Class: Biblionet::Extractors::PublisherExtractor

Inherits:
Base
  • Object
show all
Defined in:
lib/bookshark/extractors/publisher_extractor.rb

Instance Attribute Summary collapse

Attributes inherited from Base

#biblionet_id, #filepath, #page, #url

Instance Method Summary collapse

Methods inherited from Base

#decode_text, decode_text, #load_page, #load_page_from_file, #load_page_from_url, #present?, #save_page

Methods included from FileManager

#list_directories, #list_files, #save_to

Constructor Details

#initialize(uri = nil) ⇒ PublisherExtractor

Returns a new instance of PublisherExtractor.



12
13
14
15
 
# File 'lib/bookshark/extractors/publisher_extractor.rb', line 12

def initialize(uri=nil)
  super(uri)        
  extract_publisher unless uri.nil? or @page.nil?        
end

Instance Attribute Details

#publisher ⇒ Object (readonly)

Returns the value of attribute publisher.



10
11
12
 
# File 'lib/bookshark/extractors/publisher_extractor.rb', line 10

def publisher
  @publisher
end

Instance Method Details

#extract_publisher(biblionet_id = @biblionet_id, publisher_page = @page) ⇒ Object 



23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
 
# File 'lib/bookshark/extractors/publisher_extractor.rb', line 23

def extract_publisher(biblionet_id=@biblionet_id, publisher_page=@page)
  puts "Extracting publisher: #{biblionet_id}"
  page = PublisherDataExtractor.new(publisher_page)
  
  return nil if page.nodeset.nil?
          
  headquarters                    = page.headquarters
  bookstores                      = page.bookstores
  bookstores['Έδρα']              = headquarters unless headquarters.all? {|k,v| v.nil? or v.empty?}

  publisher_hash = {}
  publisher_hash[:name]          = page.name
  publisher_hash[:owner]         = page.owner       
  publisher_hash[:bookstores]    = bookstores
  publisher_hash[:b_id]          = biblionet_id

  return @publisher = publisher_hash
end

#load_and_extract_publisher(uri = nil) ⇒ Object 



18
19
20
21
 
# File 'lib/bookshark/extractors/publisher_extractor.rb', line 18

def load_and_extract_publisher(uri=nil)
  load_page(uri)
  extract_publisher unless uri.nil? or @page.nil?
end