カメラバカにつける薬 in デジカメ Watch を PDF 化する

camera-baka2pdf.rb

#!/usr/bin/env ruby

# requirement
# - nokogiri (rubygems)
# - img2pdf
# - pngquant (optional)

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

ROOT_PATH = Pathname.new(__FILE__).parent
IMAGE_CACHE_PATH = ROOT_PATH.join('image_caches')
IMAGE_CACHE_PATH.mkpath

INDEX_RANGE = 2017..(Time.now.year)

SLEEP_WAIT = 1

def get_entry_urls(index_range)
  path_re = %r{dc.watch.impress.co.jp/docs/comic/clinic/\d+.html}

  index_range.map do |year|
    index_url = "https://dc.watch.impress.co.jp/docs/comic/clinic/index#{year}.html"
    body = open(index_url).read
    links = Nokogiri(body).css('a').map {|e| e.attr('href') }
    links.find_all {|url| url.match(path_re) }
  end.flatten
end

entry_urls = get_entry_urls(INDEX_RANGE).flatten.sort.uniq

entry_urls.each do |url|
  puts "process: #{url}"
  _, path1, path2 = *url.match(/(\d{4})(\d{3})\.html/)
  (1..99).each do |i|
    cache_path = IMAGE_CACHE_PATH.join "#{path1}_#{path2}_%02d.png" % i
    image_url = "https://dc.watch.impress.co.jp/img/dcw/docs/#{path1}/#{path2}/%02d.png" % i
    if cache_path.exist?
      puts "cache found!"
      break
    else
      puts "get image: #{image_url}"
      begin
        image = open(image_url).read
        cache_path.open('w') {|f| f.puts image }
      rescue OpenURI::HTTPError => e
        warn "image can't load, next"
        break
      end
      sleep SLEEP_WAIT
    end
  end
end

image_files = Pathname.glob('./image_caches/*.png').map(&:to_s).sort

pngquant_cmd_base = %w(pngquant --ext .png --force)
pngquant_cmd = [pngquant_cmd_base + image_files].flatten
puts pngquant_cmd_base.join(' ') + " [...#{image_files.size} png files]"
system *pngquant_cmd

pdfname = Time.now.strftime('camera-baka-%Y_%m%d_%H%M%S') + ".pdf"
pdf_cmd_base = ['img2pdf', '--output', "./dist/#{pdfname}"]
pdf_cmd = [pdf_cmd_base + image_files].flatten
puts pdf_cmd_base.join(' ') + " [...#{image_files.size} png files]"
system *pdf_cmd