hydra-derivatives 0.0.2 → 0.0.3

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 74b15261c78629ff97614bc0090102ceed0c5fa7
-  data.tar.gz: 268d962f2ea5607db494c67455878d8cb8eab25d
+  metadata.gz: 6cb23951238b5918e0ea720629b80abbbf7c36f9
+  data.tar.gz: c518b9251a6869b1f76dc6095fcaa818f447e959
 SHA512:
-  metadata.gz: 5c7c0aa06d5fd7059e4bb9886d2f241ff5a7f00508941bb7650907464e0a3d59e1d884fddc61208068930a263c420e952534bb91c5931deadd896695ba8fd607
-  data.tar.gz: 5b5b6f7800798afe8dd7688638bdf88f643c6990d1cb0dad011d2f36c54238f2fc48e97f5fdaef6664d84daca620ed0b9ac9a0d8b2b0c0a51eeab843a7ecd84e
+  metadata.gz: fcc714f73eeda9db4e5c9d8bc3cfd62675d7f8af210abca8876d68ca9315479ff751624ac2c78c65b3314f08e1e5c369cff4c9edae21ec1c51fc20fb82ff977b
+  data.tar.gz: 08f350ea7eb44750e37e19c1e3082badc71c01258541f3fe5533887b4718517cad0f8591f538ce54959ed543d9991509fdafbeb0d819fe2738fc6b3e092aec5b

data/.rspec

@@ -0,0 +1 @@
+--colour

data/README.md

@@ -10,15 +10,47 @@ If you have an ActiveFedora class like this:
         
         has_file_datastream :content
         attr_accessor :mime_type
+        
+        # Use a block to declare which derivatives you want to generate
+        makes_derivatives do |obj| 
+          case obj.mime_type
+          when 'application/pdf'
+            obj.transform_datastream :content, { :thumb => "100x100>" }
+          when 'audio/wav'
+            obj.transform_datastream :content, { :mp3 => {format: 'mp3'}, :ogg => {format: 'ogg'} }, processor: :audio
+          when 'video/avi'
+            obj.transform_datastream :content, { :mp4 => {format: 'mp4'}, :webm => {format: 'webm'} }, processor: :video
+          when 'image/png', 'image/jpg'
+            obj.transform_datastream :content, { :medium => "300x300>", :thumb => "100x100>" }
+          end
+        end
+    end
+```
+
+Or a class like this:
 
-        makes_derivatives_of :content, when: :mime_type, is_one_of: ['image/tiff', 'image/jpeg],
-             derivatives: { :thumbnail => {size: "200x150>", datastream: 'thumbnail'} }
-        makes_derivatives_of :content, when: :mime_type, is: 'application/pdf',
-             derivatives: { :thumbnail => {size: "338x493", datastream: 'thumbnail'} }
-        makes_derivatives_of :content, when: :mime_type, is_one_of: ['video/mpeg', 'video/avi'],
-             derivatives: { :webm => {format: "webm", datastream: 'webm'}, :mp4 => {format: "mp4", datastream: 'mp4'} }, processors: :video
-        makes_derivatives_of :content, when: :mime_type, is_one_of: ['audio/wav', 'audio/mpeg'],
-             derivatives: {  :mp3 => {format: 'mp3', datastream: 'mp3'}, :ogg => {format: 'ogg', datastream: 'ogg'} }, processors: :audio
+```ruby
+    class GenericFile < ActiveFedora::Base
+        include Hydra::Derivatives
+    
+        has_file_datastream :content
+        attr_accessor :mime_type
+
+        # Use a callback method to declare which derivatives you want
+        makes_derivatives :generate_derivatives
+        
+        def generate_derivatives
+          case mime_type
+          when 'application/pdf'
+            transform_datastream :content, { :thumb => "100x100>" }
+          when 'audio/wav'
+            transform_datastream :content, { :mp3 => {format: 'mp3'}, :ogg => {format: 'ogg'} }, processor: :audio
+          when 'video/avi'
+            transform_datastream :content, { :mp4 => {format: 'mp4'}, :webm => {format: 'webm'} }, processor: :video
+          when 'image/png', 'image/jpg'
+            transform_datastream :content, { :medium => "300x300>", :thumb => "100x100>" }
+          end
+        end
     end
 ```
 
@@ -34,3 +66,9 @@ And you add some content to it:
 Then when you call `obj.create_derivatives` a new datastream, called 'thumbnail' will have been created, with a downsized image in it.
 
 We recommend you run `obj.create_derivatives` in a background worker, because some derivative creation (especially videos) can take a long time.
+
+You may want to adjust your path to add LibreOffice and Fits.sh support:
+```bash
+# in .bash_profile
+export PATH=${PATH}:/Users/justin/workspace/fits-0.6.2:/Applications/LibreOffice.app/Contents/MacOS
+```

data/VERSION

@@ -1 +1 @@
-0.0.2
+0.0.3

data/lib/hydra/derivatives.rb

@@ -10,105 +10,104 @@ module Hydra
     autoload :Ffmpeg
     autoload :Video
     autoload :Audio
+    autoload :Config
+    autoload :Document
     autoload :ExtractMetadata
+    autoload :ShellBasedProcessor
 
-    def self.ffmpeg_path=(val)
-      @ffmpeg_path = val
-    end
-    def self.ffmpeg_path
-      #Sufia.config.ffmpeg_path
-      @ffmpeg_path ||= 'ffmpeg'
+    def self.config
+      @config ||= reset_config!
     end
 
-    def self.temp_file_base=(val)
-      @temp_file_base = val
-    end
-    def self.temp_file_base
-      #Sufia.config.temp_file_base
-      @temp_file_base ||= '/tmp'
-    end
-
-    def self.fits_path=(val)
-      @fits_path = val
-    end
-    def self.fits_path
-      #Sufia.config.fits_path
-      @fits_path ||= 'fits.sh'
+    def self.reset_config!
+      @config = Config.new
     end
 
-    def self.enable_ffmpeg=(val)
-      @enable_ffmpeg = val
-    end
-    def self.enable_ffmpeg
-      @enable_ffmpeg ||= true
+    [:ffmpeg_path, :libreoffice_path, :temp_file_base, :fits_path, :enable_ffmpeg].each do |method|
+      module_eval <<-RUBY
+        def self.#{method.to_s}
+          config.#{method.to_s}
+        end
+        def self.#{method.to_s}= val
+          config.#{method.to_s}= val
+        end
+      RUBY
     end
 
     included do
-      class_attribute :transformation_scheme
+      class_attribute :transformation_schemes
     end
 
 
     # Runs all of the transformations immediately.
     # You may want to run this job in the background as it may take a long time.
     def create_derivatives 
-      if transformation_scheme.present?
-        transformation_scheme.each do |datastream, value|
-          transform_datastream(datastream, value) if self.datastreams[datastream.to_s].has_content?
+      if transformation_schemes.present?
+        transformation_schemes.each do |transform_scheme|
+          if transform_scheme.instance_of?(Proc)
+            transform_scheme.call(self) 
+          else
+            send(transform_scheme)
+          end
         end
       else
         logger.warn "`create_derivatives' was called on an instance of #{self.class}, but no derivatives have been requested"
       end
     end
 
-    # Transform a single datastream
-    def transform_datastream(datastream, directive_list)
-      directive_list.each do |directive|
-        if directive.applies?(self)
-          processor = directive.processors ? Array(directive.processors).first : :image
-          "Hydra::Derivatives::#{processor.to_s.classify}".constantize.new(self, datastream, directive.derivatives).process
-        end
-      end
-
+    # Create derivatives from a datastream according to transformation directives
+    # @param datastream_name 
+    # @param [Hash] transform_directives - each key corresponds to a desired derivative.  Associated values vary according to processor being used.
+    # @param [Hash] opts for specifying things like choice of :processor (processor defaults to :image)
+    # 
+    # @example This will create content_thumb
+    #   transform_datastream :content, { :thumb => "100x100>" }
+    #
+    # @example Specify the dsid for the output datastream
+    #   transform_datastream :content, { :thumb => {size: "200x300>", datastream: 'thumbnail'} }
+    #
+    # @example Create multiple derivatives with one set of directives.  This will create content_thumb and content_medium
+    #   transform_datastream :content, { :medium => "300x300>", :thumb => "100x100>" }
+    #
+    # @example Specify which processor you want to use (defaults to :image)
+    #   transform_datastream :content, { :mp3 => {format: 'mp3'}, :ogg => {format: 'ogg'} }, processor: :audio
+    #   transform_datastream :content, { :mp4 => {format: 'mp4'}, :webm => {format: 'webm'} }, processor: :video
+    #
+    def transform_datastream(datastream_name, transform_directives, opts={})
+      processor = opts[:processor] ? opts[:processor] : :image
+      "Hydra::Derivatives::#{processor.to_s.classify}".constantize.new(self, datastream_name, transform_directives).process
     end
 
-    class TransformationDirective
-      attr_accessor :differentiator, :selector, :derivatives, :processors
-      # @param [Hash] args the options 
-      # @option args [Symbol] :when the method that holds the differentiator column
-      # @option args [String, Array] :is_one_of activates this set of derivatives when the the differentiator column is includes one of these. 
-      # @option args [String, Array] :is alias for :is_one_of 
-      # @option args [Hash] :derivatives the derivatives to be produced
-      # @option args [Symbol, Array] :processors the processors to run to produce the derivatives 
-      def initialize(args)
-        self.differentiator = args[:when]
-        self.selector = args[:is_one_of] || args[:is]
-        self.derivatives = args[:derivatives]
-        self.processors = args[:processors]
-      end
-
-      def applies?(object)
-        selector.include?(object.send(differentiator))
-      end
-    end
 
     module ClassMethods
-      # @param [Symbol, String] datastream the datastream to operate on
-      # @param [Hash] args the options 
-      # @option args [Symbol] :when the method that holds the differentiator column
-      # @option args [String, Array] :is_one_of activates this set of derivatives when the the differentiator column is includes one of these. 
-      # @option args [String, Array] :is alias for :is_one_of 
-      # @option args [Hash] :derivatives the derivatives to be produced
-      # @option args [Symbol, Array] :processors the processors to run to produce the derivatives 
-      # @example
-      #    makes_derivatives_of :content, when: :mime_type, is: 'text/pdf',
-      #        derivatives: { :text => { :quality => :better }, processors: [:ocr]}
+      # Register transformation schemes for generating derivatives.  
+      # You can do this using a block or by defining a callback method.
       #
-      #    makes_derivatives_of :content, when: :mime_type, is_one_of: ['image/png', 'image/jpg'],
-      #        derivatives: { :medium => "300x300>", :thumb => "100x100>" }
-      def makes_derivatives_of(datastream, args = {})
-        self.transformation_scheme ||= {}
-        self.transformation_scheme[datastream.to_sym] ||= []
-        self.transformation_scheme[datastream.to_sym] << TransformationDirective.new(args)
+      # @example Define transformation scheme using a block
+      #    makes_derivatives do |obj| 
+      #      case obj.mime_type
+      #      when 'application/pdf'
+      #        obj.transform_datastream :content, { :thumb => "100x100>" }
+      #      when 'audio/wav'
+      #        obj.transform_datastream :content, { :mp3 => {format: 'mp3'}, :ogg => {format: 'ogg'} }, processor: :audio
+      #
+      # @example Define transformation scheme using a callback method
+      #      makes_derivatives :generate_image_derivatives
+      #
+      #      def generate_image_derivatives
+      #        case mime_type
+      #        when 'image/png', 'image/jpg'
+      #          transform_datastream :content, { :medium => "300x300>", :thumb => "100x100>" }
+      #        end
+      #      end
+      def makes_derivatives(*callback_method_names, &block)
+        self.transformation_schemes ||= []
+        if block_given?
+          self.transformation_schemes << block
+        end
+        callback_method_names.each do |callback_name|
+          self.transformation_schemes << callback_name
+        end
       end
     end
   end

data/lib/hydra/derivatives/audio.rb

@@ -5,9 +5,6 @@ module Hydra
 
       protected
 
-      def options_for(format)
-      end
-
       def new_mime_type(format)
         case format
         when 'mp3'

data/lib/hydra/derivatives/config.rb

@@ -0,0 +1,27 @@
+module Hydra
+  module Derivatives
+    class Config
+      attr_writer :ffmpeg_path, :libreoffice_path, :temp_file_base, :fits_path, :enable_ffmpeg
+      def ffmpeg_path
+        @ffmpeg_path ||= 'ffmpeg'
+      end
+
+      def libreoffice_path
+        @libreoffice_path ||= 'soffice'
+      end
+
+      def temp_file_base
+        @temp_file_base ||= '/tmp'
+      end
+
+      def fits_path
+        @fits_path ||= 'fits.sh'
+      end
+
+      def enable_ffmpeg
+        @enable_ffmpeg ||= true
+      end
+
+    end
+  end
+end

data/lib/hydra/derivatives/document.rb

@@ -0,0 +1,36 @@
+module Hydra
+  module Derivatives
+    class Document < Processor
+      include ShellBasedProcessor
+      def self.encode(path, options, output_file)
+        format = File.extname(output_file).sub('.', '')
+        outdir = File.dirname(output_file)
+        execute "#{Hydra::Derivatives.libreoffice_path} --invisible --headless --convert-to #{format} --outdir #{outdir} #{path}"
+      end
+
+      def encode_datastream(dest_dsid, file_suffix, mime_type, options = '')
+        output_file = Dir::Tmpname.create(['sufia', ".#{file_suffix}"], Hydra::Derivatives.temp_file_base){}
+        new_output = ''
+        source_datastream.to_tempfile do |f|
+          self.class.encode(f.path, options, output_file)
+          new_output = File.join(Hydra::Derivatives.temp_file_base, [File.basename(f.path).sub(File.extname(f.path), ''), file_suffix].join('.'))
+        end
+        out_file = File.open(new_output, "rb")
+        object.add_file_datastream(out_file.read, :dsid=>dest_dsid, :mimeType=>mime_type)
+        File.unlink(out_file)
+      end
+
+
+      def new_mime_type(format)
+        case format
+        when 'pdf'
+          'application/pdf'
+        when 'odf'
+          'application/vnd.oasis.opendocument.text'
+        else
+          raise "I don't know about the format '#{format}'"
+        end
+      end
+    end
+  end
+end

data/lib/hydra/derivatives/extract_metadata.rb

@@ -14,8 +14,10 @@ module Hydra
 
       def to_tempfile &block
         return unless has_content?
-        tmp_base = Hydra::Derivatives.temp_file_base
-        f = Tempfile.new("#{pid}-#{dsVersionID}")
+        list_of_extensions = MIME::Types[mimeType].first.extensions
+        extension = ".#{list_of_extensions.first}" if list_of_extensions
+
+        f = Tempfile.new(["#{pid}-#{dsVersionID}", extension])
         f.binmode
         if content.respond_to? :read
           f.write(content.read)

data/lib/hydra/derivatives/ffmpeg.rb

@@ -9,41 +9,14 @@ module Hydra
       extend ActiveSupport::Concern
 
       included do
-        extend Open3
+        include ShellBasedProcessor
       end
 
-      def process
-        directives.each do |name, args| 
-          format = args[:format]
-          raise ArgumentError, "You must provide the :format you want to transcode into. You provided #{args}" unless format
-          # TODO if the source is in the correct format, we could just copy it and skip transcoding.
-          output_datastream_name = args[:datastream] || output_datastream_id(name)
-          encode_datastream(output_datastream_name, format, new_mime_type(format), options_for(format))
-        end
-      end
-
-      def encode_datastream(dest_dsid, file_suffix, mime_type, options = '')
-        out_file = nil
-        output_file = Dir::Tmpname.create(['sufia', ".#{file_suffix}"], Hydra::Derivatives.temp_file_base){}
-        source_datastream.to_tempfile do |f|
-          self.class.encode(f.path, options, output_file)
-        end
-        out_file = File.open(output_file, "rb")
-        object.add_file_datastream(out_file.read, :dsid=>dest_dsid, :mimeType=>mime_type)
-        File.unlink(output_file)
-      end
 
       module ClassMethods
 
         def encode(path, options, output_file)
-          command = "#{Hydra::Derivatives.ffmpeg_path} -y -i \"#{path}\" #{options} #{output_file}"
-          stdin, stdout, stderr, wait_thr = popen3(command)
-          stdin.close
-          out = stdout.read
-          stdout.close
-          err = stderr.read
-          stderr.close
-          raise "Unable to execute command \"#{command}\"\n#{err}" unless wait_thr.value.success?
+          execute "#{Hydra::Derivatives.ffmpeg_path} -y -i \"#{path}\" #{options} #{output_file}"
         end
       end
 

data/lib/hydra/derivatives/shell_based_processor.rb

@@ -0,0 +1,58 @@
+# An abstract class for asyncronous jobs that transcode files using FFMpeg
+
+require 'tmpdir'
+require 'open3'
+
+module Hydra
+  module Derivatives
+    module ShellBasedProcessor 
+      extend ActiveSupport::Concern
+
+      included do
+        extend Open3
+      end
+
+      def process
+        directives.each do |name, args| 
+          format = args[:format]
+          raise ArgumentError, "You must provide the :format you want to transcode into. You provided #{args}" unless format
+          # TODO if the source is in the correct format, we could just copy it and skip transcoding.
+          output_datastream_name = args[:datastream] || output_datastream_id(name)
+          encode_datastream(output_datastream_name, format, new_mime_type(format), options_for(format))
+        end
+      end
+
+      # override this method in subclass if you want to provide specific options.
+      def options_for(format)
+      end
+
+      def encode_datastream(dest_dsid, file_suffix, mime_type, options = '')
+        out_file = nil
+        output_file = Dir::Tmpname.create(['sufia', ".#{file_suffix}"], Hydra::Derivatives.temp_file_base){}
+        source_datastream.to_tempfile do |f|
+          self.class.encode(f.path, options, output_file)
+        end
+        out_file = File.open(output_file, "rb")
+        object.add_file_datastream(out_file.read, :dsid=>dest_dsid, :mimeType=>mime_type)
+        File.unlink(output_file)
+      end
+
+      module ClassMethods
+
+        def execute(command)
+          stdin, stdout, stderr, wait_thr = popen3(command)
+          stdin.close
+          out = stdout.read
+          stdout.close
+          err = stderr.read
+          stderr.close
+          raise "Unable to execute command \"#{command}\"\n#{err}" unless wait_thr.value.success?
+        end
+
+      end
+
+    end
+  end
+end
+
+

data/spec/fixtures/sample.rtf

@@ -0,0 +1,68 @@
+{\rtf1\ansi\ansicpg1252\uc1\deff0\stshfdbch0\stshfloch0\stshfhich0\stshfbi0\deflang1033\deflangfe1033{\fonttbl{\f0\froman\fcharset0\fprq2{\*\panose 02020603050405020304}Times New Roman;}{\f1\fswiss\fcharset0\fprq2{\*\panose 020b0604020202020204}Arial;}
+{\f2\fmodern\fcharset0\fprq1{\*\panose 02070309020205020404}Courier New;}{\f3\froman\fcharset2\fprq2{\*\panose 05050102010706020507}Symbol;}{\f10\fnil\fcharset2\fprq2{\*\panose 05000000000000000000}Wingdings;}
+{\f121\froman\fcharset238\fprq2 Times New Roman CE;}{\f122\froman\fcharset204\fprq2 Times New Roman Cyr;}{\f124\froman\fcharset161\fprq2 Times New Roman Greek;}{\f125\froman\fcharset162\fprq2 Times New Roman Tur;}
+{\f126\froman\fcharset177\fprq2 Times New Roman (Hebrew);}{\f127\froman\fcharset178\fprq2 Times New Roman (Arabic);}{\f128\froman\fcharset186\fprq2 Times New Roman Baltic;}{\f129\froman\fcharset163\fprq2 Times New Roman (Vietnamese);}
+{\f131\fswiss\fcharset238\fprq2 Arial CE;}{\f132\fswiss\fcharset204\fprq2 Arial Cyr;}{\f134\fswiss\fcharset161\fprq2 Arial Greek;}{\f135\fswiss\fcharset162\fprq2 Arial Tur;}{\f136\fswiss\fcharset177\fprq2 Arial (Hebrew);}
+{\f137\fswiss\fcharset178\fprq2 Arial (Arabic);}{\f138\fswiss\fcharset186\fprq2 Arial Baltic;}{\f139\fswiss\fcharset163\fprq2 Arial (Vietnamese);}{\f141\fmodern\fcharset238\fprq1 Courier New CE;}{\f142\fmodern\fcharset204\fprq1 Courier New Cyr;}
+{\f144\fmodern\fcharset161\fprq1 Courier New Greek;}{\f145\fmodern\fcharset162\fprq1 Courier New Tur;}{\f146\fmodern\fcharset177\fprq1 Courier New (Hebrew);}{\f147\fmodern\fcharset178\fprq1 Courier New (Arabic);}
+{\f148\fmodern\fcharset186\fprq1 Courier New Baltic;}{\f149\fmodern\fcharset163\fprq1 Courier New (Vietnamese);}}{\colortbl;\red0\green0\blue0;\red0\green0\blue255;\red0\green255\blue255;\red0\green255\blue0;\red255\green0\blue255;\red255\green0\blue0;
+\red255\green255\blue0;\red255\green255\blue255;\red0\green0\blue128;\red0\green128\blue128;\red0\green128\blue0;\red128\green0\blue128;\red128\green0\blue0;\red128\green128\blue0;\red128\green128\blue128;\red192\green192\blue192;}{\stylesheet{
+\ql \li0\ri0\widctlpar\aspalpha\aspnum\faauto\adjustright\rin0\lin0\itap0 \fs24\lang1033\langfe1033\cgrid\langnp1033\langfenp1033 \snext0 Normal;}{\s1\ql \li0\ri0\sb240\sa60\keepn\widctlpar\aspalpha\aspnum\faauto\outlinelevel0\adjustright\rin0\lin0\itap0 
+\b\f1\fs32\lang1033\langfe1033\kerning32\cgrid\langnp1033\langfenp1033 \sbasedon0 \snext0 \styrsid2294299 heading 1;}{\*\cs10 \additive \ssemihidden Default Paragraph Font;}{\*
+\ts11\tsrowd\trftsWidthB3\trpaddl108\trpaddr108\trpaddfl3\trpaddft3\trpaddfb3\trpaddfr3\tscellwidthfts0\tsvertalt\tsbrdrt\tsbrdrl\tsbrdrb\tsbrdrr\tsbrdrdgl\tsbrdrdgr\tsbrdrh\tsbrdrv 
+\ql \li0\ri0\widctlpar\aspalpha\aspnum\faauto\adjustright\rin0\lin0\itap0 \fs20\lang1024\langfe1024\cgrid\langnp1024\langfenp1024 \snext11 \ssemihidden Normal Table;}{\*\ts15\tsrowd\trbrdrt\brdrs\brdrw10 \trbrdrl\brdrs\brdrw10 \trbrdrb\brdrs\brdrw10 
+\trbrdrr\brdrs\brdrw10 \trbrdrh\brdrs\brdrw10 \trbrdrv\brdrs\brdrw10 \trftsWidthB3\trpaddl108\trpaddr108\trpaddfl3\trpaddft3\trpaddfb3\trpaddfr3\tscellwidthfts0\tsvertalt\tsbrdrt\tsbrdrl\tsbrdrb\tsbrdrr\tsbrdrdgl\tsbrdrdgr\tsbrdrh\tsbrdrv 
+\ql \li0\ri0\widctlpar\aspalpha\aspnum\faauto\adjustright\rin0\lin0\itap0 \fs20\lang1024\langfe1024\cgrid\langnp1024\langfenp1024 \sbasedon11 \snext15 \styrsid2294299 Table Grid;}{
+\s16\ql \li0\ri0\widctlpar\aspalpha\aspnum\faauto\adjustright\rin0\lin0\itap0 \fs20\lang1033\langfe1033\cgrid\langnp1033\langfenp1033 \sbasedon0 \snext16 \ssemihidden \styrsid1792631 footnote text;}{\*\cs17 \additive \super 
+\sbasedon10 \ssemihidden \styrsid1792631 footnote reference;}}{\*\latentstyles\lsdstimax156\lsdlockeddef0}{\*\listtable{\list\listtemplateid-767292450\listhybrid{\listlevel\levelnfc23\levelnfcn23\leveljc0\leveljcn0\levelfollow0\levelstartat1\levelspace360
+\levelindent0{\leveltext\leveltemplateid67698689\'01\u-3913 ?;}{\levelnumbers;}\f3\fbias0 \fi-360\li720\jclisttab\tx720\lin720 }{\listlevel\levelnfc23\levelnfcn23\leveljc0\leveljcn0\levelfollow0\levelstartat1\levelspace360\levelindent0{\leveltext
+\leveltemplateid67698691\'01o;}{\levelnumbers;}\f2\fbias0 \fi-360\li1440\jclisttab\tx1440\lin1440 }{\listlevel\levelnfc23\levelnfcn23\leveljc0\leveljcn0\levelfollow0\levelstartat1\levelspace360\levelindent0{\leveltext\leveltemplateid67698693
+\'01\u-3929 ?;}{\levelnumbers;}\f10\fbias0 \fi-360\li2160\jclisttab\tx2160\lin2160 }{\listlevel\levelnfc23\levelnfcn23\leveljc0\leveljcn0\levelfollow0\levelstartat1\levelspace360\levelindent0{\leveltext\leveltemplateid67698689\'01\u-3913 ?;}{\levelnumbers
+;}\f3\fbias0 \fi-360\li2880\jclisttab\tx2880\lin2880 }{\listlevel\levelnfc23\levelnfcn23\leveljc0\leveljcn0\levelfollow0\levelstartat1\levelspace360\levelindent0{\leveltext\leveltemplateid67698691\'01o;}{\levelnumbers;}\f2\fbias0 \fi-360\li3600
+\jclisttab\tx3600\lin3600 }{\listlevel\levelnfc23\levelnfcn23\leveljc0\leveljcn0\levelfollow0\levelstartat1\levelspace360\levelindent0{\leveltext\leveltemplateid67698693\'01\u-3929 ?;}{\levelnumbers;}\f10\fbias0 \fi-360\li4320\jclisttab\tx4320\lin4320 }
+{\listlevel\levelnfc23\levelnfcn23\leveljc0\leveljcn0\levelfollow0\levelstartat1\levelspace360\levelindent0{\leveltext\leveltemplateid67698689\'01\u-3913 ?;}{\levelnumbers;}\f3\fbias0 \fi-360\li5040\jclisttab\tx5040\lin5040 }{\listlevel\levelnfc23
+\levelnfcn23\leveljc0\leveljcn0\levelfollow0\levelstartat1\levelspace360\levelindent0{\leveltext\leveltemplateid67698691\'01o;}{\levelnumbers;}\f2\fbias0 \fi-360\li5760\jclisttab\tx5760\lin5760 }{\listlevel\levelnfc23\levelnfcn23\leveljc0\leveljcn0
+\levelfollow0\levelstartat1\levelspace360\levelindent0{\leveltext\leveltemplateid67698693\'01\u-3929 ?;}{\levelnumbers;}\f10\fbias0 \fi-360\li6480\jclisttab\tx6480\lin6480 }{\listname ;}\listid687222349}}{\*\listoverridetable{\listoverride\listid687222349
+\listoverridecount0\ls1}}{\*\rsidtbl \rsid1792631\rsid2294299}{\*\generator Microsoft Word 11.0.6113;}{\info{\title This is a test RTF}{\author Nate}{\operator Nate}{\version2}}\widowctrl\ftnbj\aenddoc\noxlattoyen\expshrtn\noultrlspc\dntblnsbdb\nospaceforul\hyphcaps0\formshade\horzdoc\dgmargin\dghspace180\dgvspace180\dghorigin1800\dgvorigin1440
+\dghshow1\dgvshow1\jexpand\viewkind1\viewscale80\pgbrdrhead\pgbrdrfoot\splytwnine\ftnlytwnine\htmautsp\nolnhtadjtbl\useltbaln\alntblind\lytcalctblwd\lyttblrtgr\lnbrkrule\nobrkwrptbl\snaptogridincell\allowfieldendsel\wrppunct
+\asianbrkrule\rsidroot2294299\newtblstyruls\nogrowautofit \fet0{\*\ftnsep \pard\plain \ql \li0\ri0\widctlpar\aspalpha\aspnum\faauto\adjustright\rin0\lin0\itap0 \fs24\lang1033\langfe1033\cgrid\langnp1033\langfenp1033 {\insrsid1792631 \chftnsep 
+\par }}{\*\ftnsepc \pard\plain \ql \li0\ri0\widctlpar\aspalpha\aspnum\faauto\adjustright\rin0\lin0\itap0 \fs24\lang1033\langfe1033\cgrid\langnp1033\langfenp1033 {\insrsid1792631 \chftnsepc 
+\par }}{\*\aftnsep \pard\plain \ql \li0\ri0\widctlpar\aspalpha\aspnum\faauto\adjustright\rin0\lin0\itap0 \fs24\lang1033\langfe1033\cgrid\langnp1033\langfenp1033 {\insrsid1792631 \chftnsep 
+\par }}{\*\aftnsepc \pard\plain \ql \li0\ri0\widctlpar\aspalpha\aspnum\faauto\adjustright\rin0\lin0\itap0 \fs24\lang1033\langfe1033\cgrid\langnp1033\langfenp1033 {\insrsid1792631 \chftnsepc 
+\par }}\sectd \linex0\endnhere\sectlinegrid360\sectdefaultcl\sftnbj {\*\pnseclvl1\pnucrm\pnstart1\pnindent720\pnhang {\pntxta .}}{\*\pnseclvl2\pnucltr\pnstart1\pnindent720\pnhang {\pntxta .}}{\*\pnseclvl3\pndec\pnstart1\pnindent720\pnhang {\pntxta .}}
+{\*\pnseclvl4\pnlcltr\pnstart1\pnindent720\pnhang {\pntxta )}}{\*\pnseclvl5\pndec\pnstart1\pnindent720\pnhang {\pntxtb (}{\pntxta )}}{\*\pnseclvl6\pnlcltr\pnstart1\pnindent720\pnhang {\pntxtb (}{\pntxta )}}{\*\pnseclvl7\pnlcrm\pnstart1\pnindent720\pnhang 
+{\pntxtb (}{\pntxta )}}{\*\pnseclvl8\pnlcltr\pnstart1\pnindent720\pnhang {\pntxtb (}{\pntxta )}}{\*\pnseclvl9\pnlcrm\pnstart1\pnindent720\pnhang {\pntxtb (}{\pntxta )}}\pard\plain 
+\s1\ql \li0\ri0\sb240\sa60\keepn\widctlpar\aspalpha\aspnum\faauto\outlinelevel0\adjustright\rin0\lin0\itap0\pararsid2294299 \b\f1\fs32\lang1033\langfe1033\kerning32\cgrid\langnp1033\langfenp1033 {\insrsid2294299 This is a test RTF
+\par }\pard\plain \ql \li0\ri0\widctlpar\aspalpha\aspnum\faauto\adjustright\rin0\lin0\itap0\pararsid2294299 \fs24\lang1033\langfe1033\cgrid\langnp1033\langfenp1033 {\insrsid2294299 Hi! I\rquote m a test file. This is some }{\b\insrsid2294299 bold}{
+\insrsid2294299  text, and some }{\i\insrsid2294299 italic}{\insrsid2294299  text, as well as some }{\ul\insrsid2294299 underline}{\insrsid2294299  text. And a bit of }{\v\insrsid2294299\charrsid2294299 hidden}{\insrsid2294299  text. So we\rquote 
+re going to end this paragraph here and go on to a nice little list:
+\par 
+\par {\listtext\pard\plain\f3\insrsid2294299 \loch\af3\dbch\af0\hich\f3 \'b7\tab}}\pard \ql \fi-360\li720\ri0\widctlpar\jclisttab\tx720\aspalpha\aspnum\faauto\ls1\adjustright\rin0\lin720\itap0\pararsid2294299 {\insrsid2294299 Item 1
+\par {\listtext\pard\plain\f3\insrsid2294299 \loch\af3\dbch\af0\hich\f3 \'b7\tab}Item 2
+\par {\listtext\pard\plain\f3\insrsid2294299 \loch\af3\dbch\af0\hich\f3 \'b7\tab}Item 3
+\par {\listtext\pard\plain\f3\insrsid2294299 \loch\af3\dbch\af0\hich\f3 \'b7\tab}Item 4
+\par }\pard \ql \li0\ri0\widctlpar\aspalpha\aspnum\faauto\adjustright\rin0\lin0\itap0\pararsid2294299 {\insrsid2294299 
+\par And now comes a fun table:
+\par 
+\par }\trowd \irow0\irowband0\ts15\trgaph108\trleft-108\trbrdrt\brdrs\brdrw10 \trbrdrl\brdrs\brdrw10 \trbrdrb\brdrs\brdrw10 \trbrdrr\brdrs\brdrw10 \trbrdrh\brdrs\brdrw10 \trbrdrv\brdrs\brdrw10 
+\trftsWidth1\trftsWidthB3\trautofit1\trpaddl108\trpaddr108\trpaddfl3\trpaddft3\trpaddfb3\trpaddfr3\tbllkhdrrows\tbllklastrow\tbllkhdrcols\tbllklastcol \clvertalt\clbrdrt\brdrs\brdrw10 \clbrdrl\brdrs\brdrw10 \clbrdrb\brdrs\brdrw10 \clbrdrr\brdrs\brdrw10 
+\cltxlrtb\clftsWidth3\clwWidth2952\clshdrawnil \cellx2844\clvertalt\clbrdrt\brdrs\brdrw10 \clbrdrl\brdrs\brdrw10 \clbrdrb\brdrs\brdrw10 \clbrdrr\brdrs\brdrw10 \cltxlrtb\clftsWidth3\clwWidth2952\clshdrawnil \cellx5796\clvertalt\clbrdrt\brdrs\brdrw10 
+\clbrdrl\brdrs\brdrw10 \clbrdrb\brdrs\brdrw10 \clbrdrr\brdrs\brdrw10 \cltxlrtb\clftsWidth3\clwWidth2952\clshdrawnil \cellx8748\pard\plain \ql \li0\ri0\widctlpar\intbl\aspalpha\aspnum\faauto\adjustright\rin0\lin0\pararsid2294299\yts15 
+\fs24\lang1033\langfe1033\cgrid\langnp1033\langfenp1033 {\insrsid2294299 Cell 1\cell Cell 2
+\par More in cell 2\cell Cell 3\cell }\pard\plain \ql \li0\ri0\widctlpar\intbl\aspalpha\aspnum\faauto\adjustright\rin0\lin0 \fs24\lang1033\langfe1033\cgrid\langnp1033\langfenp1033 {\insrsid2294299 \trowd \irow0\irowband0\ts15\trgaph108\trleft-108\trbrdrt
+\brdrs\brdrw10 \trbrdrl\brdrs\brdrw10 \trbrdrb\brdrs\brdrw10 \trbrdrr\brdrs\brdrw10 \trbrdrh\brdrs\brdrw10 \trbrdrv\brdrs\brdrw10 
+\trftsWidth1\trftsWidthB3\trautofit1\trpaddl108\trpaddr108\trpaddfl3\trpaddft3\trpaddfb3\trpaddfr3\tbllkhdrrows\tbllklastrow\tbllkhdrcols\tbllklastcol \clvertalt\clbrdrt\brdrs\brdrw10 \clbrdrl\brdrs\brdrw10 \clbrdrb\brdrs\brdrw10 \clbrdrr\brdrs\brdrw10 
+\cltxlrtb\clftsWidth3\clwWidth2952\clshdrawnil \cellx2844\clvertalt\clbrdrt\brdrs\brdrw10 \clbrdrl\brdrs\brdrw10 \clbrdrb\brdrs\brdrw10 \clbrdrr\brdrs\brdrw10 \cltxlrtb\clftsWidth3\clwWidth2952\clshdrawnil \cellx5796\clvertalt\clbrdrt\brdrs\brdrw10 
+\clbrdrl\brdrs\brdrw10 \clbrdrb\brdrs\brdrw10 \clbrdrr\brdrs\brdrw10 \cltxlrtb\clftsWidth3\clwWidth2952\clshdrawnil \cellx8748\row }\pard\plain \ql \li0\ri0\widctlpar\intbl\aspalpha\aspnum\faauto\adjustright\rin0\lin0\pararsid2294299\yts15 
+\fs24\lang1033\langfe1033\cgrid\langnp1033\langfenp1033 {\insrsid2294299 Next row\cell Next row \cell Next row\cell }\pard\plain \ql \li0\ri0\widctlpar\intbl\aspalpha\aspnum\faauto\adjustright\rin0\lin0 
+\fs24\lang1033\langfe1033\cgrid\langnp1033\langfenp1033 {\insrsid2294299 \trowd \irow1\irowband1\lastrow \ts15\trgaph108\trleft-108\trbrdrt\brdrs\brdrw10 \trbrdrl\brdrs\brdrw10 \trbrdrb\brdrs\brdrw10 \trbrdrr\brdrs\brdrw10 \trbrdrh\brdrs\brdrw10 \trbrdrv
+\brdrs\brdrw10 \trftsWidth1\trftsWidthB3\trautofit1\trpaddl108\trpaddr108\trpaddfl3\trpaddft3\trpaddfb3\trpaddfr3\tbllkhdrrows\tbllklastrow\tbllkhdrcols\tbllklastcol \clvertalt\clbrdrt\brdrs\brdrw10 \clbrdrl\brdrs\brdrw10 \clbrdrb\brdrs\brdrw10 \clbrdrr
+\brdrs\brdrw10 \cltxlrtb\clftsWidth3\clwWidth2952\clshdrawnil \cellx2844\clvertalt\clbrdrt\brdrs\brdrw10 \clbrdrl\brdrs\brdrw10 \clbrdrb\brdrs\brdrw10 \clbrdrr\brdrs\brdrw10 \cltxlrtb\clftsWidth3\clwWidth2952\clshdrawnil \cellx5796\clvertalt\clbrdrt
+\brdrs\brdrw10 \clbrdrl\brdrs\brdrw10 \clbrdrb\brdrs\brdrw10 \clbrdrr\brdrs\brdrw10 \cltxlrtb\clftsWidth3\clwWidth2952\clshdrawnil \cellx8748\row }\pard \ql \li0\ri0\widctlpar\aspalpha\aspnum\faauto\adjustright\rin0\lin0\itap0\pararsid2294299 {
+\insrsid2294299 
+\par A page break:
+\par \page And here we\rquote re on the next page.}{\insrsid1792631  }{\insrsid2294299 
+\par }{\insrsid1792631 This para has a }{\cs17\super\insrsid1792631 \chftn {\footnote \pard\plain \s16\ql \li0\ri0\widctlpar\aspalpha\aspnum\faauto\adjustright\rin0\lin0\itap0 \fs20\lang1033\langfe1033\cgrid\langnp1033\langfenp1033 {\cs17\super\insrsid1792631 
+\chftn }{\insrsid1792631  This is the actual content of the footnote.}}}{\insrsid1792631 footnote.
+\par And here\rquote s yet another paragraph. }{\insrsid1792631\charrsid2294299 
+\par }}

data/spec/units/config_spec.rb

@@ -0,0 +1,25 @@
+require 'spec_helper'
+
+describe "the configuration" do
+  subject {Hydra::Derivatives }
+
+  it "should have some configuration defaults" do
+    subject.ffmpeg_path.should == 'ffmpeg'
+    subject.enable_ffmpeg.should be_true
+    subject.libreoffice_path.should == 'soffice'
+    subject.temp_file_base.should == '/tmp'
+    subject.fits_path.should == 'fits.sh'
+  end
+
+  it "should let you change the configuration" do
+    subject.ffmpeg_path = '/usr/local/ffmpeg-1.0/bin/ffmpeg'
+    subject.ffmpeg_path.should == '/usr/local/ffmpeg-1.0/bin/ffmpeg'
+  end
+
+  it "should let you reset the configuration" do
+    subject.ffmpeg_path = '/usr/local/ffmpeg-1.0/bin/ffmpeg'
+    subject.reset_config!
+    subject.ffmpeg_path.should == 'ffmpeg'
+  end
+
+end

data/spec/units/transcoding_spec.rb

@@ -14,25 +14,34 @@ describe "Transcoder" do
       delegate :mime_type, :to => :characterization, :unique => true
       has_file_datastream 'content', type: ContentDatastream
 
-      makes_derivatives_of :content, when: :mime_type, is_one_of: 'application/pdf',
-            derivatives: { :thumb => "100x100>" }
-
-      makes_derivatives_of :content, when: :mime_type, is_one_of: 'audio/wav',
-            derivatives: { :mp3 => {format: 'mp3'}, :ogg => {format: 'ogg'} }, processors: :audio
-
-      # -g 30 enforces keyframe generation every second (30fps)
-      # -b:v is the video bitrate
-      # -acodec is the audio codec
-      size_attributes = "-s 320x240"
-      audio_attributes = "-ac 2 -ab 96k -ar 44100"
-      makes_derivatives_of :content, when: :mime_type, is: 'video/avi',
-            derivatives: { :mp4 => {format: 'mp4'}, :webm => {format: 'webm'} }, processors: :video
-
-      makes_derivatives_of :content, when: :mime_type, is_one_of: ['image/png', 'image/jpg'],
-             derivatives: { :medium => "300x300>", :thumb => "100x100>" }
+      makes_derivatives do |obj| 
+        case obj.mime_type
+        when 'application/pdf'
+          obj.transform_datastream :content, { :thumb => "100x100>" }
+        when 'audio/wav'
+          obj.transform_datastream :content, { :mp3 => {format: 'mp3'}, :ogg => {format: 'ogg'} }, processor: :audio
+        when 'video/avi'
+          obj.transform_datastream :content, { :mp4 => {format: 'mp4'}, :webm => {format: 'webm'} }, processor: :video
+        when 'image/png', 'image/jpg'
+          obj.transform_datastream :content, { :medium => "300x300>", :thumb => "100x100>" }
+        when 'application/vnd.ms-powerpoint'
+          obj.transform_datastream :content, { :access => { :format=>'pdf' } }, processor: 'document'
+        when 'text/rtf'
+          obj.transform_datastream :content, { :preservation=> {:format => 'odf' }, :access => { :format=>'pdf' } }, processor: 'document'
+        end
+      end
+      
+      makes_derivatives :generate_special_derivatives
       
+      def generate_special_derivatives
+        if label == "special" && mime_type == 'image/png'
+          transform_datastream :content, { :medium => {size: "200x300>", datastream: 'special_ds'} }
+        end
+      end
+
     end
   end
+  
   describe "with an attached image" do
     let(:attachment) { File.open(File.expand_path('../../fixtures/world.png', __FILE__))}
     let(:file) { GenericFile.new(mime_type: 'image/png').tap { |t| t.content.content = attachment; t.save } }
@@ -85,4 +94,41 @@ describe "Transcoder" do
       file.datastreams['content_webm'].mimeType.should == 'video/webm'
     end
   end
+  
+  describe "using callback methods" do
+    let(:attachment) { File.open(File.expand_path('../../fixtures/world.png', __FILE__))}
+    let(:file) { GenericFile.new(mime_type: 'image/png', label: "special").tap { |t| t.content.content = attachment; t.save } }
+
+    it "should transcode" do
+      file.datastreams.key?('special_ds').should be_false
+      file.create_derivatives
+      file.datastreams['special_ds'].should have_content      
+      file.datastreams['special_ds'].mimeType.should == 'image/png'
+      file.datastreams['special_ds'].should have_content 
+    end
+  end
+
+  describe "with an attached Powerpoint", unless: ENV['TRAVIS'] == 'true' do
+    let(:attachment) { File.open(File.expand_path('../../fixtures/FlashPix.ppt', __FILE__))}
+    let(:file) { GenericFile.new(mime_type: 'application/vnd.ms-powerpoint').tap { |t| t.content.content = attachment; t.save } }
+
+    it "should transcode" do
+      file.create_derivatives
+      file.datastreams['content_access'].should have_content
+      file.datastreams['content_access'].mimeType.should == 'application/pdf'
+    end
+  end
+
+  describe "with an attached rich text format", unless: ENV['TRAVIS'] == 'true' do
+    let(:attachment) { File.open(File.expand_path('../../fixtures/sample.rtf', __FILE__))}
+    let(:file) { GenericFile.new(mime_type: 'text/rtf').tap { |t| t.content.content = attachment; t.save } }
+
+    it "should transcode" do
+      file.create_derivatives
+      file.datastreams['content_access'].should have_content
+      file.datastreams['content_access'].mimeType.should == 'application/pdf'
+      file.datastreams['content_preservation'].should have_content
+      file.datastreams['content_preservation'].mimeType.should == 'application/vnd.oasis.opendocument.text'
+    end
+  end
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: hydra-derivatives
 version: !ruby/object:Gem::Version
-  version: 0.0.2
+  version: 0.0.3
 platform: ruby
 authors:
 - Justin Coyne
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2013-07-24 00:00:00.000000000 Z
+date: 2013-07-25 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: bundler
@@ -122,6 +122,7 @@ extensions: []
 extra_rdoc_files: []
 files:
 - .gitignore
+- .rspec
 - .travis.yml
 - CONTRIBUTING.md
 - Gemfile
@@ -134,16 +135,22 @@ files:
 - hydra-derivatives.gemspec
 - lib/hydra/derivatives.rb
 - lib/hydra/derivatives/audio.rb
+- lib/hydra/derivatives/config.rb
+- lib/hydra/derivatives/document.rb
 - lib/hydra/derivatives/extract_metadata.rb
 - lib/hydra/derivatives/ffmpeg.rb
 - lib/hydra/derivatives/image.rb
 - lib/hydra/derivatives/processor.rb
+- lib/hydra/derivatives/shell_based_processor.rb
 - lib/hydra/derivatives/video.rb
+- spec/fixtures/FlashPix.ppt
 - spec/fixtures/countdown.avi
 - spec/fixtures/piano_note.wav
+- spec/fixtures/sample.rtf
 - spec/fixtures/test.pdf
 - spec/fixtures/world.png
 - spec/spec_helper.rb
+- spec/units/config_spec.rb
 - spec/units/image_spec.rb
 - spec/units/transcoding_spec.rb
 - spec/units/video_spec.rb
@@ -172,11 +179,14 @@ signing_key:
 specification_version: 4
 summary: Derivative generation plugin for hydra
 test_files:
+- spec/fixtures/FlashPix.ppt
 - spec/fixtures/countdown.avi
 - spec/fixtures/piano_note.wav
+- spec/fixtures/sample.rtf
 - spec/fixtures/test.pdf
 - spec/fixtures/world.png
 - spec/spec_helper.rb
+- spec/units/config_spec.rb
 - spec/units/image_spec.rb
 - spec/units/transcoding_spec.rb
 - spec/units/video_spec.rb