Class: EpubBook::Book

Inherits:
Object
  • Object
show all
Defined in:
lib/epub_book/book.rb

Constant Summary collapse

UserAgent = 
"Mozilla/5.0 (Windows NT 6.1; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/27.0.1453.116 Safari/537.36"
Referer = 
"http://www.baidu.com/"
Reg = 
/<script.*?>.*?<\/script>/m

Instance Attribute Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize(index_url, des_url = nil) {|_self| ... } ⇒ Book

Returns a new instance of Book.

Yields:

  • (_self)

Yield Parameters:



29
30
31
32
33
34
35
36
37
38
39
40
41
42
 
# File 'lib/epub_book/book.rb', line 29

def initialize(index_url,des_url=nil )
  @index_url = index_url
  @des_url = des_url
  @user_agent = UserAgent
  @referer = Referer
  @folder_name = Base64.urlsafe_encode64(Digest::MD5.digest(@index_url))[0..-3]
  @creator = 'javy_liu'
  @title_css = '.wrapper h1.title1'
  @index_item_css = 'ul.list3>li>a'
  @cover = 'cover.jpg'
  @body_css = '.articlebody'
  @item_attr = "href"
  yield self if block_given?
end

Instance Attribute Details

#body_css ⇒ Object

Returns the value of attribute body_css.



24
25
26
 
# File 'lib/epub_book/book.rb', line 24

def body_css
  @body_css
end

#cover ⇒ Object

Returns the value of attribute cover.



24
25
26
 
# File 'lib/epub_book/book.rb', line 24

def cover
  @cover
end

#cover_css ⇒ Object

Returns the value of attribute cover_css.



24
25
26
 
# File 'lib/epub_book/book.rb', line 24

def cover_css
  @cover_css
end

#creator ⇒ Object

Returns the value of attribute creator.



24
25
26
 
# File 'lib/epub_book/book.rb', line 24

def creator
  @creator
end

#des_url ⇒ Object

Returns the value of attribute des_url.



24
25
26
 
# File 'lib/epub_book/book.rb', line 24

def des_url
  @des_url
end

#description_css ⇒ Object

Returns the value of attribute description_css.



24
25
26
 
# File 'lib/epub_book/book.rb', line 24

def description_css
  @description_css
end

#folder_name ⇒ Object

Returns the value of attribute folder_name.



24
25
26
 
# File 'lib/epub_book/book.rb', line 24

def folder_name
  @folder_name
end

#index_item_css ⇒ Object

Returns the value of attribute index_item_css.



24
25
26
 
# File 'lib/epub_book/book.rb', line 24

def index_item_css
  @index_item_css
end

#item_attr ⇒ Object

Returns the value of attribute item_attr.



24
25
26
 
# File 'lib/epub_book/book.rb', line 24

def item_attr
  @item_attr
end

#limit ⇒ Object

Returns the value of attribute limit.



24
25
26
 
# File 'lib/epub_book/book.rb', line 24

def limit
  @limit
end

#mail_to ⇒ Object

Returns the value of attribute mail_to.



24
25
26
 
# File 'lib/epub_book/book.rb', line 24

def mail_to
  @mail_to
end

#page_attr ⇒ Object

Returns the value of attribute page_attr.



24
25
26
 
# File 'lib/epub_book/book.rb', line 24

def page_attr
  @page_attr
end

#page_css ⇒ Object

Returns the value of attribute page_css.



24
25
26
 
# File 'lib/epub_book/book.rb', line 24

def page_css
  @page_css
end

#path ⇒ Object

Returns the value of attribute path.



24
25
26
 
# File 'lib/epub_book/book.rb', line 24

def path
  @path
end

#referer ⇒ Object

Returns the value of attribute referer.



24
25
26
 
# File 'lib/epub_book/book.rb', line 24

def referer
  @referer
end

#title_css ⇒ Object

Returns the value of attribute title_css.



24
25
26
 
# File 'lib/epub_book/book.rb', line 24

def title_css
  @title_css
end

#user_agent ⇒ Object

Returns the value of attribute user_agent.



24
25
26
 
# File 'lib/epub_book/book.rb', line 24

def user_agent
  @user_agent
end

Instance Method Details

#book ⇒ Object 



52
53
54
55
56
 
# File 'lib/epub_book/book.rb', line 52

def book
  return @book if @book
  Dir.mkdir(book_path) unless test(?d,book_path)
  @book = test(?s,File.join(book_path,'index.yml')) ? YAML.load(File.open(File.join(book_path,'index.yml'))) : {files: []}
end

#book_path ⇒ Object 



44
45
46
 
# File 'lib/epub_book/book.rb', line 44

def book_path
  @book_path ||= File.join((@path || `pwd`.strip), @folder_name)
end

#fetch_book ⇒ Object 



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
 
# File 'lib/epub_book/book.rb', line 160

def fetch_book
  #重新得到书目，如果不存在或重新索引的话
  fetch_index  if !test(?s,File.join(book_path,'index.yml'))
  EpubBook.logger.info "------Fetch book----------"
  book[:files].each_with_index do |item,index|
    break if limit && index >= limit

    content_path = File.join(book_path,item[:content])

    #如果文件存在且长度不为0则获取下一个
    next if test(?s,content_path)

    begin
      doc_file = Nokogiri::HTML(judge_encoding(HTTP.headers("User-Agent" => @user_agent,'Referer'=> @referer).get(item[:url]).to_s))

      File.open(content_path,'w') do |f|
        f.write("<h3>#{item[:label]}</h3>")
        f.write(doc_file.css(@body_css).to_s.gsub(Reg,''))
      end
    rescue  Exception => e
      EpubBook.logger.info "Error:#{e.message},#{item.inspect}"
      #EpubBook.logger.info e.backtrace
      next
    end
  end

end

#fetch_index(url = nil) ⇒ Object

得到书目索引



121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
 
# File 'lib/epub_book/book.rb', line 121

def fetch_index(url=nil)
  book[:files] = []
  url ||= @index_url
  doc = Nokogiri::HTML(judge_encoding(HTTP.headers("User-Agent" => @user_agent ,'Referer'=> @referer).get(URI.encode(url)).to_s))
  #generate index.yml

  if !book[:title]
    doc1 = if @des_url.nil?
             doc
           else
             Nokogiri::HTML(judge_encoding(HTTP.headers("User-Agent" => @user_agent ,'Referer'=> @referer).get(URI.encode(generate_abs_url(doc.css(@des_url).attr("href").to_s))).to_s))
           end
    get_des(doc1)
  end

  doc.css(@index_item_css).each do |item|
    _href = URI.encode(item.attr(@item_attr).to_s)
    next if _href.start_with?('javascript') || _href.start_with?('#')

    _href = generate_abs_url(_href)

    book[:files] << {label: item.text, url: _href}
  end

  #如果有分页
  if @page_css && @page_attr
    if next_page = doc.css(@page_css).attr(@page_attr).to_s
      fetch_index(generate_abs_url(next_page))
    else
      return
    end
  end

  book[:files].each_with_index{|item,index| item[:content] = "#{index}.html"}

  #保存书目
  save_book
end

#generate_book(book_name = nil) {|_self| ... } ⇒ Object

创建书本

Yields:

  • (_self)

Yield Parameters:



66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
 
# File 'lib/epub_book/book.rb', line 66

def generate_book(book_name=nil)
  #获取epub源数据
  fetch_book
  if  !@cover_css && @cover
    generate_cover = <<-eof
    convert #{File.expand_path("../../../#{@cover}",__FILE__)} -font tsxc.ttf -gravity center -fill red -pointsize 16 -draw "text 0,0 '#{book[:title]}'"  #{File.join(book_path,@cover)}
    eof
    system(generate_cover)
  end

  epub = EeePub.make

  epub.title book[:title]
  epub.creator @creator
  epub.publisher @creator
  epub.date Time.now
  epub.identifier "http://javy_liu.com/book/#{@folder_name}", :scheme => 'URL'
  epub.uid "http://javy_liu.com/book/#{@folder_name}"
  epub.cover @cover
  epub.subject book[:title]
  epub.description book[:description] if book[:description]

  book[:files] = book[:files][0...limit] if limit
  _files = []
  book[:files].collect! do |item|
    _file = File.join(book_path,item[:content])
    if test(?f, _file)
      _files.push(_file)
      item
    end
  end
  book[:files].compact!

  epub.files _files.push(File.join(book_path,@cover))
  epub.nav book[:files]

  book[:epub_file] = File.join(book_path,"#{book_name || @folder_name}.epub")

  yield self if block_given?

  epub.save(book[:epub_file])

  #send mail

  if mail_to
    mailer = Mailer.new
    mailer.to = mail_to
    mailer.add_file book[:epub_file]
    mailer.body = "您创建的电子书[#{book[:title]}]见附件\n"
    mailer.send_mail
  end

end

#link_host ⇒ Object 



48
49
50
 
# File 'lib/epub_book/book.rb', line 48

def link_host
  @link_host ||= @index_url[/\A(http:\/\/.*?)\/\w+/,1]
end

#save_book ⇒ Object 



58
59
60
61
62
 
# File 'lib/epub_book/book.rb', line 58

def save_book
  File.open(File.join(book_path,'index.yml' ),'w') do |f|
    f.write(@book.to_yaml)
  end
end