paperclip-document 0.0.2 → 0.0.5

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: e34d0f58cb3b703bf031a5bd79ea8e44ec7c1a54
-  data.tar.gz: 71f1f58a05a92bb9b9356e7ccd9ffc5756749ed3
+  metadata.gz: 0eea80135b657f2b47ecc5948d4443be118cec19
+  data.tar.gz: f9b90dc80e759fb9809388ef49e1d0f1bfb30d3d
 SHA512:
-  metadata.gz: b025b1102d126c9de8996907fcff766727a37f9883c1517faad68ca95fb4ebc5fe5017995ee354934d182028bb942d11a9a40dc3da741d044104a05372ad7f72
-  data.tar.gz: 9461dd9fc91488397191adf30611e28a7d3d442f60b328d064a884aa040cadbcc5d9ba6f8279ed7b740ea2fb9fe836dd2e3eb6ff0e1dbde6e41fc16788ef24a9
+  metadata.gz: ede250c619b01cfa830a5b99a48d7c6153173244134782a1b53902b1f3d8b9e6f00c0ab036d5ccc660db4b7d6a5e5ec025afb27f8c5e642bae44cfedd2e0118a
+  data.tar.gz: f0e8f980c2eca47d43d19334f2119a0f6c211ed4df1e550c43086a5540238a5adb106c9091f1fcf54aa3a4d7be2111c3e9e75653292a166e173cefc40951bd49

data/lib/paperclip/document.rb

@@ -1,46 +1,20 @@
 require "paperclip/document/version"
+require "paperclip/document/attachment_extension"
 require "docsplit"
 require "pathname"
 
 module Paperclip
-  # Main processor
-  class DocumentProcessor < Processor
-    
-    attr_reader :instance, :tmp_dir
-
-    def initialize(file, options = {}, attachment = nil)
-      super(file, options, attachment)
-      @instance = @attachment.instance
-      # @tmp_dir = Pathname.new(Dir.tmpdir).join("paperclip-document", instance.class.name, attachment.name.to_s, basename, Time.now.to_i.to_s(36) + "-" + rand(1_000_000).to_s(36))
-      # @tmp_dir = Rails.root.join("paperclip-document", instance.class.name, attachment.name.to_s, basename, Time.now.to_i.to_s(36) + "-" + rand(1_000_000).to_s(36))
-      @tmp_dir = Pathname.new(Dir.tmpdir).join("paperclip-document-" + rand(1_000_000_000).to_s(36) + "-" + rand(1_000_000).to_s(36) + "-" + Time.now.to_i.to_s(36) + "-" + rand(1_000_000).to_s(36))
-    end
-    
-    def file_path
-      Pathname.new(@file.path)
-    end
-
-    def basename
-      file_path.basename.to_s.gsub(/\.[^\.]+/, '')
-    end
-    
+  module Document
+    autoload :Processor,  'paperclip/document/processor'
+    autoload :Processors, 'paperclip/document/processors'
   end
 
-  class Attachment
-
-    # Returns the content_text of the file as originally extracted, and lives in the <attachment>_content_text attribute of the model.
-    def content_text
-      instance_read(:content_text)
-    end
-
-    # Returns the pages_count of the file as originally computed, and lives in the <attachment>_pages_count attribute of the model.
-    def pages_count
-      instance_read(:pages_count)
-    end
-
+  configure do |c|
+    c.register_processor :sketcher, Document::Processors::Sketcher
+    c.register_processor :reader,   Document::Processors::Reader
+    c.register_processor :freezer,  Document::Processors::Freezer
+    c.register_processor :counter,  Document::Processors::Counter
   end
-  
 
 end
 
-require "paperclip/document/processors"

data/lib/paperclip/document/attachment_extension.rb

@@ -0,0 +1,17 @@
+module Paperclip
+
+  class Attachment
+    
+    # Returns the content_text of the file as originally extracted, and lives in the <attachment>_content_text attribute of the model.
+    def content_text
+      instance_read(:content_text)
+    end
+    
+    # Returns the pages_count of the file as originally computed, and lives in the <attachment>_pages_count attribute of the model.
+    def pages_count
+      instance_read(:pages_count)
+    end
+    
+  end
+  
+end

data/lib/paperclip/document/processor.rb

@@ -0,0 +1,25 @@
+module Paperclip
+  module Document
+    
+    # Main processor
+    class Processor < Paperclip::Processor
+      
+      attr_reader :instance, :tmp_dir
+      
+      def initialize(file, options = {}, attachment = nil)
+        super(file, options, attachment)
+        @instance = @attachment.instance
+        @tmp_dir = Pathname.new(Dir.tmpdir).join("paperclip-document-" + Time.now.to_i.to_s(36) + rand(1_000_000_000).to_s(36))
+      end
+      
+      def file_path
+        Pathname.new(@file.path)
+      end
+
+      def basename
+        file_path.basename.to_s.gsub(/\.[^\.]+/, '')
+      end
+      
+    end
+  end
+end

data/lib/paperclip/document/processors.rb

@@ -1,4 +1,11 @@
-require "paperclip/document/processors/sketcher"
-require "paperclip/document/processors/reader"
-require "paperclip/document/processors/freezer"
-require "paperclip/document/processors/counter"
+module Paperclip
+  module Document
+    module Processors
+      autoload :Sketcher, "paperclip/document/processors/sketcher"
+      autoload :Reader,   "paperclip/document/processors/reader"
+      autoload :Freezer,  "paperclip/document/processors/freezer"
+      autoload :Counter,  "paperclip/document/processors/counter"
+    end
+  end
+end
+

data/lib/paperclip/document/processors/counter.rb

@@ -1,45 +1,49 @@
 module Paperclip
+  module Document
+    module Processors
+
+      # This processor extract the OCR text of the file
+      class Counter < Paperclip::Document::Processor
+
+        attr_accessor :pages_count_column
+
+        def initialize(file, options = {}, attachment = nil)
+          super(file, options, attachment)
+          if @options[:pages_count_column].nil? and pages_count_column?
+            @options[:pages_count_column] = default_pages_count_column
+          end
+          @pages_count_column = @options[:pages_count_column]
+
+          unless @pages_count_column
+            raise Paperclip::Error, "No pages count column given"
+          end
+        end
+
+        # Extract the pages count of all the document
+        def make
+          count = Docsplit.extract_length(file_path.to_s)
+
+          instance[pages_count_column] = count
+          instance.run_callbacks(:save) { false }
+
+          return file
+        end
+        
+        # Check if a pages count column is present
+        def pages_count_column?
+          expected_column = default_pages_count_column
+          return @attachment.instance.class.columns.detect do |column|
+            column.name.to_s == expected_column
+          end
+        end
+
+        # Returns the name of the default pages count column
+        def default_pages_count_column
+          @attachment.name.to_s + "_pages_count"
+        end
 
-  # This processor extract the OCR text of the file
-  class Counter < DocumentProcessor
-
-    attr_accessor :pages_count_column
-
-    def initialize(file, options = {}, attachment = nil)
-      super(file, options, attachment)
-      if @options[:pages_count_column].nil? and pages_count_column?
-        @options[:pages_count_column] = default_pages_count_column
-      end
-      @pages_count_column = @options[:pages_count_column]
-
-      unless @pages_count_column
-        raise Paperclip::Error, "No pages count column given"
       end
-    end
-
-    # Extract the pages count of all the document
-    def make
-      count = Docsplit.extract_length(file_path.to_s)
-
-      instance[pages_count_column] = count
-      instance.run_callbacks(:save) { false }
 
-      return file
     end
-    
-    # Check if a pages count column is present
-    def pages_count_column?
-      expected_column = default_pages_count_column
-      return @attachment.instance.class.columns.detect do |column|
-        column.name.to_s == expected_column
-      end
-    end
-
-    # Returns the name of the default pages count column
-    def default_pages_count_column
-      @attachment.name.to_s + "_pages_count"
-    end
-
   end
-
 end

data/lib/paperclip/document/processors/freezer.rb

@@ -1,38 +1,36 @@
-require 'filemagic'
-
 module Paperclip
+  module Document
+    module Processors
+
+      # This processor converts document to PDF
+      class Freezer < Paperclip::Document::Processor
+        def initialize(file, options = {}, attachment = nil)
+          super
+          @format = options[:format]
+          unless @format == :pdf
+            raise Paperclip::Error, "Valid format (pdf) must be specified"
+          end
+        end
+
+        # Convert the document to pdf
+        def make
+          destination_path = tmp_dir.to_s
+          destination_file = File.join(destination_path, basename + ".#{@format}")
+          if pdf_format?
+            destination_file = file_path.to_s
+          else
+            Docsplit.extract_pdf(file_path.to_s, :output => destination_path)
+          end
+          return File.open(destination_file)
+        end
+
+
+        def pdf_format?
+          File.open(file_path, "rb", &:readline) =~ /\A\%PDF-\d+(\.\d+)?$/
+        end
 
-  # This processor extract first page as thumbnail
-  class Freezer < DocumentProcessor
-    def initialize(file, options = {}, attachment = nil)
-      super
-      @format = options[:format]
-      unless @format == :pdf
-        raise Paperclip::Error, "Valid format (pdf) must be specified"
       end
-    end
 
-    # Convert the document to pdf
-    def make
-      destination_path = tmp_dir.to_s
-      destination_file = File.join(destination_path, basename + ".#{@format}")
-      if pdf_format?
-        destination_file = file_path.to_s
-      else
-        Docsplit.extract_pdf(file_path.to_s, :output => destination_path)
-      end
-      return File.open(destination_file)
     end
-
-
-    def pdf_format?
-      file_magic = FileMagic.new
-      type = file_magic.file(file_path.to_s)
-      file_magic.close
-      type =~ /pdf/i
-    end
-
   end
-
-
 end

data/lib/paperclip/document/processors/reader.rb

@@ -1,53 +1,57 @@
 module Paperclip
+  module Document
+    module Processors
+
+      # This processor extract the OCR text of the file
+      class Reader < Paperclip::Document::Processor
+
+        attr_accessor :clean, :text_column, :language
+
+        def initialize(file, options = {}, attachment = nil)
+          super(file, options, attachment)
+          if @options[:text_column].nil? and text_column?
+            @options[:text_column] = default_text_column
+          end
+          @language = @options[:language]
+          @text_column = @options[:text_column]
+          unless @text_column
+            raise Paperclip::Error, "No content text column given"
+          end
+          @clean = (RUBY_VERSION >= "2.0" ? false : options.has_key?(:clean) ? !!options[:clean] : true)
+        end
+
+        # Extract the text of all the document
+        def make
+          destination_path = tmp_dir.to_s
+          options = {output: destination_path, clean: @clean}
+          options[:language] = (language.is_a?(Proc) ? language.call(attachment.instance) : language)
+          Docsplit.extract_text(file_path.to_s, options)
+          
+          destination_file = File.join(destination_path, basename + ".txt")
+          instance = @attachment.instance
+          f = File.open(destination_file)
+          instance[text_column] = f.read
+          instance.run_callbacks(:save) { false }
+          f.close
+
+          return file
+        end
+        
+        # Check if the default text column is present
+        def text_column?
+          expected_column = default_text_column
+          return instance.class.columns.detect do |column|
+            column.name.to_s == expected_column
+          end
+        end
+
+        # Returns the name of the default text column
+        def default_text_column
+          @attachment.name.to_s + "_content_text"
+        end
 
-  # This processor extract the OCR text of the file
-  class Reader < DocumentProcessor
-
-    attr_accessor :clean, :text_column, :language
-
-    def initialize(file, options = {}, attachment = nil)
-      super(file, options, attachment)
-      if @options[:text_column].nil? and text_column?
-        @options[:text_column] = default_text_column
-      end
-      @language = @options[:language]
-      @text_column = @options[:text_column]
-      unless @text_column
-        raise Paperclip::Error, "No content text column given"
-      end
-      @clean = (RUBY_VERSION >= "2.0" ? false : options.has_key?(:clean) ? !!options[:clean] : true)
-    end
-
-    # Extract the text of all the document
-    def make
-      destination_path = tmp_dir.to_s
-      options = {output: destination_path, clean: @clean}
-      options[:language] = (language.is_a?(Proc) ? language.call(attachment.instance) : language)
-      Docsplit.extract_text(file_path.to_s, options)
-      
-      destination_file = File.join(destination_path, basename + ".txt")
-      instance = @attachment.instance
-      f = File.open(destination_file)
-      instance[text_column] = f.read
-      instance.run_callbacks(:save) { false }
-      f.close
-
-      return file
-    end
-    
-    # Check if the default text column is present
-    def text_column?
-      expected_column = default_text_column
-      return instance.class.columns.detect do |column|
-        column.name.to_s == expected_column
       end
-    end
 
-    # Returns the name of the default text column
-    def default_text_column
-      @attachment.name.to_s + "_content_text"
     end
-
   end
-
 end

data/lib/paperclip/document/processors/sketcher.rb

@@ -1,39 +1,43 @@
 module Paperclip
+  module Document
+    module Processors
 
-  # This processor extract first page as thumbnail
-  class Sketcher < DocumentProcessor
+      # This processor extract first page as thumbnail
+      class Sketcher < Paperclip::Document::Processor
 
-    attr_accessor :format, :density, :format
+        attr_accessor :format, :density, :format
 
-    def initialize(file, options = {}, attachment = nil)
-      super(file, options, attachment)
-      @format  = (options[:format] || :jpg).to_sym
-      unless [:jpg, :png].include?(@format)
-        raise Paperclip::Error, "Valid format must be specified"
-      end
-      unless @size = options[:size]
-        @density = (options[:density] || 150).to_f
-      end
-    end
+        def initialize(file, options = {}, attachment = nil)
+          super(file, options, attachment)
+          @format  = (options[:format] || :jpg).to_sym
+          unless [:jpg, :png].include?(@format)
+            raise Paperclip::Error, "Valid format must be specified"
+          end
+          unless @size = options[:size]
+            @density = (options[:density] || 150).to_f
+          end
+        end
 
-    # Extract the page
-    def make
-      destination_path = tmp_dir.to_s
-      options = {:output => destination_path, :pages => [1], :format => [@format]}
-      if @size
-        options[:size] = @size
-      elsif @density
-        options[:density] = @density
-      end
-      Docsplit.extract_images(file_path.to_s, options)
-      begin
-      rescue
-        raise Paperclip::Error, "There was an error extracting the first thumbnail from #{basename}"
-      end
-      return File.open(File.join(destination_path, basename + "_1.#{@format}"))
-    end
+        # Extract the page
+        def make
+          destination_path = tmp_dir.to_s
+          options = {:output => destination_path, :pages => [1], :format => [@format]}
+          if @size
+            options[:size] = @size
+          elsif @density
+            options[:density] = @density
+          end
+          Docsplit.extract_images(file_path.to_s, options)
+          begin
+          rescue
+            raise Paperclip::Error, "There was an error extracting the first thumbnail from #{basename}"
+          end
+          return File.open(File.join(destination_path, basename + "_1.#{@format}"))
+        end
 
-  end
+      end
 
 
+    end
+  end
 end

data/lib/paperclip/document/version.rb

@@ -1,5 +1,5 @@
 module Paperclip
   module Document
-    VERSION = "0.0.2"
+    VERSION = "0.0.5"
   end
 end

metadata

@@ -1,73 +1,99 @@
 --- !ruby/object:Gem::Specification
 name: paperclip-document
 version: !ruby/object:Gem::Version
-  version: 0.0.2
+  version: 0.0.5
 platform: ruby
 authors:
 - Brice Texier
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2014-02-07 00:00:00.000000000 Z
+date: 2014-03-28 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: paperclip
   requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - ~>
+    - - '>='
       - !ruby/object:Gem::Version
         version: '3.1'
+    - - <
+      - !ruby/object:Gem::Version
+        version: '5'
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - ~>
+    - - '>='
       - !ruby/object:Gem::Version
         version: '3.1'
+    - - <
+      - !ruby/object:Gem::Version
+        version: '5'
 - !ruby/object:Gem::Dependency
-  name: docsplit
+  name: burisu-docsplit
   requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - ~>
+    - - '>='
       - !ruby/object:Gem::Version
-        version: 0.7.2
+        version: 0.7.6
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - ~>
+    - - '>='
       - !ruby/object:Gem::Version
-        version: 0.7.2
+        version: 0.7.6
 - !ruby/object:Gem::Dependency
-  name: ruby-filemagic
+  name: bundler
   requirement: !ruby/object:Gem::Requirement
     requirements:
     - - ~>
       - !ruby/object:Gem::Version
-        version: 0.4.2
-  type: :runtime
+        version: '1.3'
+  type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - ~>
       - !ruby/object:Gem::Version
-        version: 0.4.2
+        version: '1.3'
 - !ruby/object:Gem::Dependency
-  name: bundler
+  name: rake
   requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - ~>
+    - - '>='
       - !ruby/object:Gem::Version
-        version: '1.3'
+        version: '0'
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - ~>
+    - - '>='
       - !ruby/object:Gem::Version
-        version: '1.3'
+        version: '0'
 - !ruby/object:Gem::Dependency
-  name: rake
+  name: rails
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - '>='
+      - !ruby/object:Gem::Version
+        version: '3.2'
+    - - <
+      - !ruby/object:Gem::Version
+        version: '5'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - '>='
+      - !ruby/object:Gem::Version
+        version: '3.2'
+    - - <
+      - !ruby/object:Gem::Version
+        version: '5'
+- !ruby/object:Gem::Dependency
+  name: coveralls
   requirement: !ruby/object:Gem::Requirement
     requirements:
     - - '>='
@@ -118,6 +144,8 @@ files:
 - LICENSE.txt
 - README.md
 - lib/paperclip/document.rb
+- lib/paperclip/document/attachment_extension.rb
+- lib/paperclip/document/processor.rb
 - lib/paperclip/document/processors.rb
 - lib/paperclip/document/processors/counter.rb
 - lib/paperclip/document/processors/freezer.rb