Class: PostAnalyzer

Inherits:
Object
  • Object
show all
Defined in:
app/models/post_analyzer.rb

Class Method Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize(raw, topic_id) ⇒ PostAnalyzer

Returns a new instance of PostAnalyzer.



4
5
6
7
8
9
# File 'app/models/post_analyzer.rb', line 4

def initialize(raw, topic_id)
  @raw = raw
  @topic_id = topic_id
  @onebox_urls = []
  @found_oneboxes = false
end

Class Method Details

.parse_uri_rfc2396(uri) ⇒ Object

from rack … compat with ruby 2.2



92
93
94
95
# File 'app/models/post_analyzer.rb', line 92

def self.parse_uri_rfc2396(uri)
  @parser ||= defined?(URI::RFC2396_Parser) ? URI::RFC2396_Parser.new : URI
  @parser.parse(uri)
end

Instance Method Details

#attachment_countObject

How many attachments are present in the post



73
74
75
76
77
78
79
80
81
82
83
# File 'app/models/post_analyzer.rb', line 73

def attachment_count
  return 0 unless @raw.present?

  attachments =
    cooked_stripped.css("a.attachment[href^=\"#{Discourse.store.absolute_base_url}\"]")
  attachments +=
    cooked_stripped.css(
      "a.attachment[href^=\"#{Discourse.store.relative_base_url}\"]",
    ) if Discourse.store.internal?
  attachments.count
end

#cook(raw, opts = {}) ⇒ Object

What we use to cook posts



23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
# File 'app/models/post_analyzer.rb', line 23

def cook(raw, opts = {})
  cook_method = opts[:cook_method]
  return raw if cook_method == Post.cook_methods[:raw_html]

  if cook_method == Post.cook_methods[:email]
    cooked = EmailCook.new(raw).cook(opts)
  else
    cooked = PrettyText.cook(raw, opts)
  end

  limit = SiteSetting.max_oneboxes_per_post
  result =
    Oneboxer.apply(cooked, extra_paths: ".inline-onebox-loading") do |url, element|
      if opts[:invalidate_oneboxes]
        Oneboxer.invalidate(url)
        InlineOneboxer.invalidate(url)
      end
      next if element["class"] != Oneboxer::ONEBOX_CSS_CLASS
      next if limit <= 0
      limit -= 1
      @onebox_urls << url
      onebox = Oneboxer.cached_onebox(url)
      @found_oneboxes = true if onebox.present?
      onebox
    end

  if result.changed?
    PrettyText.sanitize_hotlinked_media(result.doc)
    cooked = result.to_html
  end

  cooked
end

#cooked_strippedObject



146
147
148
149
150
151
152
153
154
155
# File 'app/models/post_analyzer.rb', line 146

def cooked_stripped
  @cooked_stripped ||=
    begin
      doc = Nokogiri::HTML5.fragment(cook(@raw, topic_id: @topic_id))
      doc.css(
        "pre .mention, aside.quote > .title, aside.quote .mention, aside.quote .mention-group, .onebox, .elided",
      ).remove
      doc
    end
end

#embedded_media_countObject

How many images are present in the post



58
59
60
61
62
63
64
65
66
67
68
69
70
# File 'app/models/post_analyzer.rb', line 58

def embedded_media_count
  return 0 unless @raw.present?

  # TODO - do we need to look for tags other than img, video and audio?
  cooked_stripped
    .css("img", "video", "audio")
    .reject do |t|
      if dom_class = t["class"]
        (Post.allowed_image_classes & dom_class.split).count > 0
      end
    end
    .count
end

#found_oneboxes?Boolean

Returns:

  • (Boolean)


11
12
13
# File 'app/models/post_analyzer.rb', line 11

def found_oneboxes?
  @found_oneboxes
end

#has_oneboxes?Boolean

Returns:

  • (Boolean)


15
16
17
18
19
20
# File 'app/models/post_analyzer.rb', line 15

def has_oneboxes?
  return false unless @raw.present?

  cooked_stripped
  found_oneboxes?
end

How many links are present in the post



142
143
144
# File 'app/models/post_analyzer.rb', line 142

def link_count
  raw_links.size + @onebox_urls.size
end

#linked_hostsObject

Count how many hosts are linked in the post



98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
# File 'app/models/post_analyzer.rb', line 98

def linked_hosts
  all_links = raw_links + @onebox_urls

  return {} if all_links.blank?
  return @linked_hosts if @linked_hosts.present?

  @linked_hosts = {}

  all_links.each do |u|
    begin
      uri = self.class.parse_uri_rfc2396(u)
      host = uri.host
      @linked_hosts[host] ||= 1 unless host.nil?
    rescue URI::Error
      # An invalid URI does not count as a host
      next
    end
  end

  @linked_hosts
end

Returns an array of all links in a post excluding mentions



121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
# File 'app/models/post_analyzer.rb', line 121

def raw_links
  return [] unless @raw.present?
  return @raw_links if @raw_links.present?

  @raw_links = []
  cooked_stripped
    .css("a")
    .each do |l|
      # Don't include @mentions in the link count
      next if link_is_a_mention?(l)
      # Don't include heading anchor in the link count
      next if link_is_an_anchor?(l)
      # Don't include hashtags in the link count
      next if link_is_a_hashtag?(l)
      @raw_links << l["href"].to_s
    end

  @raw_links
end

#raw_mentionsObject



85
86
87
88
89
# File 'app/models/post_analyzer.rb', line 85

def raw_mentions
  return [] if @raw.blank?
  return @raw_mentions if @raw_mentions.present?
  @raw_mentions = PrettyText.extract_mentions(cooked_stripped)
end