure everything still passes by running rake.
  • If necessary, rebase your commits into logical chunks, without errors.
  • Push the branch up (git push origin my_awesome_feature).
  • Create a pull request against mime-types/ruby-mime-types and describe what your change does and the why you think it should be merged.


  • Austin Ziegler created mime-types.

Thanks to everyone else who has contributed to mime-types:

  • Aaron Patterson
  • Aggelos Avgerinos
  • Andre Pankratz
  • Andy Brody
  • Arnaud Meuret
  • Brandon Galbraith
  • Chris Gat
  • David Genord
  • Eric Marden
  • Garret Alfert
  • Godfrey Chan
  • Greg Brockman
  • Hans de Graaff
  • Henrik Hodne
  • Jeremy Evans
  • Juanito Fatas
  • Łukasz Śliwa
  • Keerthi Siva
  • Ken Ip
  • Martin d’Allens
  • Mauricio Linhares
  • nycvotes-dev
  • Postmodern
  • Richard Hirner
  • Richard Hurt
  • Richard Schneeman
  • Tibor Szolár
  • Todd Carrico


  • Copyright 2003–2015 Austin Ziegler and other contributors.

The software in this repository is made available under the MIT license.

MIT License

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.