renzaburo.jpにあるHTML形式の文書をまとめて青空文庫形式に

OBSOLETE

#
# This command was obsolete. see also https://github.com/tdtds/aozoragen
#

#!/usr/bin/env ruby
# -*- coding: utf-8; -*-
#
# usage: renzaburo2aozora <URL>
#    URL: the index of a novel in renzaburo, only HTML version NOT Flash.
#

require 'open-uri'
require 'pathname'
require 'nokogiri'

def get_content( uri, book_title = '' )
	text = ''
	html = open( uri, 'r:CP932', &:read ).encode( 'UTF-8' )
	html = html.gsub( /\&mdash;/, "\u2500" ).gsub( /\&quot;/, "\u201D" )
	(Nokogiri( html ) / 'div#mainContent' ).each do |content|
		(content / 'h3').each do |t|
			title = t.text.sub( /^『#{book_title}』　/, '' )
			text << "［＃改ページ］\n\n" << title << "\n\n"
		end
		(content / 'div.textBlock p' ).each do |para|
			next if /＜次回につづく＞/ =~ para.text
			text << '　' * 10 if (para.attr('class') || '').index( 'txtAlignC' )
			text << para.text.gsub( /<br>/, "\n" ) << "\n\n"
		end
	end
	text
end

while i = ARGV.shift do
	index = URI( i.sub( /index\.html$/, '' ) )
	bookname = Pathname( index.to_s ).basename

	title = ''
	text = ''
	html = Nokogiri( open( index, 'r:CP932', &:read ) )
	(html / 'title').each do |t|
		title = t.text.sub( /｜.*/, '' )
		text << title << "\n"
	end
	(html / 'div.textBlock strong' ).each do |st|
		text << st.text << "\n\n\n"
	end
	(html / 'ul.btnList li.withDate a' ).each do |li|
		text << get_content( index.merge( li.attr( :href ) ), title )
	end

	open( "#{bookname}.txt", 'w' ) do |w|
		w.write text
	end
end