fileinfo 0.2.0 → 0.3.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 3003ef10828a21e938e6329915aae768e7231565
-  data.tar.gz: 5d32b96bce981dc1cbcbd0efb437760ba8d318e8
+  metadata.gz: 1fd22a80c28289631c040bde751eee82a7b51369
+  data.tar.gz: 38320adc6dc77ad1e6e38e6d4d1b4c49052da480
 SHA512:
-  metadata.gz: 0ef4dff4dfe51878aa0a0d71ecfe6c1b95498f7e65048b38dd735a8dd523d371c91fc912943d698bafcc19d35b2616db94443738fe2a456468a9fdcf1f393708
-  data.tar.gz: ef3b64ace53578d25f5d64d6a9cb08a1e59eb7f169497e61b2ec0f0c112e714535ba5b96300fea40f1fbea50c240cb60ee01d07f0a86470ca21a27d34451cbe3
+  metadata.gz: 24f820540d61e3e613b3c511eb3d5e8cebc7663de1c8828d6b4456a7215835ec7c18e1574411315e8961c0a7b1de2c63f7eddf1654b675a45139a0cf0ac8b59c
+  data.tar.gz: 2823c8c4c042db5d8b8aecbfffd38b3c552c7c2b20f1300bbddd4aacf5b08ff8139127339fd20d19043d795e1152dc6c3a88f32195aa86ec8f9d2eaf2f466e54

data/README.md

@@ -6,7 +6,7 @@
 [![Coverage Status](https://coveralls.io/repos/rafBM/fileinfo/badge.png?branch=master)](https://coveralls.io/r/rafBM/fileinfo?branch=master)
 [![Dependency Status](https://gemnasium.com/rafBM/fileinfo.png)](https://gemnasium.com/rafBM/fileinfo)
 
-FileInfo detects encoding from strings and files using the [wonderful Unix `file` command](http://en.wikipedia.org/wiki/File_\(command\)).
+FileInfo detects file encodings and MIME types using the [wonderful Unix `file` command](http://en.wikipedia.org/wiki/File_\(command\)).
 
 ## Installation
 
@@ -30,18 +30,43 @@ $ gem install fileinfo
 
 ## Usage
 
-Use `FileInfo.parse` with a string:
+### Detect encoding
+
+Use `FileInfo.parse` with a string or `FileInfo.load` with a filename.
 
 ```ruby
 FileInfo.parse('foo bar baz').encoding # => #<Encoding:US-ASCII>
 FileInfo.parse('föø bår bàz').encoding # => #<Encoding:UTF-8>
+
+filename = '/Users/rafbm/Downloads/some_crap_coming_from_windows.csv'
+FileInfo.load(filename).encoding # => #<Encoding:ISO-8859-1>
 ```
 
-Use `FileInfo.load` with a filename:
+### Detect MIME type
+
+A bunch of methods are available depending on your needs.
 
 ```ruby
-filename = '/Users/rafbm/Downloads/some_crap_coming_from_windows.csv'
-FileInfo.load(filename).encoding # => #<Encoding:ISO-8859-1>
+info = FileInfo.load('picture.jpg')
+info.type       # => "image/jpeg"
+info.media_type # => "image"
+info.sub_type   # => "jpeg"
+```
+
+The `#mime_type` method can also return a `MIME::Type` instance.
+
+```ruby
+info = FileInfo.parse('Hello world')
+info.mime_type # => #<MIME::Type:0x007f81a1a36090 @content_type="text/plain" ...>
+```
+
+**NOTE:** You must install the `mime-types` gem yourself as it is not a dependency of `fileinfo`.
+
+Finally, `#content_type` can be used to get the full string returned by `file --mime --brief`.
+
+```ruby
+info = FileInfo.parse('Hëllø wõrld')
+info.content_type # => "text/plain; charset=utf-8"
 ```
 
 ## Contributing
@@ -54,4 +79,4 @@ FileInfo.load(filename).encoding # => #<Encoding:ISO-8859-1>
 
 ---
 
-© 2013 [Rafaël Blais Masson](http://heliom.ca). FileInfo is released under the MIT license.
+© 2014 [Rafaël Blais Masson](http://rafbm.com). FileInfo is released under the MIT license.

data/fileinfo.gemspec

@@ -8,8 +8,8 @@ Gem::Specification.new do |spec|
   spec.version       = FileInfo::VERSION
   spec.authors       = ['Rafaël Blais Masson']
   spec.email         = ['rafael@heliom.ca']
-  spec.description   = "FileInfo detects encoding from strings and files using the wonderful Unix `file` command."
-  spec.summary       = "FileInfo detects encoding from strings and files using the wonderful Unix `file` command."
+  spec.description   = "FileInfo detects file encodings and MIME types using the wonderful Unix `file` command."
+  spec.summary       = "FileInfo detects file encodings and MIME types using the wonderful Unix `file` command."
   spec.homepage      = 'http://github.com/rafBM/fileinfo'
   spec.license       = 'MIT'
 
@@ -21,4 +21,5 @@ Gem::Specification.new do |spec|
   spec.add_development_dependency 'bundler', '~> 1.3'
   spec.add_development_dependency 'rake'
   spec.add_development_dependency 'rspec', '~> 2.14'
+  spec.add_development_dependency 'mime-types'
 end

data/lib/file_info.rb

@@ -6,42 +6,60 @@ require 'tempfile'
 class FileInfo
   class UnknownEncodingError < StandardError; end
 
-  STRING_REGEX    = /: ([^:]+)$/
-  ENCODING_REGEX  = /charset=(\S+)/
+  MIME_TYPE_REGEX = /^[^;]+/
+  CHARSET_REGEX = /charset=(\S+)/
+
+  MIME_TYPE_ERROR_MESSAGE = 'You must install the "mime-types" gem to use FileInfo#mime_type'
+
+  attr_reader :content_type
 
   def initialize(output)
-    @output = output
+    @content_type = output.strip
+  end
+
+  def type
+    @type ||= content_type.match(MIME_TYPE_REGEX)[0]
+  end
+
+  def media_type
+    @media_type ||= type.split('/')[0]
+  end
+
+  def sub_type
+    @sub_type ||= type.split('/')[1]
+  end
+
+  def mime_type
+    @mime_type ||= begin
+      require 'mime/types' unless defined? MIME::Types
+      MIME::Types[type][0]
+    rescue LoadError
+      raise LoadError, MIME_TYPE_ERROR_MESSAGE
+    end
+  end
+
+  def charset
+    @charset ||= content_type.match(CHARSET_REGEX)[1]
   end
 
   def encoding
-    @encoding ||= ::Encoding.find(encoding_string)
+    @encoding ||= ::Encoding.find(charset)
   rescue ArgumentError => e
     raise UnknownEncodingError, e.message
   end
 
   def self.load(filename)
     raise ArgumentError, "File '#{filename}' does not exist." if !File.exists? filename
-    new `file --mime #{Shellwords.escape(filename)}`
+    new `file --mime --brief #{Shellwords.escape(filename)}`
   end
 
   def self.parse(content)
     file = Tempfile.new(rand.to_s)
     file.write(content)
     file.rewind
-    output = `file --mime #{file.path}`
+    new `file --mime --brief #{file.path}`
+  ensure
     file.close
     file.unlink
-
-    new output
-  end
-
-private
-
-  def string
-    @string ||= @output.match(STRING_REGEX)[1].strip
-  end
-
-  def encoding_string
-    @encoding_string ||= string.match(ENCODING_REGEX)[1]
   end
 end

data/lib/file_info/version.rb

@@ -1,3 +1,3 @@
 class FileInfo
-  VERSION = '0.2.0'
+  VERSION = '0.3.0'
 end

data/spec/file_info_spec.rb

@@ -7,15 +7,33 @@ describe FileInfo do
   let(:macroman_file)   { fixture('encoding_macroman.csv') }
   let(:utf8_file)       { fixture('encoding_utf8.csv') }
 
-  describe '.load' do
-    it 'detects encoding from file' do
+  let(:photoshop_file)  { fixture('mockup.psd') }
+  let(:binary_file)     { fixture('bytes') }
+
+  describe '#charset' do
+    it 'returns encoding string' do
+      expect(FileInfo.load(ascii_file.path).charset).to      eq 'us-ascii'
+      expect(FileInfo.load(isolatin_file.path).charset).to   eq 'iso-8859-1'
+      expect(FileInfo.load(isowindows_file.path).charset).to eq 'iso-8859-1'
+      expect(FileInfo.load(utf8_file.path).charset).to       eq 'utf-8'
+    end
+  end
+
+  describe '#encoding' do
+    it 'returns Encoding instance' do
       expect(FileInfo.load(ascii_file.path).encoding).to      eq Encoding::US_ASCII
       expect(FileInfo.load(isolatin_file.path).encoding).to   eq Encoding::ISO_8859_1
       expect(FileInfo.load(isowindows_file.path).encoding).to eq Encoding::ISO_8859_1
       expect(FileInfo.load(utf8_file.path).encoding).to       eq Encoding::UTF_8
     end
 
-    it 'detects encoding from file with space in filename' do
+    it 'raises UnknownEncodingError' do
+      expect { FileInfo.load(macroman_file.path).encoding }.to raise_error(FileInfo::UnknownEncodingError)
+    end
+  end
+
+  describe '.load' do
+    it 'accepts filename with space' do
       old_filename = fixture('encoding_utf8.csv').path
       new_filename = old_filename.sub('_', ' ')
       FileUtils.cp(old_filename, new_filename)
@@ -25,7 +43,7 @@ describe FileInfo do
       FileUtils.rm(new_filename)
     end
 
-    it 'detects encoding from file with space and quote in filename' do
+    it 'accepts filename with space and quote' do
       old_filename = fixture('encoding_utf8.csv').path
       new_filename = old_filename.sub('_', ' " ')
       FileUtils.cp(old_filename, new_filename)
@@ -41,7 +59,7 @@ describe FileInfo do
   end
 
   describe '.parse' do
-    it 'detects encoding from string' do
+    it 'accepts a string' do
       expect(FileInfo.parse(ascii_file.read).encoding).to      eq Encoding::US_ASCII
       expect(FileInfo.parse(isolatin_file.read).encoding).to   eq Encoding::ISO_8859_1
       expect(FileInfo.parse(isowindows_file.read).encoding).to eq Encoding::ISO_8859_1
@@ -49,9 +67,72 @@ describe FileInfo do
     end
   end
 
-  describe '#encoding' do
-    it 'raises UnknownEncodingError' do
-      expect { FileInfo.load(macroman_file.path).encoding }.to raise_error(FileInfo::UnknownEncodingError)
+  let(:txt) { FileInfo.parse(FileInfo.parse('Hello, world!')) }
+  let(:csv) { FileInfo.load(utf8_file.path) }
+  let(:psd) { FileInfo.load(photoshop_file.path) }
+  let(:empty) { FileInfo.parse('') }
+  let(:bytes) { FileInfo.load(binary_file.path) }
+
+  describe '#content_type' do
+    it 'returns full Content-Type string' do
+      expect(txt.content_type).to eq 'text/plain; charset=us-ascii'
+      expect(csv.content_type).to eq 'text/plain; charset=utf-8'
+      expect(psd.content_type).to eq 'image/vnd.adobe.photoshop; charset=binary'
+      expect(empty.content_type).to eq 'application/x-empty; charset=binary'
+      expect(bytes.content_type).to eq 'application/octet-stream; charset=binary'
+    end
+  end
+
+  describe '#type' do
+    it 'returns MIME type string' do
+      expect(txt.type).to eq 'text/plain'
+      expect(csv.type).to eq 'text/plain'
+      expect(psd.type).to eq 'image/vnd.adobe.photoshop'
+      expect(empty.type).to eq 'application/x-empty'
+      expect(bytes.type).to eq 'application/octet-stream'
+    end
+  end
+
+  describe '#media_type' do
+    it 'returns MIME media type string' do
+      expect(txt.media_type).to eq 'text'
+      expect(csv.media_type).to eq 'text'
+      expect(psd.media_type).to eq 'image'
+      expect(empty.media_type).to eq 'application'
+      expect(bytes.media_type).to eq 'application'
+    end
+  end
+
+  describe '#sub_type' do
+    it 'returns MIME sub type string' do
+      expect(txt.sub_type).to eq 'plain'
+      expect(csv.sub_type).to eq 'plain'
+      expect(psd.sub_type).to eq 'vnd.adobe.photoshop'
+      expect(empty.sub_type).to eq 'x-empty'
+      expect(bytes.sub_type).to eq 'octet-stream'
+    end
+  end
+
+  describe '#mime_type' do
+    it 'returns a MIME::Type instance' do
+      [txt, csv, psd, bytes].each do |fileinfo|
+        expect(fileinfo.mime_type).to be_a MIME::Type
+      end
+    end
+
+    it 'returns nil for empty file' do
+      expect(empty.mime_type).to eq nil
+    end
+
+    context 'when "mime-types" gem isn’t available' do
+      before do
+        MIME.send(:remove_const, :Types)
+        allow_any_instance_of(FileInfo).to receive(:require).with('mime/types').and_raise(LoadError)
+      end
+
+      it 'raises LoadError with custom message' do
+        expect { txt.mime_type }.to raise_error(LoadError, FileInfo::MIME_TYPE_ERROR_MESSAGE)
+      end
     end
   end
 end

data/spec/fixtures/bytes

@@ -0,0 +1 @@
+�Ɋ���z��S��Q_

metadata

@@ -1,68 +1,82 @@
 --- !ruby/object:Gem::Specification
 name: fileinfo
 version: !ruby/object:Gem::Version
-  version: 0.2.0
+  version: 0.3.0
 platform: ruby
 authors:
 - Rafaël Blais Masson
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2013-10-30 00:00:00.000000000 Z
+date: 2014-04-10 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: bundler
   requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - ~>
+    - - "~>"
       - !ruby/object:Gem::Version
         version: '1.3'
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - ~>
+    - - "~>"
       - !ruby/object:Gem::Version
         version: '1.3'
 - !ruby/object:Gem::Dependency
   name: rake
   requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - '>='
+    - - ">="
       - !ruby/object:Gem::Version
         version: '0'
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - '>='
+    - - ">="
       - !ruby/object:Gem::Version
         version: '0'
 - !ruby/object:Gem::Dependency
   name: rspec
   requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - ~>
+    - - "~>"
       - !ruby/object:Gem::Version
         version: '2.14'
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - ~>
+    - - "~>"
       - !ruby/object:Gem::Version
         version: '2.14'
-description: FileInfo detects encoding from strings and files using the wonderful
-  Unix `file` command.
+- !ruby/object:Gem::Dependency
+  name: mime-types
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
+description: FileInfo detects file encodings and MIME types using the wonderful Unix
+  `file` command.
 email:
 - rafael@heliom.ca
 executables: []
 extensions: []
 extra_rdoc_files: []
 files:
-- .gitignore
-- .rspec
-- .travis.yml
+- ".gitignore"
+- ".rspec"
+- ".travis.yml"
 - Gemfile
 - LICENSE.txt
 - README.md
@@ -72,11 +86,13 @@ files:
 - lib/file_info/version.rb
 - lib/fileinfo.rb
 - spec/file_info_spec.rb
+- spec/fixtures/bytes
 - spec/fixtures/encoding_ascii.csv
 - spec/fixtures/encoding_isolatin.csv
 - spec/fixtures/encoding_isowindows.csv
 - spec/fixtures/encoding_macroman.csv
 - spec/fixtures/encoding_utf8.csv
+- spec/fixtures/mockup.psd
 - spec/spec_helper.rb
 homepage: http://github.com/rafBM/fileinfo
 licenses:
@@ -88,26 +104,28 @@ require_paths:
 - lib
 required_ruby_version: !ruby/object:Gem::Requirement
   requirements:
-  - - '>='
+  - - ">="
     - !ruby/object:Gem::Version
       version: '0'
 required_rubygems_version: !ruby/object:Gem::Requirement
   requirements:
-  - - '>='
+  - - ">="
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
 rubyforge_project: 
-rubygems_version: 2.1.9
+rubygems_version: 2.2.2
 signing_key: 
 specification_version: 4
-summary: FileInfo detects encoding from strings and files using the wonderful Unix
-  `file` command.
+summary: FileInfo detects file encodings and MIME types using the wonderful Unix `file`
+  command.
 test_files:
 - spec/file_info_spec.rb
+- spec/fixtures/bytes
 - spec/fixtures/encoding_ascii.csv
 - spec/fixtures/encoding_isolatin.csv
 - spec/fixtures/encoding_isowindows.csv
 - spec/fixtures/encoding_macroman.csv
 - spec/fixtures/encoding_utf8.csv
+- spec/fixtures/mockup.psd
 - spec/spec_helper.rb