epub_book 0.1.17 → 0.1.23

checksums.yaml

@@ -1,7 +1,7 @@
 ---
-SHA1:
-  metadata.gz: da90462dcefd8f95b3591adec93ca673f192c981
-  data.tar.gz: b0af4b9ed6486b59c222cff218a1edd8d6e37c64
+SHA256:
+  metadata.gz: 44b412b721dc1e443c1a9bb1a685f2e6968ff0c8506bf9be113f94e4827f4da0
+  data.tar.gz: 8e5344dd88fd9e3f550e6962912dae4fe67057f214f08852188642c9d92bfc6d
 SHA512:
-  metadata.gz: 951efc7f602fc6a63d000a116a010605154dcbd53e88187ac6e9189868fe0609f39f8a64f67c2e94925ae720521eaab40e892c2cef2989c7f8bbbe4a7f787baa
-  data.tar.gz: ba37d2daddb9c2cf60029cda72cd50df8137a9c7c6cd1e6b3116ef40fe696e07b34caecdce7aab62b99090b2eb86e069d10a978367e38cf103aef72f5cdea9eb
+  metadata.gz: cee0921eee24cc47105b595e850612c7cf34f9a8ebd2de1d447867f59eaaf1624847f3526a6313ecb61ef447477df26e6a2d11d2fb1a5827b5b72697bfdf0c9b
+  data.tar.gz: 6ef82d375814323d227f8c210ae14f2d498aff710ae9268b38c78af32b0723fd8a367cebb40a4f00c7346ebd6ca7789f17be2d6100883b0e99564a2baaa5dbee

data/Gemfile

@@ -1,4 +1,5 @@
-source 'https://ruby.taobao.org'
+#source 'https://ruby.taobao.org'
+source 'https://gems.ruby-china.com/'
 
 # Specify your gem's dependencies in epub_book.gemspec
 gemspec

data/README.md

@@ -29,6 +29,7 @@ Or install it yourself as:
     book.body_css = '.wrapper #content'
     book.creator = 'javy_liu'
     book.path  = '/tmp'
+    book.ext_name  = 'txt'
   end
 ``` 
 
@@ -44,6 +45,7 @@ Or install it yourself as:
     config.mail_user_name = 'ex@example.com'
     config.mail_password = 'password'
     config.mail_from = 'yourmail@example.com'
+    config.ext_name = 'epub'
   end
 ```
 Or use a ./default_setting.yml file have following content
@@ -69,6 +71,7 @@ Or use a ./default_setting.yml file have following content
     creator: 'user name'
     path: '/'
     mail_to: 'yourmail@example.com'
+    ext_name: 'txt'
 
   #special host book setting(the key is book_url's host which replacing the dot with underline)
   #http://www.piaotian.net/html/0/431/index.html
@@ -111,6 +114,7 @@ if you have setting file , you can create book like following
     book.creator         #epub creator
     book.path            #epub book save path
     book.mail_to         #if your want send by email when epub created, set this to your email
+    book.ext_name        #can create epub or txt file
 ```
 ## Perform `create_book` in your terminal
 ```bash

data/epub_book.gemspec

@@ -28,15 +28,16 @@ Gem::Specification.new do |spec|
   spec.require_paths = ["lib"]
 
   spec.add_dependency 'http','~> 2.1'
-  spec.add_dependency 'nokogiri','~> 1.6'
-  spec.add_dependency 'eeepub'
-  spec.add_dependency 'zip-zip'
-  spec.add_dependency 'mail'
+  spec.add_runtime_dependency 'nokogiri', '~> 1.11', '>= 1.11.4'
+  spec.add_dependency 'eeepub', '~> 0.8.1'
+  spec.add_dependency 'zip-zip', '~> 0.3'
+  spec.add_dependency 'mail', '~>2.7.0'
 
 
-  spec.add_development_dependency "bundler", "~> 1.12"
-  spec.add_development_dependency "rake", "~> 10.0"
+  spec.add_development_dependency 'bundler', '~> 2.2', '>= 2.2.27'
+  spec.add_development_dependency "rake",  '~> 13.0','>= 13.0.6'
   spec.add_development_dependency "rspec", "~> 3.0"
-  spec.add_development_dependency "pry"
-  spec.add_development_dependency "pry-doc"
+  spec.add_development_dependency "pry", '~> 0.14', '>= 0.14.1'
+  spec.add_development_dependency "pry-byebug", '~> 3.8', '>= 3.8.0'
+  spec.add_development_dependency "pry-doc", '~> 1.2', '>= 1.2.0'
 end

data/lib/epub_book/book.rb

@@ -16,12 +16,13 @@ require 'yaml'
 #user_agent 访问代理
 #referer 访问原地址
 #creator 责任人
-
+#ext_name 扩展名 epub,txt
+#ignore_txt 忽略字符，带有ignore_txt的行将被删除
 module EpubBook
   class Book
     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/"
-    attr_accessor :title_css, :index_item_css, :body_css, :limit, :item_attr, :page_css, :page_attr,:cover,:cover_css, :description_css,:path,:user_agent,:referer,:creator,:mail_to, :folder_name,:des_url
+    attr_accessor :title_css, :index_item_css, :body_css, :limit, :item_attr, :page_css, :page_attr,:cover,:cover_css, :description_css,:path,:user_agent,:referer,:creator,:mail_to, :folder_name,:des_url,:ext_name,:ignore_txt
 
 
     Reg = /<script.*?>.*?<\/script>/m
@@ -38,6 +39,7 @@ module EpubBook
       @cover = 'cover.jpg'
       @body_css = '.articlebody'
       @item_attr = "href"
+      @ext_name = 'epub'
       yield self if block_given?
     end
 
@@ -46,7 +48,7 @@ module EpubBook
     end
 
     def link_host
-      @link_host ||= @index_url[/\A(http:\/\/.*?)\/\w+/,1]
+      @link_host ||= @index_url[/\A(https?:\/\/.*?)\/\w+/,1]
     end
 
     def book
@@ -55,6 +57,7 @@ module EpubBook
       @book = test(?s,File.join(book_path,'index.yml')) ? YAML.load(File.open(File.join(book_path,'index.yml'))) : {files: []}
     end
 
+    #save catalog file
     def save_book
       File.open(File.join(book_path,'index.yml' ),'w') do |f|
         f.write(@book.to_yaml)
@@ -65,52 +68,54 @@ module EpubBook
     #创建书本
     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!
+      fetch_index  if !test(?s,File.join(book_path,'index.yml'))
 
-      epub.files _files.push(File.join(book_path,@cover))
-      epub.nav book[:files]
+      book[:file_abs_name] = File.join(book_path,"#{book[:title]}.#{ext_name}")
 
-      book[:epub_file] = File.join(book_path,"#{book_name || @folder_name}.epub")
+      fetch_book
+      if ext_name == 'epub'
+        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
 
-      yield self if block_given?
+        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.save(book[:epub_file])
+        epub.files _files.push(File.join(book_path,@cover))
+        epub.nav book[:files]
+        yield self if block_given?
 
+        epub.save(book[:file_abs_name])
+      end
       #send mail
 
       if mail_to
         mailer = Mailer.new
         mailer.to = mail_to
-        mailer.add_file book[:epub_file]
+        mailer.add_file book[:file_abs_name]
         mailer.body = "您创建的电子书[#{book[:title]}]见附件\n"
         mailer.send_mail
       end
@@ -121,20 +126,22 @@ module EpubBook
     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))
+      doc = Nokogiri::HTML(judge_encoding(HTTP.headers("User-Agent" => @user_agent ,'Referer'=> @referer).get(url).to_s))
       #generate index.yml
+      EpubBook.logger.info "------Fetch index--#{url}---------------"
 
       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))
+                 Nokogiri::HTML(judge_encoding(HTTP.headers("User-Agent" => @user_agent ,'Referer'=> @referer).get(generate_abs_url(doc.css(@des_url).attr("href").to_s)).to_s))
                end
         get_des(doc1)
       end
 
+      #binding.pry
       doc.css(@index_item_css).each do |item|
-        _href = URI.encode(item.attr(@item_attr).to_s)
+        _href = item.attr(@item_attr).to_s
         next if _href.start_with?('javascript') || _href.start_with?('#')
 
         _href = generate_abs_url(_href)
@@ -159,22 +166,40 @@ module EpubBook
 
     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'))
+      #fetch_index  if !test(?s,File.join(book_path,'index.yml'))
       EpubBook.logger.info "------Fetch book----------"
+      #open a txt file to write
+      if ext_name == 'txt'
+        txt_file = File.open(book[:file_abs_name], 'a')
+        txt_file.write("简介\n\n")
+        txt_file.write('  ')
+        txt_file.write(book[:description] || " ")
+      end
+
       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
+        #binding.pry
         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,''))
+          EpubBook.logger.info item[:label]
+          #binding.pry
+          if ext_name == 'pub'
+            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
+          else
+            txt_file.write("\n\n")
+            txt_file.write(item[:label])
+            txt_file.write("\n  ")
+            txt_file.write(doc_file.css(@body_css).text)
           end
         rescue  Exception => e
           EpubBook.logger.info "Error:#{e.message},#{item.inspect}"
@@ -182,6 +207,15 @@ module EpubBook
           next
         end
       end
+      if ext_name == 'txt'
+        txt_file.close
+        EpubBook.logger.info "=============去除包含指定忽略字符的行======="
+        EpubBook.logger.info ignore_txt
+        if ignore_txt
+          system("sed -i -r '/#{ignore_txt}/d' #{book[:file_abs_name]}")
+        end
+
+      end
 
     end
 
@@ -189,15 +223,23 @@ module EpubBook
     private
     #is valid encoding
     def judge_encoding(str)
-      /<meta.*?charset\s*=[\s\"\']?utf-8/i =~ str ? str : str.force_encoding('gbk').encode('utf-8',invalid: :replace, undef: :replace)
+      EpubBook.logger.info str.encoding
+      /<meta.*?charset\s*=[\s\"\']?utf-8/i =~ str ? str : str.force_encoding('gbk').encode!('utf-8',invalid: :replace, undef: :replace)
       str.scrub! unless str.valid_encoding?
+
+      EpubBook.logger.info "-------encode 后 #{str.encoding}"
       str
     end
 
     #得到书名，介绍，及封面
     def get_des(doc)
       book[:title] = doc.css(@title_css).text.strip
-      if @cover_css && !book[:cover]
+
+      #EpubBook.logger.info doc
+      #EpubBook.logger.info @title_css
+
+      #binding.pry
+      if @cover_css && !book[:cover] && ext_name == 'epub'
         cover_url = doc.css(@cover_css).attr("src").to_s
         cover_url = generate_abs_url(cover_url) #link_host + cover_url unless cover_url.start_with?("http")
         cover_path = File.join(book_path,@cover)

data/lib/epub_book/loggable.rb

@@ -1,3 +1,4 @@
+require 'logger'
 module EpubBook
 
   # Contains logging behaviour.
@@ -46,7 +47,7 @@ module EpubBook
     #
     # @since 3.0.0
     def default_logger
-      logger = Logger.new($stdout)
+      logger = ::Logger.new($stdout)
       logger.level = EpubBook.config.log_level
       logger
     end

data/lib/epub_book/version.rb

@@ -1,3 +1,3 @@
 module EpubBook
-  VERSION = "0.1.17"
+  VERSION = "0.1.23"
 end

data/lib/epub_book.rb

@@ -1,6 +1,6 @@
-require "epub_book/version"
-require 'epub_book/book'
-require 'epub_book/loggable'
+require_relative "epub_book/version"
+require_relative 'epub_book/book'
+require_relative 'epub_book/loggable'
 require 'mail'
 
 module EpubBook
@@ -8,7 +8,7 @@ module EpubBook
   autoload :Book, "epub_book/book"
   autoload :Mailer, "epub_book/mailer"
 
-  Config = Struct.new(:setting_file,:mail_from,:mail_subject,:mail_body,:mail_address,:mail_port,:mail_user_name,:mail_password,:log_level) do
+  Config = Struct.new(:setting_file,:mail_from,:mail_subject,:mail_body,:mail_address,:mail_port,:mail_user_name,:mail_password,:log_level, :ext_name) do
     include Singleton
     def initialize
       self.mail_subject =  'epub 电子书'

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: epub_book
 version: !ruby/object:Gem::Version
-  version: 0.1.17
+  version: 0.1.23
 platform: ruby
 authors:
 - qmliu
-autorequire: 
+autorequire:
 bindir: exe
 cert_chain: []
-date: 2016-12-27 00:00:00.000000000 Z
+date: 2021-11-15 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: http
@@ -30,84 +30,102 @@ dependencies:
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '1.6'
+        version: '1.11'
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: 1.11.4
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '1.6'
+        version: '1.11'
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: 1.11.4
 - !ruby/object:Gem::Dependency
   name: eeepub
   requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - ">="
+    - - "~>"
       - !ruby/object:Gem::Version
-        version: '0'
+        version: 0.8.1
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - ">="
+    - - "~>"
       - !ruby/object:Gem::Version
-        version: '0'
+        version: 0.8.1
 - !ruby/object:Gem::Dependency
   name: zip-zip
   requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - ">="
+    - - "~>"
       - !ruby/object:Gem::Version
-        version: '0'
+        version: '0.3'
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - ">="
+    - - "~>"
       - !ruby/object:Gem::Version
-        version: '0'
+        version: '0.3'
 - !ruby/object:Gem::Dependency
   name: mail
   requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - ">="
+    - - "~>"
       - !ruby/object:Gem::Version
-        version: '0'
+        version: 2.7.0
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - ">="
+    - - "~>"
       - !ruby/object:Gem::Version
-        version: '0'
+        version: 2.7.0
 - !ruby/object:Gem::Dependency
   name: bundler
   requirement: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '1.12'
+        version: '2.2'
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: 2.2.27
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '1.12'
+        version: '2.2'
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: 2.2.27
 - !ruby/object:Gem::Dependency
   name: rake
   requirement: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '10.0'
+        version: '13.0'
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: 13.0.6
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '10.0'
+        version: '13.0'
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: 13.0.6
 - !ruby/object:Gem::Dependency
   name: rspec
   requirement: !ruby/object:Gem::Requirement
@@ -126,30 +144,62 @@ dependencies:
   name: pry
   requirement: !ruby/object:Gem::Requirement
     requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '0.14'
     - - ">="
       - !ruby/object:Gem::Version
-        version: '0'
+        version: 0.14.1
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '0.14'
     - - ">="
       - !ruby/object:Gem::Version
-        version: '0'
+        version: 0.14.1
+- !ruby/object:Gem::Dependency
+  name: pry-byebug
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '3.8'
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: 3.8.0
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '3.8'
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: 3.8.0
 - !ruby/object:Gem::Dependency
   name: pry-doc
   requirement: !ruby/object:Gem::Requirement
     requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '1.2'
     - - ">="
       - !ruby/object:Gem::Version
-        version: '0'
+        version: 1.2.0
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '1.2'
     - - ">="
       - !ruby/object:Gem::Version
-        version: '0'
+        version: 1.2.0
 description: 'create epub from a book index url and mail to you. by setting a default_setting.yml
   you can use it in shell. '
 email:
@@ -167,8 +217,6 @@ files:
 - LICENSE.txt
 - README.md
 - Rakefile
-- bin/console
-- bin/setup
 - cover.jpg
 - epub_book.gemspec
 - exe/create_book
@@ -183,7 +231,7 @@ licenses:
 - MIT
 metadata:
   allowed_push_host: https://rubygems.org
-post_install_message: 
+post_install_message:
 rdoc_options: []
 require_paths:
 - lib
@@ -198,9 +246,8 @@ required_rubygems_version: !ruby/object:Gem::Requirement
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
-rubyforge_project: 
-rubygems_version: 2.4.8
-signing_key: 
+rubygems_version: 3.2.3
+signing_key:
 specification_version: 4
 summary: create epub from a book index url and mail to you.
 test_files: []

data/bin/console

@@ -1,14 +0,0 @@
-#!/usr/bin/env ruby
-
-require "bundler/setup"
-require "epub_book"
-
-# You can add fixtures and/or initialization code here to make experimenting
-# with your gem easier. You can also use a different console, if you like.
-
-# (If you use this, don't forget to add pry to your Gemfile!)
-require "pry"
-Pry.start
-
-#require "irb"
-#IRB.start

data/bin/setup

@@ -1,8 +0,0 @@
-#!/usr/bin/env bash
-set -euo pipefail
-IFS=$'\n\t'
-set -vx
-
-bundle install
-
-# Do any other automated setup that you need to do here