Class: HTML::Proofer

Inherits:
Object
  • Object
show all
Includes:
Yell::Loggable
Defined in:
lib/html/proofer.rb,
lib/html/proofer/checks.rb,
lib/html/proofer/version.rb,
lib/html/proofer/checkable.rb

Defined Under Namespace

Classes: Checkable, Checks

Constant Summary collapse

VERSION =
"1.5.4"

Instance Attribute Summary collapse

Class Method Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize(src, opts = {}) ⇒ Proofer

Returns a new instance of Proofer.


32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
# File 'lib/html/proofer.rb', line 32

def initialize(src, opts={})
  @src = src

  @proofer_opts = {
    :ext => ".html",
    :favicon => false,
    :href_swap => [],
    :href_ignore => [],
    :check_external_hash => false,
    :alt_ignore => [],
    :disable_external => false,
    :verbose => false,
    :only_4xx => false,
    :directory_index_file => "index.html",
    :validate_html => false,
    :error_sort => :path
  }

  @typhoeus_opts = {
    :followlocation => true
  }

  # fall back to parallel defaults
  @parallel_opts = opts[:parallel] || {}
  opts.delete(:parallel)

  # Typhoeus won't let you pass in any non-Typhoeus option; if the option is not
  # a proofer_opt, it must be for Typhoeus
  opts.keys.each do |key|
    if @proofer_opts[key].nil?
      @typhoeus_opts[key] = opts[key]
    end
  end

  @options = @proofer_opts.merge(@typhoeus_opts).merge(opts)

  @failed_tests = []

  Yell.new({ :format => false, :name => "HTML::Proofer", :level => "gte.#{log_level}" }) do |l|
    l.adapter :stdout, level: [:debug, :info, :warn]
    l.adapter :stderr, level: [:error, :fatal]
  end
end

Instance Attribute Details

#optionsObject (readonly)

Returns the value of attribute options


30
31
32
# File 'lib/html/proofer.rb', line 30

def options
  @options
end

#parallel_optsObject (readonly)

Returns the value of attribute parallel_opts


30
31
32
# File 'lib/html/proofer.rb', line 30

def parallel_opts
  @parallel_opts
end

#typhoeus_optsObject (readonly)

Returns the value of attribute typhoeus_opts


30
31
32
# File 'lib/html/proofer.rb', line 30

def typhoeus_opts
  @typhoeus_opts
end

Class Method Details

.create_nokogiri(path) ⇒ Object


231
232
233
234
# File 'lib/html/proofer.rb', line 231

def self.create_nokogiri(path)
  content = File.open(path).read
  Nokogiri::HTML(content)
end

Instance Method Details

#add_failed_tests(filenames, desc, status = nil) ⇒ Object


255
256
257
258
259
260
261
262
263
# File 'lib/html/proofer.rb', line 255

def add_failed_tests(filenames, desc, status = nil)
  if filenames.nil?
    @failed_tests << Checks::Issue.new("", desc, status)
  elsif
    filenames.each { |f|
      @failed_tests << Checks::Issue.new(f, desc, status)
    }
  end
end

the hypothesis is that Proofer runs way faster if we pull out all the external URLs and run the checks at the end. Otherwise, we're halting the consuming process for every file. In addition, sorting the list lets libcurl keep connections to hosts alive. Finally, we'll make a HEAD request, rather than GETing all the contents


152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
# File 'lib/html/proofer.rb', line 152

def external_link_checker(external_urls)
  external_urls = Hash[external_urls.sort]

  logger.info HTML::colorize :yellow, "Checking #{external_urls.length} external links..."

  Ethon.logger = logger # log from Typhoeus/Ethon

  external_urls.each_pair do |href, filenames|
    if has_hash? href && @options[:check_external_hash]
      queue_request(:get, href, filenames)
    else
      queue_request(:head, href, filenames)
    end
  end
  logger.debug HTML::colorize :yellow, "Running requests for all #{hydra.queued_requests.size} external URLs..."
  hydra.run
end

#failed_testsObject


265
266
267
268
269
270
# File 'lib/html/proofer.rb', line 265

def failed_tests
  return [] if @failed_tests.empty?
  result = []
  @failed_tests.each { |f| result << f.to_s }
  result
end

#filesObject


220
221
222
223
224
225
226
227
228
229
# File 'lib/html/proofer.rb', line 220

def files
  if File.directory? @src
    pattern = File.join @src, "**", "*#{@options[:ext]}"
    Dir.glob(pattern).select { |fn| File.file? fn }
  elsif File.extname(@src) == @options[:ext]
    [@src]
  else
    []
  end
end

#get_checksObject


236
237
238
239
240
241
# File 'lib/html/proofer.rb', line 236

def get_checks
  checks = HTML::Proofer::Checks::Check.subclasses.map { |c| c.name }
  checks.delete("Favicons") unless @options[:favicon]
  checks.delete("Html") unless @options[:validate_html]
  checks
end

#has_hash?(url) ⇒ Boolean

Returns:

  • (Boolean)

243
244
245
246
247
248
249
# File 'lib/html/proofer.rb', line 243

def has_hash?(url)
  begin
    URI.parse(url).fragment
  rescue URI::InvalidURIError
    nil
  end
end

#hydraObject


216
217
218
# File 'lib/html/proofer.rb', line 216

def hydra
  @hydra ||= Typhoeus::Hydra.new
end

#log_levelObject


251
252
253
# File 'lib/html/proofer.rb', line 251

def log_level
  @options[:verbose] ? :debug : :info
end

#queue_request(method, href, filenames) ⇒ Object


170
171
172
173
174
# File 'lib/html/proofer.rb', line 170

def queue_request(method, href, filenames)
  request = Typhoeus::Request.new(href, @typhoeus_opts.merge({:method => method}))
  request.on_complete { |response| response_handler(response, filenames) }
  hydra.queue request
end

#response_handler(response, filenames) ⇒ Object


176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
# File 'lib/html/proofer.rb', line 176

def response_handler(response, filenames)
  effective_url = response.options[:effective_url]
  href = response.request.base_url
  method = response.request.options[:method]
  response_code = response.code

  debug_msg = "Received a #{response_code} for #{href}"
  debug_msg << " in #{filenames.join(' ')}" unless filenames.nil?
  logger.debug debug_msg

  if response_code.between?(200, 299)
    return if @options[:only_4xx]
    return unless @options[:check_external_hash]
    if hash = has_hash?(href)
      body_doc = Nokogiri::HTML(response.body)
      # user-content is a special addition by GitHub.
      xpath = %$//*[@name="#{hash}"]|//*[@id="#{hash}"]$
      xpath << %$|//*[@name="user-content-#{hash}"]|//*[@id="user-content-#{hash}"]$ if URI.parse(href).host.match(/github\.com/i)
      if body_doc.xpath(xpath).empty?
        add_failed_tests filenames, "External link #{href} failed: #{effective_url} exists, but the hash '#{hash}' does not", response_code
      end
    end
  elsif response.timed_out?
    return if @options[:only_4xx]
    add_failed_tests filenames, "External link #{href} failed: got a time out", response_code
  elsif (response_code == 405 || response_code == 420 || response_code == 503) && method == :head
    # 420s usually come from rate limiting; let's ignore the query and try just the path with a GET
    uri = URI(href)
    queue_request(:get, uri.scheme + "://" + uri.host + uri.path, filenames)
  # just be lazy; perform an explicit get request. some servers are apparently not configured to
  # intercept HTTP HEAD
  elsif method == :head
    queue_request(:get, effective_url, filenames)
  else
    return if @options[:only_4xx] && !response_code.between?(400, 499)
    # Received a non-successful http response.
    add_failed_tests filenames, "External link #{href} failed: #{response_code} #{response.return_message}", response_code
  end
end

#runObject


76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
# File 'lib/html/proofer.rb', line 76

def run
  unless @src.is_a? Array
    external_urls = {}

    logger.info HTML::colorize :white, "Running #{get_checks} checks on #{@src} on *#{@options[:ext]}... \n\n"

    results = Parallel.map(files, @parallel_opts) do |path|
      html = HTML::Proofer.create_nokogiri(path)
      result = {:external_urls => {}, :failed_tests => []}

      get_checks.each do |klass|
        logger.debug HTML::colorize :blue, "Checking #{klass.to_s.downcase} on #{path} ..."
        check =  Object.const_get(klass).new(@src, path, html, @options)
        check.run
        result[:external_urls].merge!(check.external_urls)
        result[:failed_tests].concat(check.issues) if check.issues.length > 0
      end
      result
    end

    results.each do |item|
      external_urls.merge!(item[:external_urls])
      @failed_tests.concat(item[:failed_tests])
    end

    external_link_checker(external_urls) unless @options[:disable_external]

    logger.info HTML::colorize :green, "Ran on #{files.length} files!\n\n"
  else
    external_urls = Hash[*@src.map{ |s| [s, nil] }.flatten]
    external_link_checker(external_urls) unless @options[:disable_external]
  end

  if @failed_tests.empty?
    logger.info HTML::colorize :green, "HTML-Proofer finished successfully."
  else
    matcher = nil

    # always sort by the actual option, then path, to ensure consistent alphabetical (by filename) results
    @failed_tests = @failed_tests.sort do |a,b|
      comp = (a.send(@options[:error_sort]) <=> b.send(@options[:error_sort]))
      comp.zero? ? (a.path <=> b.path) : comp
    end

    @failed_tests.each do |issue|
      case @options[:error_sort]
      when :path
        if matcher != issue.path
          logger.error HTML::colorize :blue, "- #{issue.path}"
          matcher = issue.path
        end
        logger.error HTML::colorize :red, "  *  #{issue.desc}"
      when :desc
        if matcher != issue.desc
          logger.error HTML::colorize :blue, "- #{issue.desc}"
          matcher = issue.desc
        end
        logger.error HTML::colorize :red, "  *  #{issue.path}"
      when :status
        if matcher != issue.status
          logger.error HTML::colorize :blue, "- #{issue.status}"
          matcher = issue.status
        end
        logger.error HTML::colorize :red, "  *  #{issue.to_s}"
      end
    end

    raise HTML::colorize :red, "HTML-Proofer found #{@failed_tests.length} failures!"
  end
end