Module: Textract

Defined in:
lib/textract.rb,
lib/textract/version.rb

Defined Under Namespace

Classes: Client

Constant Summary collapse

TAG_WHITELIST =

attr_accessor :client

%w[
  div span p a img ul ol li blockquote table tr td h1 h2 h3 h4 h5 b em i strong
  figure
]
VERSION = 
"0.0.22"

Class Method Summary collapse

Class Method Details

.build_author(article, html) ⇒ Object 



118
119
120
121
122
123
 
# File 'lib/textract.rb', line 118

def self.build_author(article, html)
  {
    name: article.author || get_author(html),
    twitter: get_twitter(html),
  }
end

.build_site(url, html) ⇒ Object 



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
 
# File 'lib/textract.rb', line 91

def self.build_site(url, html)
  site_twitter = Nokogiri::HTML(html).search('meta[name="twitter:site"]')
  site_name = Nokogiri::HTML(html).search('meta[property="og:site_name"]')
  if site_name.empty?
    domain_regex = /^(http|ftp)s?:\/\/(([\w\d\-_]+\.)?([\w\d\-_]+)\.(\w+))\//
    site = url.match domain_regex
    site = site[2] unless site[2].nil?
    site_name = site.sub(/^www\./, '').capitalize!
  else
    site_name = site_name.attribute('content').value
  end
  if !site_twitter.empty?
    if !site_twitter.attribute('content').nil?
      site_twitter = site_twitter.attribute('content').value
    elsif !site_twitter.attribute('value').nil?
      site_twitter = site_twitter.attribute('value').value
    else
      site_twitter = nill
    end
  end
  {
    name: site_name,
    twitter: site_twitter,
  }
end

.generate_hash(text) ⇒ Object 



125
126
127
 
# File 'lib/textract.rb', line 125

def self.generate_hash(text)
  Digest::MD5.hexdigest text
end

.get_author(html) ⇒ Object 



69
70
71
72
73
74
75
 
# File 'lib/textract.rb', line 69

def self.get_author(html)
  name_meta = Nokogiri::HTML(html).search('meta[name="author"]')
  if name_meta.empty?
    name_meta = Nokogiri::HTML(html).search('meta[property="author"]')
  end
  name_meta.attribute('content').value unless name_meta.empty?
end

.get_og_tags(html, url) ⇒ Object 



19
20
21
22
23
24
25
 
# File 'lib/textract.rb', line 19

def self.get_og_tags(html, url)
  begin
    OpenGraph.new(html)
  rescue
    OpenGraph.new(url)
  end
end

.get_page_title(html) ⇒ Object 



65
66
67
 
# File 'lib/textract.rb', line 65

def self.get_page_title(html)
  Nokogiri::HTML(html).search('head').search('title').text
end

.get_text(url, selectors = nil, format = "markdown") ⇒ Object 



15
16
17
 
# File 'lib/textract.rb', line 15

def self.get_text(url, selectors=nil, format="markdown")
  @client = Client.new(url, selectors, format)
end

.get_twitter(html) ⇒ Object 



77
78
79
80
81
82
83
84
85
86
87
88
89
 
# File 'lib/textract.rb', line 77

def self.get_twitter(html)
  twitter_meta = Nokogiri::HTML(html).search('meta[name="twitter:creator"]')
  if !twitter_meta.empty?
    if !twitter_meta.attribute('content').nil?
      twitter_meta = twitter_meta.attribute('content').value
    elsif !twitter_meta.attribute('value').nil?
      twitter_meta = twitter_meta.attribute('value').value
    else
      twitter_meta = nill
    end
  end
  twitter_meta
end

.smart_extract(html, description, selectors) ⇒ Object 



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
56
57
58
59
60
61
62
63
 
# File 'lib/textract.rb', line 27

def self.smart_extract(html, description, selectors)
  doc = Nokogiri::HTML html
  if selectors.nil?
    article = doc.search('article')
  else
    article = doc.search(selectors)
  end
  if article.count == 1
    article_el = article[0]
  elsif !description.nil? and article.count == 0
    els = [1,2,3]
    i = 1
    until els.count < 2 or i > description.split(" ").length
      search_text = description.split(" ")[0..i].join(" ")
      if search_text.index "'"
        els = doc.search "[text()*=\"#{search_text}\"]"
      else
        els = doc.search "[text()*='#{search_text}']"
      end
      i += 1
    end
    if els.count == 1
      el = els[0]
      article_el = el.parent
    else
      article_el = doc
      # do something else if multiple or no matches
    end
  else
    article_el = doc
  end
  Readability::Document.new(article_el.to_s,
                            tags: TAG_WHITELIST,
                            attributes: %w[src href],
                            remove_empty_nodes: false,
                           )
end