# Copyright (C) 2007 Marco Costantini
# Copyright (C) 2014 Matijs van Zuijlen
# based on ibs_it.rb by Claudio Belotti
#
# Alexandria is free software; you can redistribute it and/or
# modify it under the terms of the GNU General Public License as
# published by the Free Software Foundation; either version 2 of the
# License, or (at your option) any later version.
#
# Alexandria is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
# General Public License for more details.
#
# You should have received a copy of the GNU General Public
# License along with Alexandria; see the file COPYING.  If not,
# write to the Free Software Foundation, Inc., 51 Franklin Street,
# Fifth Floor, Boston, MA 02110-1301 USA.

require 'fileutils'
require 'net/http'
require 'open-uri'
# require 'cgi'

module Alexandria
  class BookProviders
    class BOL_itProvider < GenericProvider
      BASE_URI = 'http://www.bol.it'
      CACHE_DIR = File.join(Alexandria::Library::DIR, '.bol_it_cache')
      REFERER = BASE_URI
      LOCALE = 'libri' # possible locales are: "libri", "inglesi", "video", "musica", "choco"
      def initialize
        super('BOL_it', 'BOL (Italy)')
        FileUtils.mkdir_p(CACHE_DIR) unless File.exist?(CACHE_DIR)
        # no preferences for the moment
        at_exit { clean_cache }
      end

      def search(criterion, type)
        criterion = criterion.convert('ISO-8859-1', 'UTF-8')
        req = BASE_URI + '/' + LOCALE + '/'
        req += case type
               when SEARCH_BY_ISBN
                 'scheda/'

               when SEARCH_BY_TITLE
                 'risultatoricerca?action=bolrisultatoricerca&skin=bol&filtro_ricerca=BOL&quick_type=Titolo&titolo='

               when SEARCH_BY_AUTHORS
                 'risultatoricerca?action=bolrisultatoricerca&skin=bol&filtro_ricerca=BOL&quick_type=Autore&titolo='

               when SEARCH_BY_KEYWORD
                 'risultatoricerca?action=bolrisultatoricerca&skin=bol&filtro_ricerca=BOL&quick_type=Parola%20chiave&titolo='

               else
                 raise InvalidSearchTypeError

               end

        ## warning: this provider uses pages like http://www.bol.it/libri/scheda/ea978888584104 with 12 numbers, without the checksum
        criterion = 'ea' + Library.canonicalise_ean(criterion)[0..-2] + '.html' if type == SEARCH_BY_ISBN
        req += CGI.escape(criterion)
        p req if $DEBUG
        data = transport.get(URI.parse(req))
        if type == SEARCH_BY_ISBN
          to_book(data) # rescue raise NoResultsError
        else
          begin
            results = []
            each_book_page(data) do |code, _title|
              results << to_book(transport.get(URI.parse(BASE_URI + "/#{LOCALE}/scheda/ea" + code)))
            end
            return results
          rescue
            raise NoResultsError
          end
        end
      end

      def url(book)
        BASE_URI + "/#{LOCALE}/scheda/ea" + Library.canonicalise_ean(book.isbn)[0..-2] + '.html'
      end

      private

      def to_book(data)
        raise NoResultsError if /Scheda libro non completa  \(TP null\)/.match(data)
        data = data.convert('UTF-8', 'ISO-8859-1')

        md = /<INPUT type =hidden name ="mailTitolo" value="([^"]+)/.match(data)
        raise 'No title' unless md
        title = CGI.unescape(md[1].strip)

        authors = []
        if (md = /<INPUT type =HIDDEN name ="mailAutore" value="([^"]+)/.match(data))
          md[1].strip.split(', ').each { |a| authors << CGI.unescape(a.strip) }
        end

        md = /<INPUT type =HIDDEN name ="mailEAN" value="([^"]+)/.match(data)
        raise 'No ISBN' unless md
        isbn = md[1].strip
        isbn += String(Library.ean_checksum(Library.extract_numbers(isbn)))

        # raise unless
        md = /<INPUT type =HIDDEN name ="mailEditore" value="([^"]+)/.match(data)
        publisher = CGI.unescape(md[1].strip) or md

        # raise unless
        md = /<INPUT type =HIDDEN name ="mailFormato" value="([^"]+)/.match(data)
        edition = CGI.unescape(md[1].strip) or md

        if (md = /#{edition}\&nbsp\;\|\&nbsp\;(\d+)\&nbsp\;\|\&nbsp\;/.match(data))
          nr_pages = CGI.unescape(md[1].strip)
        elsif (md = / (\d+) pagine \| /.match(data))
          nr_pages = CGI.unescape(md[1].strip)
        end
        if nr_pages != '0' and !nr_pages.nil?
          edition = nr_pages + ' p., ' + edition
        end

        publish_year = nil
        if (md = /<INPUT type =HIDDEN name ="mailAnnoPubbl" value="([^"]+)/.match(data))
          publish_year = CGI.unescape(md[1].strip).to_i
          publish_year = nil if publish_year == 0
        end

        cover_url = BASE_URI + '/bol/includes/tornaImmagine.jsp?cdSoc=BL&ean=' + isbn[0..11] + '&tipoOggetto=PIB&cdSito=BL' # use "FRB" instead of "PIB" for smaller images
        cover_filename = isbn + '.tmp'
        Dir.chdir(CACHE_DIR) do
          File.open(cover_filename, 'w') do |file|
            file.write open(cover_url, 'Referer' => REFERER).read
          end
        end

        medium_cover = CACHE_DIR + '/' + cover_filename
        if File.size(medium_cover) > 43 and File.size(medium_cover) != 2382 # 2382 is the size of the fake image "copertina non disponibile"
          puts medium_cover + ' has non-0 size' if $DEBUG
          return [Book.new(title, authors, isbn, publisher, publish_year, edition), medium_cover]
        end
        puts medium_cover + ' has 0 size, removing ...' if $DEBUG
        File.delete(medium_cover)
        [Book.new(title, authors, isbn, publisher, publish_year, edition)]
      end

      def each_book_page(data)
        raise if data.scan(/<a href="\/#{LOCALE}\/scheda\/ea(\d+)\.html;jsessionid=[^"]+">\s*Scheda completa\s*<\/a>/) { |a| yield a }.empty?
      end

      def clean_cache
        # FIXME begin ... rescue ... end?
        Dir.chdir(CACHE_DIR) do
          Dir.glob('*.tmp') do |file|
            puts 'removing ' + file if $DEBUG
            File.delete(file)
          end
        end
      end
    end
  end
end