Class: SiteDiff

Inherits:
Object
  • Object
show all
Defined in:
lib/sitediff.rb,
lib/sitediff/cli.rb,
lib/sitediff/diff.rb,
lib/sitediff/cache.rb,
lib/sitediff/fetch.rb,
lib/sitediff/rules.rb,
lib/sitediff/config.rb,
lib/sitediff/result.rb,
lib/sitediff/crawler.rb,
lib/sitediff/sanitize.rb,
lib/sitediff/exception.rb,
lib/sitediff/webserver.rb,
lib/sitediff/uriwrapper.rb,
lib/sitediff/config/creator.rb,
lib/sitediff/sanitize/regexp.rb,
lib/sitediff/sanitize/dom_transform.rb,
lib/sitediff/webserver/resultserver.rb

Defined Under Namespace

Modules: Diff Classes: Cache, Cli, Config, Crawler, Fetch, Result, Rules, Sanitizer, SiteDiffException, SiteDiffReadFailure, UriWrapper, Webserver

Constant Summary collapse

FILES_DIR =

path to misc. static files (e.g. erb, css files)

File.join(File.dirname(__FILE__), 'sitediff', 'files')
DIFFS_DIR =

subdirectory containing all failing diffs

'diffs'
FAILURES_FILE =

files in output

'failures.txt'
REPORT_FILE = 
'report.html'
SETTINGS_FILE = 
'settings.yaml'

Instance Attribute Summary collapse

Class Method Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize(config, cache, concurrency, interval, verbose = true, debug = false) ⇒ SiteDiff

Returns a new instance of SiteDiff.



57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
 
# File 'lib/sitediff.rb', line 57

def initialize(config, cache, concurrency, interval, verbose = true, debug = false)
  @cache = cache
  @verbose = verbose
  @debug = debug
  @interval = interval
  # Check for single-site mode
  validate_opts = {}
  if !config.before['url'] && @cache.tag?(:before)
    unless @cache.read_tags.include?(:before)
      raise SiteDiffException,
            "A cached 'before' is required for single-site mode"
    end
    validate_opts[:need_before] = false
  end
  config.validate(validate_opts)

  @concurrency = concurrency
  @config = config
end

Instance Attribute Details

#config ⇒ Object (readonly)

Returns the value of attribute config.



48
49
50
 
# File 'lib/sitediff.rb', line 48

def config
  @config
end

#results ⇒ Object (readonly)

Returns the value of attribute results.



48
49
50
 
# File 'lib/sitediff.rb', line 48

def results
  @results
end

Class Method Details

.log(str, type = :info, label = nil) ⇒ Object

label will be colorized and str will not be. type dictates the color: can be :success, :error, or :failure



25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
 
# File 'lib/sitediff.rb', line 25

def self.log(str, type = :info, label = nil)
  label = label ? "[sitediff] #{label}" : '[sitediff]'
  bg = fg = nil
  case type
  when :info
    bg = fg = nil
  when :diff_success
    bg = :green
    fg = :black
  when :diff_failure
    bg = :red
  when :warn
    bg = :yellow
    fg = :black
  when :error
    bg = :red
  end
  label = Rainbow(label)
  label = label.bg(bg) if bg
  label = label.fg(fg) if fg
  puts label + ' ' + str
end

Instance Method Details

#after ⇒ Object 



53
54
55
 
# File 'lib/sitediff.rb', line 53

def after
  @config.after['url']
end

#before ⇒ Object 



49
50
51
 
# File 'lib/sitediff.rb', line 49

def before
  @config.before['url']
end

#dump(dir, report_before, report_after) ⇒ Object

Dump results to disk



144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
 
# File 'lib/sitediff.rb', line 144

def dump(dir, report_before, report_after)
  report_before ||= before
  report_after ||= after
  dir = Pathname.new(dir)
  dir.mkpath unless dir.directory?

  # store diffs of each failing case, first wipe out existing diffs
  diff_dir = dir + DIFFS_DIR
  diff_dir.rmtree if diff_dir.exist?
  results.each { |r| r.dump(dir) if r.status == Result::STATUS_FAILURE }
  SiteDiff.log "All diff files were dumped inside #{dir.expand_path}"

  # store failing paths
  failures = dir + FAILURES_FILE
  SiteDiff.log "Writing failures to #{failures.expand_path}"
  failures.open('w') do |f|
    results.each { |r| f.puts r.path unless r.success? }
  end

  # create report of results
  report = Diff.generate_html_report(results, report_before, report_after,
                                     @cache)
  dir.+(REPORT_FILE).open('w') { |f| f.write(report) }

  # serve some settings
  settings = { 'before' => report_before, 'after' => report_after,
               'cached' => %w[before after] }
  dir.+(SETTINGS_FILE).open('w') { |f| YAML.dump(settings, f) }
end

#process_results(path, read_results) ⇒ Object

Process a set of read results



92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
 
# File 'lib/sitediff.rb', line 92

def process_results(path, read_results)
  if (error = (read_results[:before].error || read_results[:after].error))
    diff = Result.new(path, nil, nil, nil, nil, error)
  else
    begin
      diff = Result.new(path,
                        *sanitize(path, read_results),
                        read_results[:before].encoding,
                        read_results[:after].encoding,
                        nil)
    rescue => e
      raise if @debug

      Result.new(path, nil, nil, nil, nil, "Sanitization error: #{e}")
    end
  end
  @results[path] = diff

  # Print results in order!
  while (next_diff = @results[@ordered.first])
    next_diff.log(@verbose)
    @ordered.shift
  end
end

#run(curl_opts = {}, debug = true) ⇒ Object

Perform the comparison, populate @results and return the number of failing paths (paths with non-zero diff).



119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
 
# File 'lib/sitediff.rb', line 119

def run(curl_opts = {}, debug = true)
  # Map of path -> Result object, populated by process_results
  @results = {}
  @ordered = @config.paths.dup

  unless @cache.read_tags.empty?
    SiteDiff.log('Using sites from cache: ' +
      @cache.read_tags.sort.join(', '))
  end

  # TODO: Fix this after config merge refactor!
  # Not quite right. We are not passing @config.before or @config.after
  # so passing this instead but @config.after['curl_opts'] is ignored.
  config_curl_opts = @config.before['curl_opts']
  curl_opts = config_curl_opts.clone.merge(curl_opts) if config_curl_opts
  fetcher = Fetch.new(@cache, @config.paths, @interval, @concurrency, curl_opts, debug,
                      before: before, after: after)
  fetcher.run(&method(:process_results))

  # Order by original path order
  @results = @config.paths.map { |p| @results[p] }
  results.map { |r| r unless r.success? }.compact.length
end

#sanitize(path, read_results) ⇒ Object

Sanitize HTML



78
79
80
81
82
83
84
85
86
87
88
89
 
# File 'lib/sitediff.rb', line 78

def sanitize(path, read_results)
  %i[before after].map do |tag|
    html = read_results[tag].content
    encoding = read_results[tag].encoding
    if encoding
      config = @config.send(tag)
      Sanitizer.new(html, config, path: path).sanitize
    else
      html
    end
  end
end