# frozen_string_literal: true class FetchLinkCardService < BaseService URL_PATTERN = %r{ ( # $1 URL (https?:\/\/) # $2 Protocol (required) (#{Twitter::Regex[:valid_domain]}) # $3 Domain(s) (?::(#{Twitter::Regex[:valid_port_number]}))? # $4 Port number (optional) (/#{Twitter::Regex[:valid_url_path]}*)? # $5 URL Path and anchor (\?#{Twitter::Regex[:valid_url_query_chars]}*#{Twitter::Regex[:valid_url_query_ending_chars]})? # $6 Query String ) }iox def call(status) @status = status @url = parse_urls if @status.preview_cards.any? if @url.nil? detach_card return end return if @status.preview_cards.first.url == @url end return if @url.nil? @url = @url.to_s RedisLock.acquire(lock_options) do |lock| if lock.acquired? @card = PreviewCard.find_by(url: @url) process_url if @card.nil? || @card.updated_at <= 12.hours.ago || @card.missing_image? else raise GabSocial::RaceConditionError end end attach_card if @card&.persisted? rescue HTTP::Error, Addressable::URI::InvalidURIError, GabSocial::HostValidationError, GabSocial::LengthValidationError => e Rails.logger.debug "Error fetching link #{@url}: #{e}" nil end private def process_url @card ||= PreviewCard.new(url: @url) Request.new(:get, @url).add_headers('Accept' => 'text/html').perform do |res| if res.code == 200 && res.mime_type == 'text/html' @html = res.body_with_limit @html_charset = res.charset else @html = nil @html_charset = nil end end if @html.nil? detach_card return end attempt_oembed || attempt_opengraph end def attach_card @status.preview_cards = [@card] send_status_update_payload(@status) Rails.cache.delete(@status) end def detach_card @status.preview_cards = [] send_status_update_payload(@status) Rails.cache.delete(@status) end def parse_urls if @status.local? urls = @status.text.scan(URL_PATTERN).map { |array| Addressable::URI.parse(array[0]).normalize } else html = Nokogiri::HTML(@status.text) links = html.css('a') urls = links.map { |a| Addressable::URI.parse(a['href']).normalize unless skip_link?(a) }.compact end urls.reject { |uri| bad_url?(uri) }.first end def bad_url?(uri) # Avoid invalid URLs uri.host.blank? || !%w(http https).include?(uri.scheme) end def mention_link?(a) @status.mentions.any? do |mention| a['href'] == TagManager.instance.url_for(mention.account) end end def skip_link?(a) # Avoid links for hashtags and mentions (microformats) a['rel']&.include?('tag') || a['class']&.include?('u-url') || mention_link?(a) end def attempt_oembed service = FetchOEmbedService.new embed = service.call(@url, html: @html) url = Addressable::URI.parse(service.endpoint_url) return false if embed.nil? @card.type = embed[:type] @card.title = embed[:title] || '' @card.author_name = embed[:author_name] || '' @card.author_url = embed[:author_url].present? ? (url + embed[:author_url]).to_s : '' @card.provider_name = embed[:provider_name] || '' @card.provider_url = embed[:provider_url].present? ? (url + embed[:provider_url]).to_s : '' @card.width = 0 @card.height = 0 case @card.type when 'link' @card.image_remote_url = (url + embed[:thumbnail_url]).to_s if embed[:thumbnail_url].present? when 'photo' return false if embed[:url].blank? @card.embed_url = (url + embed[:url]).to_s @card.image_remote_url = (url + embed[:url]).to_s @card.width = embed[:width].presence || 0 @card.height = embed[:height].presence || 0 when 'video' @card.width = embed[:width].presence || 0 @card.height = embed[:height].presence || 0 @card.html = Formatter.instance.sanitize(embed[:html], Sanitize::Config::GABSOCIAL_OEMBED) @card.image_remote_url = (url + embed[:thumbnail_url]).to_s if embed[:thumbnail_url].present? when 'rich' # Most providers rely on