html-proofer 2.5.2 → 3.0.0

data/lib/html-proofer.rb

@@ -0,0 +1,47 @@
+def require_all(path)
+  glob = File.join(File.dirname(__FILE__), path, '*.rb')
+  Dir[glob].each do |f|
+    require f
+  end
+end
+
+require_all 'html-proofer'
+require_all 'html-proofer/check'
+
+require 'parallel'
+require 'fileutils'
+
+begin
+  require 'awesome_print'
+rescue LoadError; end
+
+module HTMLProofer
+
+  def check_file(file, options = {})
+    fail ArgumentError unless file.is_a?(String)
+    options[:type] = :file
+    HTMLProofer::Runner.new(file, options)
+  end
+  module_function :check_file
+
+  def check_directory(directory, options = {})
+    fail ArgumentError unless directory.is_a?(String)
+    options[:type] = :directory
+    HTMLProofer::Runner.new([directory], options)
+  end
+  module_function :check_directory
+
+  def check_directories(directories, options = {})
+    fail ArgumentError unless directories.is_a?(Array)
+    options[:type] = :directory
+    HTMLProofer::Runner.new(directories, options)
+  end
+  module_function :check_directories
+
+  def check_links(links, options = {})
+    fail ArgumentError unless links.is_a?(Array)
+    options[:type] = :links
+    HTMLProofer::Runner.new(links, options)
+  end
+  module_function :check_links
+end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: html-proofer
 version: !ruby/object:Gem::Version
-  version: 2.5.2
+  version: 3.0.0
 platform: ruby
 authors:
 - Garen Torikian
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2015-11-06 00:00:00.000000000 Z
+date: 2016-03-05 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: mercenary
@@ -108,6 +108,20 @@ dependencies:
     - - "~>"
       - !ruby/object:Gem::Version
         version: '2.3'
+- !ruby/object:Gem::Dependency
+  name: activesupport
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '4.2'
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '4.2'
 - !ruby/object:Gem::Dependency
   name: redcarpet
   requirement: !ruby/object:Gem::Requirement
@@ -178,30 +192,45 @@ dependencies:
     - - "~>"
       - !ruby/object:Gem::Version
         version: '2.9'
+- !ruby/object:Gem::Dependency
+  name: timecop
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '0.8'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '0.8'
 description: Test your rendered HTML files to make sure they're accurate.
 email:
 - gjtorikian@gmail.com
 executables:
-- htmlproof
+- htmlproofer
 extensions: []
 extra_rdoc_files: []
 files:
-- bin/htmlproof
-- lib/html/proofer.rb
-- lib/html/proofer/cache.rb
-- lib/html/proofer/check_runner.rb
-- lib/html/proofer/check_runner/issue.rb
-- lib/html/proofer/checkable.rb
-- lib/html/proofer/checks/favicon.rb
-- lib/html/proofer/checks/html.rb
-- lib/html/proofer/checks/images.rb
-- lib/html/proofer/checks/links.rb
-- lib/html/proofer/checks/scripts.rb
-- lib/html/proofer/log.rb
-- lib/html/proofer/url_validator.rb
-- lib/html/proofer/utils.rb
-- lib/html/proofer/version.rb
-- lib/html/proofer/xpathfunctions.rb
+- bin/htmlproofer
+- lib/html-proofer.rb
+- lib/html-proofer/cache.rb
+- lib/html-proofer/check.rb
+- lib/html-proofer/check/favicon.rb
+- lib/html-proofer/check/html.rb
+- lib/html-proofer/check/images.rb
+- lib/html-proofer/check/links.rb
+- lib/html-proofer/check/scripts.rb
+- lib/html-proofer/configuration.rb
+- lib/html-proofer/element.rb
+- lib/html-proofer/issue.rb
+- lib/html-proofer/log.rb
+- lib/html-proofer/runner.rb
+- lib/html-proofer/url_validator.rb
+- lib/html-proofer/utils.rb
+- lib/html-proofer/version.rb
 homepage: https://github.com/gjtorikian/html-proofer
 licenses:
 - MIT
@@ -222,7 +251,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
       version: '0'
 requirements: []
 rubyforge_project: 
-rubygems_version: 2.4.5
+rubygems_version: 2.4.5.1
 signing_key: 
 specification_version: 4
 summary: A set of tests to validate your HTML output. These tests check if your image

data/bin/htmlproof

@@ -1,85 +0,0 @@
-#!/usr/bin/env ruby
-STDOUT.sync = true
-
-$LOAD_PATH.unshift File.join(File.dirname(__FILE__), *%w( .. lib ))
-
-require 'html/proofer'
-require 'mercenary'
-require 'rubygems'
-
-def to_regex?(item)
-  if item.start_with?('/') && item.end_with?('/')
-    Regexp.new item[1...-1]
-  else
-    item
-  end
-end
-
-Mercenary.program(:htmlproof) do |p|
-  p.version HTML::Proofer::VERSION
-  p.description %(Test your rendered HTML files to make sure they're accurate.)
-  p.syntax 'htmlproof PATH [options]'
-
-  p.description 'Runs the HTML-Proofer suite on the files in PATH. For more details, see the README.'
-
-  p.option 'as_links', '--as-links', 'Assumes that `PATH` is a comma-separated array of links to check.'
-  p.option 'alt_ignore', '--alt-ignore image1,[image2,...]', Array, 'Comma-separated list of Strings or RegExps containing `img`s whose missing `alt` tags are safe to ignore'
-  p.option 'empty_alt_ignore', '--empty-alt-ignore', 'Ignores images with empty alt tags.'
-  p.option 'checks_to_ignore', '--checks-to-ignore check1,[check2,...]', Array, ' An array of Strings indicating which checks you\'d like to not perform.'
-  p.option 'check_external_hash', '--check-external-hash', 'Checks whether external hashes exist (even if the website exists). This slows the checker down (default: `false`).'
-  p.option 'check_favicon', '--check-favicon', 'Enables the favicon checker (default: `false`).'
-  p.option 'check_html', '--check-html', 'Enables HTML validation errors from Nokogiri (default: `false`).'
-  p.option 'directory_index_file', '--directory-index-file', String, 'Sets the file to look for when a link refers to a directory. (default: `index.html`)'
-  p.option 'disable_external', '--disable-external', 'Disables the external link checker (default: `false`)'
-  p.option 'error_sort', '--error-sort SORT', 'Defines the sort order for error output. Can be `path`, `desc`, or `status` (default: `path`).'
-  p.option 'enforce_https', '--enforce-https', 'Fails a link if it\'s not marked as `https` (default: `false`).'
-  p.option 'ext', '--ext EXT', String, 'The extension of your HTML files (default: `.html`)'
-  p.option 'file_ignore', '--file-ignore file1,[file2,...]', Array, 'Comma-separated list of Strings or RegExps containing file paths that are safe to ignore'
-  p.option 'href_ignore', '--href-ignore link1,[link2,...]', Array, 'Comma-separated list of Strings or RegExps containing `href`s that are safe to ignore.'
-  p.option 'href_swap', '--href-swap re:string,[re:string,...]', Array, 'Comma-separated list of key-value pairs of `RegExp:String`. Transforms links matching `RegExp` into `String`'
-  p.option 'ignore_script_embeds', '--ignore-script-embeds', 'Ignore `check_html` errors associated with `script`s (default: `false`)'
-  p.option 'only_4xx', '--only-4xx', 'Only reports errors for links that fall within the 4x status code range.'
-  p.option 'url_ignore', '--url-ignore link1,[link2,...]', Array, 'Comma-separated list of Strings or RegExps containing URLs that are safe to ignore.'
-  p.option 'verbose', '--verbose', 'Enables more verbose logging.'
-  p.option 'verbosity', '--verbosity', String, 'Sets the logging level, as determined by Yell'
-
-  p.action do |args, opts|
-    args = ['.'] if args.empty?
-    path = args.first
-
-    options = {}
-
-    # prepare everything to go to proofer
-    p.options.select { |o| !opts[o.config_key].nil? }.each do |option|
-      if option.return_type.to_s == 'Array' # TODO: is_a? doesn't work here?
-        opts[option.config_key] = opts[option.config_key].map { |i| to_regex?(i) }
-      end
-      options[option.config_key.to_sym] = opts[option.config_key]
-    end
-
-    # some minor manipulation of a special option
-    unless opts['href_swap'].nil?
-      options[:href_swap] = {}
-      opts['href_swap'].each do |s|
-        pair = s.split(':', 2)
-        options[:href_swap][Regexp.new(pair[0])] = pair[1]
-      end
-    end
-
-    # check for ignore_scripts_embeds as it should be set in :validation
-    unless opts['ignore_script_embeds'].nil?
-        options[:validation] = { :ignore_script_embeds => true }
-    end
-
-    options[:error_sort] = opts['error-sort'].to_sym unless opts['error-sort'].nil?
-    options[:verbosity] = opts['verbosity'].to_sym unless opts['verbosity'].nil?
-
-    # FIXME: this is gross
-    options[:validation] = {}
-    options[:validation][:ignore_script_embeds] = opts['ignore_script_embeds']
-
-    path = path.delete(' ').split(',') if opts['as_links']
-
-    HTML::Proofer.new(path, options).run
-  end
-end

data/lib/html/proofer/cache.rb

@@ -1,16 +0,0 @@
-module HTML
-  class Proofer
-    module Cache
-      def create_nokogiri(path)
-        if File.exist? path
-          content = File.open(path).read
-        else
-          content = path
-        end
-
-        Nokogiri::HTML(content)
-      end
-      module_function :create_nokogiri
-    end
-  end
-end

data/lib/html/proofer/check_runner.rb

@@ -1,69 +0,0 @@
-# encoding: utf-8
-
-module HTML
-  class Proofer
-    # Mostly handles issue management and collecting of external URLs.
-    class CheckRunner
-
-      attr_reader :issues, :src, :path, :options, :typhoeus_opts, :hydra_opts, :parallel_opts, \
-                  :validation_opts, :external_urls, :href_ignores, :url_ignores, :alt_ignores, \
-                  :empty_alt_ignore
-
-      def initialize(src, path, html, options, typhoeus_opts, hydra_opts, parallel_opts, validation_opts)
-        @src    = src
-        @path   = path
-        @html   = remove_ignored(html)
-        @options = options
-        @typhoeus_opts = typhoeus_opts
-        @hydra_opts = hydra_opts
-        @parallel_opts = parallel_opts
-        @validation_opts = validation_opts
-        @issues = []
-        @href_ignores = @options[:href_ignore]
-        @url_ignores = @options[:url_ignore]
-        @alt_ignores = @options[:alt_ignore]
-        @empty_alt_ignore = @options[:empty_alt_ignore]
-        @external_urls = {}
-      end
-
-      def run
-        fail NotImplementedError, 'HTML::Proofer::CheckRunner subclasses must implement #run'
-      end
-
-      def add_issue(desc, line_number = nil, status = -1)
-        @issues << Issue.new(@path, desc, line_number, status)
-      end
-
-      def add_to_external_urls(url, line)
-        return if @external_urls[url]
-        add_path_for_url(url)
-      end
-
-      def add_path_for_url(url)
-        if @external_urls[url]
-          @external_urls[url] << @path
-        else
-          @external_urls[url] = [@path]
-        end
-      end
-
-      def self.checks
-        classes = []
-
-        ObjectSpace.each_object(Class) do |c|
-          next unless c.superclass == self
-          classes << c
-        end
-
-        classes
-      end
-
-      private
-
-      def remove_ignored(html)
-        html.css('code, pre, tt').each(&:unlink)
-        html
-      end
-    end
-  end
-end

data/lib/html/proofer/checkable.rb

@@ -1,164 +0,0 @@
-require 'addressable/uri'
-require_relative './utils'
-
-module HTML
-  class Proofer
-    # Represents the superclass from which all checks derive.
-    class Checkable
-      include HTML::Proofer::Utils
-
-      attr_reader :line
-
-      def initialize(obj, check)
-        obj.attributes.each_pair do |attribute, value|
-          instance_variable_set("@#{attribute.tr('-:.', '_')}".to_sym, value.value)
-        end
-
-        @text = obj.content
-        @check = check
-        @checked_paths = {}
-        @type = self.class.name
-        @line = obj.line
-
-        if @href && @check.options[:href_swap]
-          @href = swap(@href, @check.options[:href_swap])
-        end
-
-        # fix up missing protocols
-        @href.insert 0, 'http:' if @href =~ %r{^//}
-        @src.insert 0, 'http:' if @src =~ %r{^//}
-      end
-
-      def url
-        @src || @srcset || @href || ''
-      end
-
-      def valid?
-        !parts.nil?
-      end
-
-      def parts
-        @parts ||= Addressable::URI.parse url
-      rescue URI::Error, Addressable::URI::InvalidURIError
-        @parts = nil
-      end
-
-      def path
-        Addressable::URI.unencode parts.path unless parts.nil?
-      end
-
-      def hash
-        parts.fragment unless parts.nil?
-      end
-
-      def scheme
-        parts.scheme unless parts.nil?
-      end
-
-      # path is to an external server
-      def remote?
-        %w( http https ).include? scheme
-      end
-
-      def non_http_remote?
-        !scheme.nil? && !remote?
-      end
-
-      def ignore?
-        return true if @data_proofer_ignore
-
-        # ignore base64 encoded images
-        if %w(ImageCheckable FaviconCheckable).include? @type
-          return true if url.match(/^data:image/)
-        end
-
-        # ignore user defined URLs
-        return true if ignores_pattern_check(@check.url_ignores)
-
-        # ignore user defined hrefs
-        if 'LinkCheckable' == @type
-          return true if ignores_pattern_check(@check.href_ignores)
-        end
-
-        # ignore user defined alts
-        if 'ImageCheckable' == @type
-          return true if ignores_pattern_check(@check.alt_ignores)
-        end
-      end
-
-      def ignore_empty_alt?
-        @check.empty_alt_ignore
-      end
-
-      # path is external to the file
-      def external?
-        !internal?
-      end
-
-      # path is an anchor or a query
-      def internal?
-        url.start_with? '#', '?'
-      end
-
-      def file_path
-        return if path.nil?
-
-        if path =~ %r{^/} # path relative to root
-          base = File.directory?(@check.src) ? @check.src : File.dirname(@check.src)
-        elsif File.exist?(File.expand_path path, @check.src) # relative links, path is a file
-          base = File.dirname @check.path
-        elsif File.exist?(File.join(File.dirname(@check.path), path)) # relative links in nested dir, path is a file
-          base = File.dirname @check.path
-        else # relative link, path is a directory
-          base = @check.path
-        end
-
-        file = File.join base, path
-
-        # implicit index support
-        if File.directory?(file) && !unslashed_directory?(file)
-          file = File.join file, @check.options[:directory_index_file]
-        end
-
-        file
-      end
-
-      # checks if a file exists relative to the current pwd
-      def exists?
-        return @checked_paths[absolute_path] if @checked_paths.key? absolute_path
-        @checked_paths[absolute_path] = File.exist? absolute_path
-      end
-
-      def absolute_path
-        path = file_path || @check.path
-        File.expand_path path, Dir.pwd
-      end
-
-      def ignores_pattern_check(links)
-        links.each do |ignore|
-          if ignore.is_a? String
-            return true if ignore == url
-          elsif ignore.is_a? Regexp
-            return true if ignore =~ url
-          end
-        end
-
-        false
-      end
-
-      def unslashed_directory?(file)
-        File.directory?(file) && !file.end_with?(File::SEPARATOR) && !follow_location?
-      end
-
-      def follow_location?
-        @check.typhoeus_opts && @check.typhoeus_opts[:followlocation]
-      end
-
-      private
-
-      def real_attr(attr)
-        attr.to_s unless attr.nil? || attr.empty?
-      end
-    end
-  end
-end

data/lib/html/proofer/checks/html.rb

@@ -1,46 +0,0 @@
-# encoding: utf-8
-
-class HtmlCheck < ::HTML::Proofer::CheckRunner
-  # new html5 tags (source: http://www.w3schools.com/html/html5_new_elements.asp)
-  # and svg child tags (source: https://developer.mozilla.org/en-US/docs/Web/SVG/Element)
-  HTML5_TAGS = %w(article aside bdi details dialog figcaption
-                  figure footer header main mark menuitem meter
-                  nav progress rp rt ruby section summary
-                  time wbr datalist keygen output color date
-                  datetime datetime-local email month number
-                  range search tel time url week canvas
-                  svg audio embed source track video
-                  altGlyph altGlyphDef altGlyphItem animate
-                  animateColor animateMotion animateTransform
-                  circle clipPath color-profile cursor defs
-                  desc ellipse feBlend feColorMatrix
-                  feComponentTransfer feComposite feConvolveMatrix
-                  feDiffuseLighting feDisplacementMap feDistantLight
-                  feFlood feFuncA feFuncB feFuncG feFuncR feGaussianBlur
-                  feImage feMerge feMergeNode feMorphology feOffset
-                  fePointLight feSpecularLighting feSpotLight feTile
-                  feTurbulence filter font font-face font-face-format
-                  font-face-name font-face-src font-face-uri
-                  foreignObject g glyph glyphRef hkern image line
-                  linearGradient marker mask metadata missing-glyph
-                  mpath path pattern polygon polyline radialGradient
-                  rect set stop switch symbol text textPath tref tspan use
-                  view vkern)
-
-  SCRIPT_EMBEDS_MSG = /Element script embeds close tag/
-
-  def run
-    @html.errors.each do |error|
-      message = error.message
-      line    = error.line
-      # Nokogiri (or rather libxml2 underhood) only recognizes html4 tags,
-      # so we need to skip errors caused by the new tags in html5
-      next if HTML5_TAGS.include? message[/Tag ([\w-]+) invalid/o, 1]
-
-      # tags embedded in scripts are used in templating languages: http://git.io/vOovv
-      next if @validation_opts[:ignore_script_embeds] && message =~ SCRIPT_EMBEDS_MSG
-
-      add_issue(message, line)
-    end
-  end
-end

data/lib/html/proofer/checks/images.rb

@@ -1,54 +0,0 @@
-# encoding: utf-8
-
-class ImageCheckable < ::HTML::Proofer::Checkable
-  SCREEN_SHOT_REGEX = /Screen(?: |%20)Shot(?: |%20)\d+-\d+-\d+(?: |%20)at(?: |%20)\d+.\d+.\d+/
-
-  attr_reader :alt
-
-  def empty_alt_tag?
-    alt.strip.empty?
-  end
-
-  def terrible_filename?
-    src =~ SCREEN_SHOT_REGEX
-  end
-
-  def src
-    real_attr(@src) || real_attr(@srcset)
-  end
-
-  def missing_src?
-    !src
-  end
-end
-
-class ImageCheck < ::HTML::Proofer::CheckRunner
-  def run
-    @html.css('img').each do |node|
-      img = ImageCheckable.new(node, self)
-      line = node.line
-
-      next if img.ignore?
-
-      # screenshot filenames should return because of terrible names
-      next add_issue("image has a terrible filename (#{img.src})", line) if img.terrible_filename?
-
-      # does the image exist?
-      if img.missing_src?
-        add_issue('image has no src or srcset attribute', line)
-      else
-        if img.remote?
-          add_to_external_urls(img.src, line)
-        else
-          add_issue("internal image #{img.src} does not exist", line) unless img.exists?
-        end
-      end
-
-      if img.alt.nil? || (img.empty_alt_tag? && !img.ignore_empty_alt?)
-        add_issue("image #{img.src} does not have an alt attribute", line)
-      end
-    end
-
-    external_urls
-  end
-end