yomu 0.1.5 → 0.1.6

data/README.md

@@ -15,13 +15,14 @@ For the complete list of supported formats, please visit the Apache Tika
 
 ## Usage
 
-Text and metadata can be extracted by calling `Yomu.read` directly:
+Text, metadata and MIME type information can be extracted by calling `Yomu.read` directly:
 
     require 'yomu'
 
     data = File.read 'sample.pages'
     text = Yomu.read :text, data
     metadata = Yomu.read :metadata, data
+    mimetype = Yomu.read :mimetype, data
 
 ### Reading text from a given filename
 
@@ -53,6 +54,14 @@ Metadata is returned as a hash.
     yomu = Yomu.new 'sample.pages'
     yomu.metadata['Content-Type'] #=> "application/vnd.apple.pages"
 
+### Reading MIME types
+
+MIME type is returned as a MIME::Type object.
+
+    yomu = Yomu.new 'sample.docx'
+    yomu.mimetype.content_type #=> "application/vnd.openxmlformats-officedocument.wordprocessingml.document"
+    yomu.mimetype.extensions #=> ['docx']
+
 ## Installation and Dependencies
 
 Add this line to your application's Gemfile:

data/lib/yomu/version.rb

@@ -1,3 +1,3 @@
 class Yomu
-  VERSION = "0.1.5"
+  VERSION = "0.1.6"
 end

data/lib/yomu.rb

@@ -1,6 +1,7 @@
 require 'yomu/version'
 
 require 'net/http'
+require 'mime/types'
 require 'yaml'
 
 class Yomu
@@ -19,6 +20,8 @@ class Yomu
       '-t'
     when :metadata
       '-m'
+    when :mimetype
+      '-m'
     end
 
     result = IO.popen "#{java} -Djava.awt.headless=true -jar #{Yomu::JARPATH} #{switch}", 'r+' do |io|
@@ -27,7 +30,14 @@ class Yomu
       io.read
     end
 
-    type == :metadata ? YAML.load(result) : result
+    case type
+    when :text
+      result
+    when :metadata
+      YAML.load enclose_metadata_fields(result)
+    when :mimetype
+      MIME::Types[YAML.load(enclose_metadata_fields(result))['Content-Type']].first
+    end
   end
 
   # Create a new instance of Yomu with a given document.
@@ -82,6 +92,18 @@ class Yomu
     @metadata = Yomu.read :metadata, data
   end
 
+  # Returns the mimetype object of the Yomu document.
+  #
+  #   yomu = Yomu.new 'sample.docx'
+  #   yomu.mimetype.content_type #=> 'application/vnd.openxmlformats-officedocument.wordprocessingml.document'
+  #   yomu.mimetype.extensions #=> ['docx']
+
+  def mimetype
+    return @mimetype if defined? @mimetype
+
+    @mimetype = MIME::Types[metadata['Content-Type']].first
+  end
+
   # Returns +true+ if the Yomu document was specified using a file path.
   #
   #   yomu = Yomu.new 'sample.pages'
@@ -129,6 +151,12 @@ class Yomu
     @data
   end
 
+  def self.enclose_metadata_fields metadata
+    metadata.each_line.map do |l|
+      l.gsub(/: (.*)/,': "\1"')
+    end.join
+  end
+
   def self.java
     ENV['JAVA_HOME'] ? ENV['JAVA_HOME'] + '/bin/java' : 'java'
   end

data/test/specs/yomu.rb

@@ -1,7 +1,7 @@
 require_relative '../helper.rb'
 
 describe Yomu do
-  let(:data) { File.read 'test/samples/sample.pages' }
+  let(:data) { File.read 'test/samples/sample.docx' }
 
   before do
     ENV['JAVA_HOME'] = nil
@@ -17,7 +17,21 @@ describe Yomu do
     it 'reads metadata' do
       metadata = Yomu.read :metadata, data
 
-      assert_equal 'application/vnd.apple.pages', metadata['Content-Type']
+      assert_equal 'application/vnd.openxmlformats-officedocument.wordprocessingml.document', metadata['Content-Type']
+    end
+
+    it 'accepts metadata with colon' do
+      doc = File.read 'test/samples/enclosure_problem.doc'
+      metadata = Yomu.read :metadata, doc
+
+      assert_equal 'problem: test', metadata['dc:title']
+    end
+
+    it 'reads mimetype' do
+      mimetype = Yomu.read :mimetype, data
+
+      assert_equal 'application/vnd.openxmlformats-officedocument.wordprocessingml.document', mimetype.content_type
+      assert_includes mimetype.extensions, 'docx'
     end
   end
 

data/yomu.gemspec

@@ -15,5 +15,7 @@ Gem::Specification.new do |gem|
   gem.require_paths = ["lib"]
   gem.version       = Yomu::VERSION
 
+  gem.add_runtime_dependency 'mime-types', '~> 1.23'
+
   gem.add_development_dependency 'minitest'
 end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: yomu
 version: !ruby/object:Gem::Version
-  version: 0.1.5
+  version: 0.1.6
   prerelease: 
 platform: ruby
 authors:
@@ -9,8 +9,24 @@ authors:
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2013-02-10 00:00:00.000000000 Z
+date: 2013-05-01 00:00:00.000000000 Z
 dependencies:
+- !ruby/object:Gem::Dependency
+  name: mime-types
+  requirement: !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ~>
+      - !ruby/object:Gem::Version
+        version: '1.23'
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ~>
+      - !ruby/object:Gem::Version
+        version: '1.23'
 - !ruby/object:Gem::Dependency
   name: minitest
   requirement: !ruby/object:Gem::Requirement
@@ -45,7 +61,9 @@ files:
 - lib/yomu.rb
 - lib/yomu/version.rb
 - test/helper.rb
+- test/samples/enclosure_problem.doc
 - test/samples/sample filename with spaces.pages
+- test/samples/sample.docx
 - test/samples/sample.pages
 - test/specs/yomu.rb
 - yomu.gemspec
@@ -76,6 +94,8 @@ summary: Read text and metadata from files and documents (.doc, .docx, .pages, .
   .rtf, .pdf)
 test_files:
 - test/helper.rb
+- test/samples/enclosure_problem.doc
 - test/samples/sample filename with spaces.pages
+- test/samples/sample.docx
 - test/samples/sample.pages
 - test/specs/yomu.rb