Class: Opener::PolarityTagger::LexiconsCache

Inherits:
Object
  • Object
show all
Includes:
MonitorMixin
Defined in:
lib/opener/polarity_tagger/lexicons_cache.rb

Instance Method Summary collapse

Constructor Details

#initialize ⇒ LexiconsCache

Returns a new instance of LexiconsCache.



7
8
9
10
11
12
13
 
# File 'lib/opener/polarity_tagger/lexicons_cache.rb', line 7

def initialize
  super #MonitorMixin

  @url   = ENV['POLARITY_LEXICON_URL']
  @path  = ENV['POLARITY_LEXICON_PATH']
  @cache = {}
end

Instance Method Details

#[](**params) ⇒ Object Also known as: get 



15
16
17
18
19
20
21
22
23
24
25
 
# File 'lib/opener/polarity_tagger/lexicons_cache.rb', line 15

def [] **params
  synchronize do
    if existing = @cache[params]
      existing.tap do
        Thread.new{ @cache[params] = cache_update existing, **params }
      end
    else
      @cache[params] = cache_update **params
    end
  end
end

#cache_update(existing = nil, **params) ⇒ Object 



28
29
30
31
32
33
34
35
36
37
 
# File 'lib/opener/polarity_tagger/lexicons_cache.rb', line 28

def cache_update existing = nil, **params
  from     = Time.now
  lexicons = load_lexicons cache: existing, **params

  return existing if existing and lexicons.blank?
  Hashie::Mash.new(
    lexicons: lexicons,
    from:     from,
  )
end

#load_from_path(lang:, **params) ⇒ Object 



55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
 
# File 'lib/opener/polarity_tagger/lexicons_cache.rb', line 55

def load_from_path lang:, **params
  @path  ||= 'core/general-lexicons'
  dir      = "#{@path}/#{lang.upcase}-lexicon"
  config   = Nokogiri::XML File.read "#{dir}/config.xml"
  lexicons = []

  config.css(:lexicon).each do |cl|
    filename = cl.at(:filename).text
    resource = cl.at(:resource).text
    xml      = Nokogiri::XML File.read "#{dir}/#{filename}"
    puts "#{lang}: loading lexicons from the file #{filename}"

    lexicons.concat(xml.css(:LexicalEntry).map do |le|
      Hashie::Mash.new(
        resource:   resource,
        identifier: le.attr(:id),
        type:       le.attr(:type),
        lemma:      le.at(:Lemma).attr(:writtenForm).downcase,
        pos:        le.attr(:partOfSpeech)&.downcase,
        aspect:     le.at(:Domain)&.attr(:aspect)&.downcase,
        polarity:   le.at(:Sentiment).attr(:polarity),
        strength:   le.at(:Sentiment).attr(:strength),
        confidence_level:   le.at(:Confidence)&.attr(:level),
        domain_conditional: le.at(:Domain)&.attr(:conditional) == 'yes',
      )
    end)
  end

  lexicons
end

#load_from_url(lang:, cache:, **params) ⇒ Object 



45
46
47
48
49
50
51
52
53
 
# File 'lib/opener/polarity_tagger/lexicons_cache.rb', line 45

def load_from_url lang:, cache:, **params
  url  = "#{@url}&language_code=#{lang}&#{params.to_query}"
  url += "&if_updated_since=#{cache.from.iso8601}" if cache
  puts "#{lang}: loading lexicons from url #{url}"

  lexicons = JSON.parse HTTPClient.new.get(url).body
  lexicons = lexicons['data'].map{ |l| Hashie::Mash.new l }
  lexicons
end

#load_lexicons(lang:, **params) ⇒ Object 



39
40
41
42
43
 
# File 'lib/opener/polarity_tagger/lexicons_cache.rb', line 39

def load_lexicons lang:, **params
  lexicons = if @url then load_from_url lang: lang, **params else load_from_path lang: lang, **params end

  LexiconMap.new lang: lang, lexicons: lexicons
end