Class: Paperless::Engine

Inherits:
Object
  • Object
show all
Defined in:
lib/paperless/engine.rb

Constant Summary collapse

PDFPEN_ENGINE = 
'pdfpen'
PDFPENPRO_ENGINE = 
'pdfpenpro'
PDFPENPRO6_ENGINE = 
'pdfpenpro6'
ACROBAT_ENGINE = 
'acrobat'
DEVONTHINKPRO_ENGINE = 
'devonthinkpro'
DEVONTHINKPRO_SERVICE = 
'devonthinkpro'
FINDER_SERVICE = 
'finder'
EVERNOTE_SERVICE = 
'evernote'

Instance Attribute Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize(options) ⇒ Engine

Returns a new instance of Engine.



28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
 
# File 'lib/paperless/engine.rb', line 28

def initialize(options)
    @destination         = nil
    @service             = options[:default_service]
    @title               = nil
    @date                = DateTime.now
    @tags                = Array.new()
    
    @file                = options[:file]
    @text_ext            = options[:text_ext]
    @pdf_ext             = [Paperless::PDF_EXT]
    @default_service     = options[:default_service]
    @date_format         = options[:date_format]
    @date_locale         = options[:date_locale]
    @date_default        = options[:date_default]
    @default_destination = options[:default_destination]
    @rules               = Array.new()

	options[:rules].each do |rule|
		@rules.push Paperless::Rule.new(rule)
	end

    @ocr_engine = options[:ocr_engine]||false
end

Instance Attribute Details

#service ⇒ Object (readonly)

Returns the value of attribute service.



26
27
28
 
# File 'lib/paperless/engine.rb', line 26

def service
  @service
end

Instance Method Details

#add_tags(tags) ⇒ Object 



70
71
72
73
74
75
 
# File 'lib/paperless/engine.rb', line 70

def add_tags(tags)
	if tags.length > 0
		@tags = (@tags + tags).collect {|x| x = x.downcase }
		@tags.uniq!
	end
end

#create(options) ⇒ Object 



197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
 
# File 'lib/paperless/engine.rb', line 197

def create(options)
    # May need to externalize this so other methods can access it.
    service = case @service.nil? ? @default_service : @service
      when /^#{EVERNOTE_SERVICE}$/i then PaperlessService::Evernote.new
      when /^#{FINDER_SERVICE}$/i then PaperlessService::Finder.new
      when /^#{DEVONTHINKPRO_SERVICE}$/i then PaperlessService::DevonThinkPro.new
      else false
    end

    if service
      self.print
      
      destination = @destination.nil? ? @default_destination                      : @destination
      title       = @title.nil?       ? File.basename(@file, File.extname(@file)) : @title

      service.create({ 
        :delete => options[:delete], 
        :destination => destination, 
        :text_ext => @text_ext, 
        :file => @file, 
        :date => @date, 
        :title => title, 
        :tags => @tags
      })
    else 
      puts "WARNING: No valid Service was defined."
    end
end

#ocr(dump = false) ⇒ Object 



158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
 
# File 'lib/paperless/engine.rb', line 158

def ocr(dump = false)
    reader = PDF::Reader.new(@file)
    if reader.pages.length > 0
      text = reader.pages[0].text
      if !text.nil? && text != ''
        puts text if dump
        puts "This doc already seems to be OCR'd. Not processing through #{@ocr_engine}"
        return
      end
    end

	puts "Running OCR on file with #{@ocr_engine}"
    ocr_engine = case @ocr_engine
      when /^#{PDFPENPRO6_ENGINE}$/i    then PaperlessOCR::PDFpenPro6.new
      when /^#{PDFPEN6_ENGINE}$/i       then PaperlessOCR::PDFpen6.new
      when /^#{PDFPENPRO_ENGINE}$/i     then PaperlessOCR::PDFpenPro.new
      when /^#{PDFPEN_ENGINE}$/i        then PaperlessOCR::PDFpen.new
      when /^#{ACROBAT_ENGINE}$/i       then PaperlessOCR::Acrobat.new
      when /^#{DEVONTHINKPRO_ENGINE}$/i then PaperlessOCR::DevonThinkPro.new
      else false
    end
    
    if ocr_engine
      ocr_engine.ocr({:file => @file})

      if dump
        puts "Dumping Page Content..."
        # Print the contents of the doc
        reader = PDF::Reader.new(@file)
        reader.pages.each do |page|
          puts page.text
        end
      end

    else
      puts "WARNING: No valid OCR engine was defined."
    end
end

#print ⇒ Object 



226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
 
# File 'lib/paperless/engine.rb', line 226

def print
    service = @service.nil? ? @default_service : @service
    title = @title.nil? ? File.basename(@file, File.extname(@file)) : @title

    destination = @destination.nil? ? @default_destination : @destination
    if destination == PaperlessService::Finder::NO_MOVE && service == PaperlessService::FINDER.downcase
      destination = File.dirname(@file)
    end

    puts "* ---------------------------------------------"
	puts "* File: #{@file}"
    puts "* Service: #{service}"
	puts "* Destination: #{destination}"
	puts "* Title: #{title}"
	puts "* Date: #{@date.strftime('%Y-%m-%d')}"
	puts "* Tags: #{@tags.join(', ')}"
    puts "* ---------------------------------------------"
end

#process_pdf ⇒ Object 



133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
 
# File 'lib/paperless/engine.rb', line 133

def process_pdf
	puts "Processing PDF pages..."

  reader = PDF::Reader.new(@file)

  # Verify that we need to search for date or just set to today
  # Need to prcess file for date in case the rules need to use it.
  # First check if there are actually any date rules
    @rules.each do |rule|
	  if rule.condition == Paperless::DATE_VAR
	    reader.pages.each do |page|
	    	break if @date = date_search(page.text,@date_locale)
	    end
        # Check for the date in the file name if not found in the content
        @date = date_search(@file,@date_locale) if @date.nil?
	    break
   	end
	end

	# Process each page and pass it through the rules engine
   reader.pages.each do |page|
      process_rules_engine(page.text)
   end
end

#process_rules ⇒ Object 



52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
 
# File 'lib/paperless/engine.rb', line 52

def process_rules
    markdown_ext = ['md','mmd']
	text_ext = @text_ext + markdown_ext
	file_ext = File.extname(@file).gsub(/\./,'')

	if file_ext == Paperless::PDF_EXT
      
      self.process_pdf

    elsif text_ext.index file_ext

      self.process_text
    
    else
    	puts "Unknown file type. No rules were processed."
    end
end

#process_rules_engine(text) ⇒ Object 



100
101
102
103
104
105
106
107
108
109
110
111
112
 
# File 'lib/paperless/engine.rb', line 100

def process_rules_engine(text)
  self.set_date_default if @date.nil?
  # Process each page and pass it through the rules engine
  @rules.each do |rule|
    rule.set_date(@date,@date_format)
    if !rule.matched && rule.match(@file, text)
      self.add_tags(rule.tags)
      self.set_destination(rule.destination)
      self.set_title(rule.title)
      self.set_service(rule.service)
    end
  end
end

#process_text ⇒ Object 



114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
 
# File 'lib/paperless/engine.rb', line 114

def process_text
  puts "Processing Text file..."

  text = File.open(@file, "rb") {|io| io.read}

  # Verify that we need to search for date or just set to today
  # Need to prcess file for date in case the rules need to use it.
  # First check if there are actually any date rules
  @rules.each do |rule|
    if rule.condition == Paperless::DATE_VAR
      @date = date_search(text,@date_locale) || date_search(@file,@date_locale)
    end
  end

  # Process each page and pass it through the rules engine
  process_rules_engine(text)
end

#set_date_default ⇒ Object 



89
90
91
92
93
94
95
96
97
98
 
# File 'lib/paperless/engine.rb', line 89

def set_date_default()
  puts "Using default date..."
  # Set the default date to the date of the file or else to now
  if @date_default == Paperless::FILEDATE
    t = File.stat(@file).mtime
    @date = Date.new(t.year,t.month,t.day) 
  else
    @date = DateTime.now
  end
end

#set_destination(destination) ⇒ Object 



77
78
79
 
# File 'lib/paperless/engine.rb', line 77

def set_destination(destination)
	@destination = destination if destination && @destination.nil?
end

#set_service(service) ⇒ Object 



85
86
87
 
# File 'lib/paperless/engine.rb', line 85

def set_service(service)
	@service = service if service && @service.nil?
end

#set_title(title) ⇒ Object 



81
82
83
 
# File 'lib/paperless/engine.rb', line 81

def set_title(title)
	@title = title if title && @title.nil?
end