#!/usr/bin/env ruby
# frozen_string_literal: true

$LOAD_PATH.unshift File.expand_path('../lib', __dir__)
require 'broken_link_finder'
require 'thor'

class BrokenLinkFinderCLI < Thor
  desc 'crawl [URL]', 'Find broken links at the URL'
  option :recursive, type: :boolean, aliases: [:r], default: false, desc: 'Crawl the entire site.'
  option :threads, type: :numeric, aliases: [:t], default: BrokenLinkFinder::DEFAULT_MAX_THREADS, desc: 'Max number of threads to use when crawling recursively; 1 thread per web page.'
  option :xpath, type: :string, aliases: [:x], default: BrokenLinkFinder::DEFAULT_LINK_XPATH, desc: 'The xpath to extract links with, before checking if broken'
  option :js, type: :boolean, default: false, desc: 'Run the Javascript on a page before crawling the HTML, requires Chrome/Chromium to be installed to $PATH'
  option :js_delay, type: :numeric, default: 1, desc: "The seconds of delay time given to a page's Javascript for it to update the DOM, requires the --js flag"
  option :html, type: :boolean, aliases: [:h], default: false, desc: 'Produce a HTML report (instead of text)'
  option :sort_by_link, type: :boolean, aliases: [:l], default: false, desc: 'Makes report more concise if there are more pages crawled than broken links found. Use with -r on medium/large sites.'
  option :verbose, type: :boolean, aliases: [:v], default: false, desc: 'Display all ignored links.'
  option :concise, type: :boolean, aliases: [:c], default: false, desc: 'Display only a summary of broken links.'
  def crawl(url)
    url = "http://#{url}" unless url.start_with?('http')

    report_type     = options[:html] ? :html : :text
    sort_by         = options[:sort_by_link] ? :link : :page
    max_threads     = options[:threads]
    broken_verbose  = !options[:concise]
    ignored_verbose = options[:verbose]
    parse_js        = options[:js]
    parse_js_delay  = options[:js_delay]

    BrokenLinkFinder.link_xpath = options[:xpath]
    finder = BrokenLinkFinder::Finder.new(sort: sort_by, max_threads:) do |crawler|
      crawler.parse_javascript = parse_js
      crawler.parse_javascript_delay = parse_js_delay
    end

    options[:recursive] ? finder.crawl_site(url) : finder.crawl_page(url)
    finder.report(type: report_type, broken_verbose:, ignored_verbose:)

    exit 0
  rescue StandardError => e
    puts "An error has occurred: #{e.message}"

    exit 1
  end

  desc 'version', 'Display the currently installed version'
  def version
    puts "broken_link_finder v#{BrokenLinkFinder::VERSION}"

    exit 0
  end
end

BrokenLinkFinderCLI.start(ARGV)