site_health.rb in site_health-0.2.0

- old
+ new
@@ -1,136 +1,112 @@
-require "spidr"
-require 'w3c_validators'
-require "site_health/version"
+# frozen_string_literal: true
 
-require "site_health/key_struct"
+require 'logger'
+require 'spidr'
 
-require 'site_health/journals/css_journal'
-require 'site_health/journals/html_journal'
-require 'site_health/journals/xml_journal'
-require 'site_health/journals/w3c_journal'
+require 'site_health/version'
+require 'site_health/configuration/configuration'
 
-require "site_health/checkers/css_page"
-require "site_health/checkers/html_page"
-require "site_health/checkers/xml_page"
+require 'site_health/key_struct'
+require 'site_health/url_map'
+require 'site_health/link'
 
+require 'site_health/checkers/checker'
+require 'site_health/nurse'
+require 'site_health/issues_report'
+
+# Top-level module/namespace
 module SiteHealth
-  def self.check(site)
-    Check.call(site: site)
+  def self.require_optional_dependency(path, gem_name: nil)
+    gem_name ||= path
+    require path
+  rescue LoadError => e
+    message_parts = [
+      e.message,
+      "unable to require file from '#{gem_name}' gem",
+      'please install it',
+    ]
+    raise(LoadError, message_parts.join(' -- '))
   end
 
-  class Check
-    def self.call(**args)
-      new(**args).call
-    end
+  # @param [Checker] klass that inherits from Checker
+  # @return [see SiteHealth#registered_checkers]
+  def self.register_checker(klass)
+    registered_checkers[klass.name.to_sym] = klass
+    registered_checkers
+  end
 
-    BrokenLinkJournal = KeyStruct.new(:url, :exists_on)
+  # @return [Hash] all registered checkers
+  def self.registered_checkers
+    @checkers ||= {}
+  end
 
-    HTTPCodeJournal = KeyStruct.new(:url, :code)
-    class HTTPCodeJournal
-      def error?
-        code >= 400
-      end
+  # @param [Symbol, String] name of the checker to be loaded
+  # @return [Checker] loaded class that should inherits from Checker
+  def self.load_checker(name)
+    name_key = name.to_sym
+    registered_checkers.fetch(name_key) do
+      require "site_health/checkers/#{name}"
+      registered_checkers[name_key]
     end
+  end
 
-    ChecksJournal = KeyStruct.new(
-      :missing_html_title,
-      :broken_urls,
-      :http_error_urls,
-      :html_error_urls,
-      :html_warning_urls,
-      :xml_error_urls,
-      :css_error_urls,
-      :css_warning_urls
-    )
+  # @param [String] site to be checked
+  # @param config [SiteHealth::Configuration] the configuration to use
+  # @yieldparam [SiteHealth::Nurse] nurse (a.k.a agent)
+  # @return [Hash] journal data
+  # @see Nurse#journal
+  def self.check(site, config: SiteHealth.config)
+    nurse = Nurse.new(config: config)
+    yield(nurse) if block_given?
 
-    attr_reader :site
-
-    def initialize(site:)
-      @site = site
+    Spidr.site(site) do |spider|
+      spider.every_failed_url { |url| nurse.check_failed_url(url) }
+      spider.every_page { |page| nurse.check_page(page) }
     end
 
-    def call
-      url_map = Hash.new { |hash, key| hash[key] = [] }
+    nurse.punch_out!
+  end
 
-      missing_html_title = []
-      http_error_urls = []
-      html_error_urls = []
-      html_warning_urls = []
-      xml_error_urls = []
-      css_error_urls = []
-      css_warning_urls = []
+  # @param [Array<String>, String] urls to be checked
+  # @param config [SiteHealth::Configuration] the configuration to use
+  # @yieldparam [SiteHealth::Nurse] nurse (a.k.a agent)
+  # @return [Hash] journal data
+  # @see Nurse#journal
+  def self.check_urls(urls, config: SiteHealth.config)
+    nurse = Nurse.new(config: config)
+    yield(nurse) if block_given?
 
-      spider = Spidr.site(site) do |spider|
-        spider.every_link do |origin, destination|
-          url_map[destination] << origin
-        end
+    agent = Spidr::Agent.new
 
-        spider.every_page do |page|
-          code_journal = HTTPCodeJournal.new(url: page.url, code: page.code)
-          http_error_urls << code_journal if code_journal.error?
+    Array(urls).each do |url|
+      page = agent.get_page(url)
 
-          if page.css?
-            result = Checkers::CSSPage.check(page)
-            xml_error_urls << result if result.errors?
-          end
-
-          if page.xml?
-            result = Checkers::XMLPage.check(page)
-            xml_error_urls << result if result.errors?
-          end
-
-          if page.html?
-            result = Checkers::HTMLPage.check(page)
-            missing_html_title << result if result.missing_title?
-            html_error_urls << result if result.errors?
-          end
-        end
+      if page.nil?
+        nurse.check_failed_url(url)
+        next
       end
 
-      http_error_urls = map_http_error_urls(http_error_urls, url_map)
-      broken_urls = broken_links(spider, url_map) + http_error_urls
-
-      ChecksJournal.new(
-        missing_html_title: missing_html_title,
-        broken_urls: broken_urls,
-        http_error_urls: http_error_urls,
-        html_error_urls: html_error_urls,
-        html_warning_urls: html_warning_urls,
-        xml_error_urls: xml_error_urls,
-        css_error_urls: css_error_urls,
-        css_warning_urls: css_warning_urls
-      )
+      nurse.check_page(page)
     end
 
-    def validate_css_page(page, errors)
-      css_checker = Checkers::CSSPage.new(page)
-      result = css_checker.check
-      return unless result.errors?
+    nurse.punch_out!
+  end
 
-      result
-    end
+  # @see Configuration#logger
+  def self.logger
+    config.logger
+  end
 
-    def map_http_error_urls(urls, url_map)
-      urls.map do |failed_url|
-        BrokenLinkJournal.new(url: failed_url, exists_on: url_map[failed_url])
-      end
-    end
+  # @return [Configuration] the current configuration
+  # @yieldparam [Configuration] the current configuration
+  def self.configure
+    @configuration ||= Configuration.new
+    yield(@configuration) if block_given?
+    @configuration
+  end
 
-    # Finds all pages which have broken links:
-    def broken_links(spider, url_map)
-      # FIXME: spider#failures only returns timeout errors etc and not HTTP error status codes..
-      #        so we need to have 2 types of "failed" URLs
-      spider.failures.map do |failed_url|
-        BrokenLinkJournal.new(url: failed_url, exists_on: url_map[failed_url])
-      end
-    end
-
-    # @return [W3CValidators::Results]
-    # @raise [W3CValidators::ValidatorUnavailable] the service is offline or returns 400 Bad Request
-    # @see https://github.com/w3c-validators/w3c_validators/issues/39 we really want to use #validate_text instead of #validate_uri but due to the linked issue thats not possible
-    def validate_html(html_url)
-      validator = W3CValidators::NuValidator.new
-      validator.validate_uri(html_url)
-    end
+  # @return [Configuration] the current configuration
+  def self.config
+    configure
   end
 end