epub_book 0.1.13 → 0.1.18

checksums.yaml

@@ -1,7 +1,7 @@
 ---
-SHA1:
-  metadata.gz: 242f6d9ee15d028f0c72cd604790292af9c9b2d9
-  data.tar.gz: 9856509153f5caafd9f2aae760dc4f8013e302a6
+SHA256:
+  metadata.gz: df55d3587ef2f535b307f63117af6dafa73597a5007ad641e25ce88b17ef78b9
+  data.tar.gz: 0f78bbb8a4dee88540915e1cb95f28cd8abf872b07a8adeb74473ef171162bb5
 SHA512:
-  metadata.gz: 088faed36f70d1b635ae0e4eff49648d0f96b985388af0b5fdb319495c62b37662a559a9dc9a06e94564f6cd18173bbbb1238657be56d36f1fffadd3afa3ee27
-  data.tar.gz: 516aee02a5ffa3f7b6fee89ff55fd2399ae7564b69510d7ea1d4c70693906784bcc5df88a012be9223d508cc17033f6cb535f59999ddf76f514bb914953c1e44
+  metadata.gz: 2337421c1b0a1b0ce4d7096f6c899091044c5a69dc9297f95ec22b212f6d0f76677deb2edf53be7a7aa7d89df633ce56cea621b39d2692579a29c0c7060ffced
+  data.tar.gz: 23b4b10d567941e7b8dbf7a8662e370fa5b2d73ec0d6f979ca2e398007cee4b44743cac691395d6583b0aad1567e3ba7acf6cbebba7de36e50281c5450259124

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

@@ -27,7 +27,8 @@ Gem::Specification.new do |spec|
   spec.executables   = spec.files.grep(%r{^exe/}) { |f| File.basename(f) }
   spec.require_paths = ["lib"]
 
-  spec.add_dependency 'nokogiri','~> 1.6.8'
+  spec.add_dependency 'http','~> 2.1'
+  spec.add_dependency 'nokogiri','1.6.8.1'
   spec.add_dependency 'eeepub'
   spec.add_dependency 'zip-zip'
   spec.add_dependency 'mail'
@@ -37,5 +38,6 @@ Gem::Specification.new do |spec|
   spec.add_development_dependency "rake", "~> 10.0"
   spec.add_development_dependency "rspec", "~> 3.0"
   spec.add_development_dependency "pry"
+  spec.add_development_dependency "pry-byebug"
   spec.add_development_dependency "pry-doc"
 end

data/lib/epub_book.rb

@@ -1,5 +1,6 @@
 require "epub_book/version"
 require 'epub_book/book'
+require 'epub_book/loggable'
 require 'mail'
 
 module EpubBook
@@ -7,15 +8,18 @@ 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) 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'
       #self.mail_body =  "您创建的电子书见附件\n"
       self.mail_port = 25
+      self.log_level = "info"
     end
   end
 
+  extend Loggable
+
   #book initialize, and the block will prior the yml setting
   def self.create_book(url,bookname=nil,des_url=nil)
 

data/lib/epub_book/book.rb

@@ -1,4 +1,4 @@
-require 'open-uri'
+require 'http'
 require 'nokogiri'
 require 'eeepub'
 require 'base64'
@@ -16,13 +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
 
 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
-    attr_accessor :cover_css, :description_css,:path,:user_agent,:referer,:creator,:mail_to
+    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
 
 
     Reg = /<script.*?>.*?<\/script>/m
@@ -39,22 +39,27 @@ module EpubBook
       @cover = 'cover.jpg'
       @body_css = '.articlebody'
       @item_attr = "href"
+      @ext_name = 'epub'
       yield self if block_given?
-      @book_path = File.join((@path || `pwd`.strip), @folder_name)
     end
 
+    def book_path
+      @book_path ||= File.join((@path || `pwd`.strip), @folder_name)
+    end
 
     def link_host
-      @link_host ||= @index_url[/\A(http:\/\/.*?)\/\w+/,1]
+      @link_host ||= @index_url[/\A(https?:\/\/.*?)\/\w+/,1]
     end
 
     def 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: []})
+      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
 
+    #save catalog file
     def save_book
-      File.open(File.join(@book_path,'index.yml' ),'w') do |f|
+      File.open(File.join(book_path,'index.yml' ),'w') do |f|
         f.write(@book.to_yaml)
       end
     end
@@ -63,68 +68,78 @@ 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
+      fetch_index  if !test(?s,File.join(book_path,'index.yml'))
 
-      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
-
-      epub.files book[:files].map{|item| File.join(@book_path,item[:content])}.push(File.join(@book_path,@cover))
-      epub.nav book[:files]
+      book[:file_abs_name] = File.join(book_path,"#{book[:title]}.#{ext_name}")
 
+      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
 
-      epub_file = File.join(@book_path,"#{book_name || @folder_name}.epub")
+        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(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
-      puts mail_to
+
       if mail_to
         mailer = Mailer.new
         mailer.to = mail_to
-        mailer.add_file 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
 
     end
 
-
-
-
     #得到书目索引
     def fetch_index(url=nil)
+      book[:files] = []
       url ||= @index_url
-      doc = Nokogiri::HTML(judge_encoding(open(URI.encode(url),"User-Agent" => @user_agent ,'Referer'=> @referer).read))
+      doc = Nokogiri::HTML(judge_encoding(HTTP.headers("User-Agent" => @user_agent ,'Referer'=> @referer).get(URI.encode(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(open(URI.encode(@des_url),"User-Agent" => @user_agent ,'Referer'=> @referer).read))
+                 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
 
-
+      #binding.pry
       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?('#')
@@ -151,49 +166,72 @@ 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')
+      end
       book[:files].each_with_index do |item,index|
         break if limit && index >= limit
 
-        content_path = File.join(@book_path,item[:content])
+        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(open(item[:url],"User-Agent" => @user_agent,'Referer'=> @referer).read))
-
-          File.open(content_path,'w') do |f|
-            f.write("<h3>#{item[:label]}</h3>")
-            f.write(doc_file.css(@body_css).to_s.gsub(Reg,''))
+          doc_file = Nokogiri::HTML(judge_encoding(HTTP.headers("User-Agent" => @user_agent,'Referer'=> @referer).get(item[:url]).to_s))
+
+          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
-
-          puts item[:label]
-
         rescue  Exception => e
-          puts "Error:#{e.message}"
-          puts e.backtrace.inspect
+          EpubBook.logger.info "Error:#{e.message},#{item.inspect}"
+          #EpubBook.logger.info e.backtrace
           next
         end
       end
 
+      txt_file.close if ext_name == 'txt'
+
     end
 
 
     private
     #is valid encoding
     def judge_encoding(str)
+      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?
-      /<meta.*?charset\s*=[\s\"\']?utf-8/i =~ str ? str : str.force_encoding('gbk').encode('utf-8')
+
+      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)
+        cover_path = File.join(book_path,@cover)
         system("curl #{cover_url} -o #{cover_path} ")
         book[:cover] = cover_path
       end
@@ -204,13 +242,13 @@ module EpubBook
     end
 
     def generate_abs_url(url)
-      if _href.start_with?("http")
+      if url.start_with?("http")
         url
-      elsif _href.start_with?("/")
-        "#{link_host}#{_href}"
+      elsif url.start_with?("/")
+        "#{link_host}#{url}"
       else
         @path_name ||= @index_url[/.*\//]
-        "#{@path_name}#{_href}"
+        "#{@path_name}#{url}"
       end
 
     end

data/lib/epub_book/loggable.rb

@@ -0,0 +1,69 @@
+require 'logger'
+module EpubBook
+
+  # Contains logging behaviour.
+  module Loggable
+
+    # Get the logger.
+    #
+    # @note Will try to grab Rails' logger first before creating a new logger
+    #   with stdout.
+    #
+    # @example Get the logger.
+    #   Loggable.logger
+    #
+    # @return [ Logger ] The logger.
+    #
+    # @since 3.0.0
+    def logger
+      return @logger if defined?(@logger)
+      @logger = rails_logger || default_logger
+    end
+
+    # Set the logger.
+    #
+    # @example Set the logger.
+    #   Loggable.logger = Logger.new($stdout)
+    #
+    # @param [ Logger ] The logger to set.
+    #
+    # @return [ Logger ] The new logger.
+    #
+    # @since 3.0.0
+    def logger=(logger)
+      @logger = logger
+    end
+
+    private
+
+    # Gets the default EpubBook logger - stdout.
+    #
+    # @api private
+    #
+    # @example Get the default logger.
+    #   Loggable.default_logger
+    #
+    # @return [ Logger ] The default logger.
+    #
+    # @since 3.0.0
+    def default_logger
+      logger = ::Logger.new($stdout)
+      logger.level = EpubBook.config.log_level
+      logger
+    end
+
+    # Get the Rails logger if it's defined.
+    #
+    # @api private
+    #
+    # @example Get Rails' logger.
+    #   Loggable.rails_logger
+    #
+    # @return [ Logger ] The Rails logger.
+    #
+    # @since 3.0.0
+    def rails_logger
+      defined?(::Rails) && ::Rails.respond_to?(:logger) && ::Rails.logger
+    end
+  end
+end

data/lib/epub_book/version.rb

@@ -1,3 +1,3 @@
 module EpubBook
-  VERSION = "0.1.13"
+  VERSION = "0.1.18"
 end

metadata

@@ -1,29 +1,43 @@
 --- !ruby/object:Gem::Specification
 name: epub_book
 version: !ruby/object:Gem::Version
-  version: 0.1.13
+  version: 0.1.18
 platform: ruby
 authors:
 - qmliu
-autorequire: 
+autorequire:
 bindir: exe
 cert_chain: []
-date: 2016-09-06 00:00:00.000000000 Z
+date: 2021-06-24 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
-  name: nokogiri
+  name: http
   requirement: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: 1.6.8
+        version: '2.1'
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: 1.6.8
+        version: '2.1'
+- !ruby/object:Gem::Dependency
+  name: nokogiri
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - '='
+      - !ruby/object:Gem::Version
+        version: 1.6.8.1
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - '='
+      - !ruby/object:Gem::Version
+        version: 1.6.8.1
 - !ruby/object:Gem::Dependency
   name: eeepub
   requirement: !ruby/object:Gem::Requirement
@@ -122,6 +136,20 @@ dependencies:
     - - ">="
       - !ruby/object:Gem::Version
         version: '0'
+- !ruby/object:Gem::Dependency
+  name: pry-byebug
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
 - !ruby/object:Gem::Dependency
   name: pry-doc
   requirement: !ruby/object:Gem::Requirement
@@ -153,13 +181,12 @@ files:
 - LICENSE.txt
 - README.md
 - Rakefile
-- bin/console
-- bin/setup
 - cover.jpg
 - epub_book.gemspec
 - exe/create_book
 - lib/epub_book.rb
 - lib/epub_book/book.rb
+- lib/epub_book/loggable.rb
 - lib/epub_book/mailer.rb
 - lib/epub_book/version.rb
 - tsxc.ttf
@@ -168,7 +195,7 @@ licenses:
 - MIT
 metadata:
   allowed_push_host: https://rubygems.org
-post_install_message: 
+post_install_message:
 rdoc_options: []
 require_paths:
 - lib
@@ -183,9 +210,8 @@ 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