EPUBChop 0.1.1 → 0.1.3

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: eae1dc4152536744c7d2a51ee3231526632c742b
-  data.tar.gz: de5b13b1bfe346629bf89ff4c9c7764db3cb1746
+  metadata.gz: 310ed2a66e86fba757f244581f9cfbe4cbf69094
+  data.tar.gz: 4243b1aae661b31a47d3c78d2e027717f1c6d5a6
 SHA512:
-  metadata.gz: 410cec8dcdda2bd1aeff728fbb22d87c12d797ca805d19eab81ee8f1152b9b8be52fb78e11cd9ca942e6544fe70c32f9ecd378fc09d06299a0c5a2b7221adb8e
-  data.tar.gz: ae763206822c42f8df9887c8ec9df52bf9e7bfc97c419a243ae9435990c82819ddb17df2c8b37cd0d0e65d8310329bdd6a12221f8c738e5357ee89b6d2bf8ab7
+  metadata.gz: 4b4afb0cc16991ced9a95117a45d974a1b58797a574eaed765fb17009dbf4b1df725fe4e0d347939fbdf5dec8316d36b03c82471b6af0478993766c7216d5022
+  data.tar.gz: 63959397ad7d9589e0c0ee4b466ec46360eb9372e43d3fe4068c4515ae56f83ef24ca2175c212a7b66d51b4a8c86251c7b34ae493bb8cdbc23e76feaa56bbc6c

data/EPUBChop.gemspec

@@ -21,7 +21,7 @@ Gem::Specification.new do |spec|
   spec.add_development_dependency "bundler", "~> 1.3"
   spec.add_development_dependency "rake"
   spec.add_development_dependency "rspec"
-  spec.add_runtime_dependency "epubinfo_with_toc"
-  spec.add_runtime_dependency "rubyzip", "~> 1.0"
+  spec.add_runtime_dependency "epubinfo_with_toc", "0.5.6"
+  spec.add_runtime_dependency "rubyzip", "~> 1.1.3"
   spec.add_runtime_dependency "nokogiri"
 end

data/bin/epubchop

@@ -35,14 +35,17 @@ begin
     chop_by = options[:chop]
     verbose = options[:verbose]
 
-    puts "loading EPUB #{filename}"
-    b=EPUBChop.get(filename, :chop_by => chop_by.to_sym, :verbose => verbose)
-    puts 'chopping EPUB'
-    c=b.chop({:base => base.to_s, :words => words, :text => text})
-    puts 'rebuilding EPUB'
-    FileUtils.move(c, "chopped_#{File.basename(filename)}")
+    Dir.glob(filename).each do |f|
+      puts "loading EPUB #{f}"
+      b=EPUBChop.get(f, :chop_by => chop_by.to_sym, :verbose => verbose)
+      puts 'chopping EPUB'
+      c=b.chop({:base => base.to_s, :words => words, :text => text})
+      puts 'rebuilding EPUB'
+      FileUtils.move(c, "chopped_#{File.basename(f)}")
+    end
 
 rescue Exception => e
   puts "An error occured\n#{e.message}"
+  puts e.backtrace.join("/n")
   exit 1
 end

data/lib/EPUBChop/chop.rb

@@ -1,4 +1,4 @@
-#encoding: UTF-8
+# coding: utf-8
 require 'nokogiri'
 require 'epubinfo'
 require 'tempfile'
@@ -39,7 +39,7 @@ module EPUBChop
 
 
       return rebuild_epub_from_tmp_dir(extract_dir)
-    rescue Zip::ZipError => e
+    rescue Zip::Error => e
       raise RuntimeError, "Error processing EPUB #{@book.table_of_contents.parser.path}.\n #{e.message}", e.backtrace
     rescue Exception => e
       puts e.backtrace.join("\n")
@@ -80,18 +80,19 @@ module EPUBChop
 
           else
             #noinspection RubyResolve
-            resource = Nokogiri::HTML(@book.table_of_contents.resources[filename]) do |config|
-            #resource = Nokogiri::HTML.parse(@book.table_of_contents.resources[filename], 'UTF-8') do |config|
+            resource = Nokogiri::HTML(@book.table_of_contents.resources[filename].force_encoding('UTF-8')) do |config|
               config.noblanks.nonet
             end
-            resource.encoding = 'UTF-8'
 
+#            resource.encoding = 'UTF-8'
             resource = chop_file(resource, processed_file_size)
 
             #persist page
+            save_options = Nokogiri::XML::Node::SaveOptions::FORMAT | Nokogiri::XML::Node::SaveOptions::NO_DECLARATION | Nokogiri::XML::Node::SaveOptions::AS_XHTML
             File.open("#{extract_dir}/#{filename}", 'w:UTF-8') do |f|
-              #  f.puts resource.to_xml(:save_with => Nokogiri::XML::Node::SaveOptions::NO_DECLARATION)
-              f.puts resource.serialize(:encoding => 'UTF-8', :save_with => Nokogiri::XML::Node::SaveOptions::NO_DECLARATION)
+              #f.puts resource.serialize(:encoding => 'ISO-8859-1', :save_with => save_options)
+              f.puts resource.serialize(:encoding => 'UTF-8', :save_with => save_options)
+              #f.puts resource.serialize(:encoding => resource.encoding, :save_with => save_options)
             end
 
           end
@@ -150,12 +151,31 @@ module EPUBChop
         end
       end
 
+      meta = Nokogiri::XML::Node.new('meta', resource)
+      meta['http-equiv'] = "Content-Type"
+      meta['content'] = "text/html; charset=UTF-8"
+
+      meta_charset = Nokogiri::XML::Node.new('meta', resource)
+      meta_charset['charset'] = 'UTF-8'
+
+      resource.css('head').first << meta
+      resource.css('head').first << meta_charset
+
       resource
     end
 
 
     def rebuild_epub_from_tmp_dir(extract_dir)
       #zip new ebook
+
+      Zip.setup do |z|
+        z.write_zip64_support = false
+        z.on_exists_proc = true
+        z.continue_on_exists_proc = true
+        z.unicode_names = true
+        z.default_compression = Zlib::BEST_COMPRESSION
+      end
+
       new_ebook_name = Tempfile.new(['epub', '.epub'], Dir.tmpdir)
       new_ebook_name_path = new_ebook_name.path
       new_ebook_name_path.gsub!('-', '')
@@ -166,7 +186,14 @@ module EPUBChop
 
       #minetype should be the first entry and should not be zipped. Else FIDO will not know that this is an EPUB
       mimetype = epub_files.delete("#{extract_dir}/mimetype")
-      mimetype_entry = Zip::Entry.new(zipfile, mimetype.sub("#{extract_dir}/", ''), '', '', 0, 0, Zip::Entry::STORED)
+      mimetype_entry = Zip::Entry.new(zipfile,                              #@zipfile
+                                      mimetype.sub("#{extract_dir}/", ''),  #@name
+                                      '',                                   #@comment
+                                      '',                                   #@extra
+                                      0,                                    #@compressed_size
+                                      0,                                    #@crc
+                                      Zip::Entry::STORED)                   #@compression_method
+
       zipfile.add(mimetype_entry, mimetype) unless mimetype.nil?
 
       #all the other files

data/lib/EPUBChop/version.rb

@@ -1,3 +1,3 @@
 module EPUBChop
-  VERSION = "0.1.1"
+  VERSION = "0.1.3"
 end

data/lib/EPUBChop/zip/entry.rb

@@ -1,10 +1,22 @@
-# Monkey patching bug in rubyzip, currently fixed on master, but not yet released
-# and EPUBChop brings in the 1.0 release.
+# Monkey patching bug in rubyzip can not reproduce this on the main branch
+require 'zip/version'
 
-module Zip
-  class Entry
-    private
-      def prep_zip64_extra(for_local_header)
+if Zip::VERSION.eql?('1.1.3')
+  module Zip
+    class Entry
+      alias_method :old_write_to_zip_output_stream, :write_to_zip_output_stream
+
+      def write_to_zip_output_stream(zip_output_stream) #:nodoc:all
+        if @ftype == :directory
+          zip_output_stream.put_next_entry(self, nil, nil, ::Zip::Entry::STORED)
+        elsif @filepath
+          zip_output_stream.put_next_entry(self, nil, nil, self.compression_method || ::Zip::Entry::DEFLATED )
+          get_input_stream { |is| ::Zip::IOExtras.copy_stream(zip_output_stream, is) }
+        else
+          zip_output_stream.copy_raw_entry(self)
+        end
       end
+
+    end
   end
-end
+end

data/spec/epubchop_spec.rb

@@ -30,4 +30,20 @@ describe 'EPUBChop' do
       deviation.should < 5
     end
 
+    context 'created epub file' do
+      before(:all) do
+        b = EPUBChop.get('./spec/epub/Verne_20000_West_pg11393.epub', {:base => :percentage, :words => 10})
+        @chopped = b.chop()
+      end
+
+      it 'should contain a minetype file' do
+        f= File.open(@chopped, 'rb')
+        first_100_bytes = f.read(100)
+        f.close
+
+        first_100_bytes.should match(/mimetypeapplication\/epub\+zip/)
+      end
+
+    end
+
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: EPUBChop
 version: !ruby/object:Gem::Version
-  version: 0.1.1
+  version: 0.1.3
 platform: ruby
 authors:
 - Mehmet Celik
 autorequire:
 bindir: bin
 cert_chain: []
-date: 2014-02-26 00:00:00.000000000 Z
+date: 2014-04-24 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: bundler
@@ -56,14 +56,14 @@ dependencies:
   name: epubinfo_with_toc
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - '>='
+    - - '='
       - !ruby/object:Gem::Version
-        version: '0'
+        version: 0.5.6
   requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - '>='
+    - - '='
       - !ruby/object:Gem::Version
-        version: '0'
+        version: 0.5.6
   prerelease: false
   type: :runtime
 - !ruby/object:Gem::Dependency
@@ -72,12 +72,12 @@ dependencies:
     requirements:
     - - ~>
       - !ruby/object:Gem::Version
-        version: '1.0'
+        version: 1.1.3
   requirement: !ruby/object:Gem::Requirement
     requirements:
     - - ~>
       - !ruby/object:Gem::Version
-        version: '1.0'
+        version: 1.1.3
   prerelease: false
   type: :runtime
 - !ruby/object:Gem::Dependency