Class: YayImdbs

Inherits:
Object
  • Object
show all
Defined in:
lib/yay_imdbs.rb

Constant Summary collapse

IMDB_BASE_URL = 
'http://www.imdb.com/'
IMDB_SEARCH_URL = 
IMDB_BASE_URL + 'find?s=tt&q='
IMDB_MOVIE_URL = 
IMDB_BASE_URL + 'title/tt'
STRIP_WHITESPACE = 
/(\s{2,}|\n|\||\302\240\302\273)/u
['See more',
 'Add/edit official sites',
 'See all certifications',
 'See full summary',
 'see all',
]
DATE_PROPERTIES = 
[:release_date]
LIST_PROPERTIES = 
[:genres, :plot_keywords, :country, :sound_mix, :language]
INT_LIST_PROPERTIES = 
[:year, :season]
PROPERTY_ALIAS = 
{:genres => :genre, 
               :taglines => :tagline, 
               :year => :years, 
               :season => :seasons,
               :language => :languages,
               :motion_picture_rating_mpaa => :mpaa,
:official_sites => :official_site}
OFFICAL_SITE_REGEX = 
/<a href="([^"]+)"[^>]*>Official site<\/a>/

Class Method Summary collapse

Class Method Details

.clean_movie_property(key, value, imdb_id) ⇒ Object 



127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
 
# File 'lib/yay_imdbs.rb', line 127

def clean_movie_property(key, value, imdb_id)
  if DATE_PROPERTIES.include?(key)
    value = Date.strptime(value, '%d %B %Y') rescue nil
  elsif key == :runtime
    if value =~ /(\d+)\smin/
      value = $1.to_i
    else
      value = nil
    end
  elsif key == :official_sites
    value = get_official_site_url(value, imdb_id)
  elsif LIST_PROPERTIES.include?(key)
    value = value.split('|').collect { |l| l.gsub(/[^a-zA-Z0-9\-]/, '') }
  elsif INT_LIST_PROPERTIES.include?(key)
    value = value.split('|').collect { |l| l.strip.to_i }.reject { |y| y <= 0 }
  end
  return value
end

.clean_title(movie_title) ⇒ Object

Remove surrounding double quotes that seems to appear on tv show name



235
236
237
238
 
# File 'lib/yay_imdbs.rb', line 235

def clean_title(movie_title)
  movie_title = $1 if movie_title =~ /^"(.*)"$/
  return movie_title.strip
end

.get_episodes_page(imdb_id) ⇒ Object 



215
216
217
 
# File 'lib/yay_imdbs.rb', line 215

def get_episodes_page(imdb_id)
  Nokogiri::HTML(open(IMDB_MOVIE_URL + imdb_id + '/episodes'))
end

.get_media_page(url_fragment) ⇒ Object 



219
220
221
 
# File 'lib/yay_imdbs.rb', line 219

def get_media_page(url_fragment)
 Nokogiri::HTML(open(IMDB_BASE_URL + url_fragment))
end

.get_movie_page(imdb_id) ⇒ Object 



207
208
209
 
# File 'lib/yay_imdbs.rb', line 207

def get_movie_page(imdb_id)
  Nokogiri::HTML(open(IMDB_MOVIE_URL + imdb_id))
end

.get_official_site_url(value, imdb_id) ⇒ Object

TODO capture all official sites, not all sites have an “Official site” link (e.g. Lost)



159
160
161
162
163
164
165
 
# File 'lib/yay_imdbs.rb', line 159

def get_official_site_url(value, imdb_id)
    value = value.match(OFFICAL_SITE_REGEX)
    if value.nil?
        value = get_official_sites_page(imdb_id).inner_html.match(OFFICAL_SITE_REGEX)
    end
    return $1
end

.get_official_sites_page(imdb_id) ⇒ Object 



211
212
213
 
# File 'lib/yay_imdbs.rb', line 211

def get_official_sites_page(imdb_id)
  Nokogiri::HTML(open(IMDB_MOVIE_URL + imdb_id + '/officialsites'	))
end

.get_search_page(name) ⇒ Object 



203
204
205
 
# File 'lib/yay_imdbs.rb', line 203

def get_search_page(name)
  Nokogiri::HTML(open(IMDB_SEARCH_URL + URI.escape(name)))
end

.get_title_and_year_from_meta(doc) ⇒ Object 



223
224
225
226
227
228
229
230
231
232
 
# File 'lib/yay_imdbs.rb', line 223

def get_title_and_year_from_meta(doc)
  title_text = doc.at_css("meta[name='title']").try(:[], 'content')
  title_text = title_text.sub(/^IMDb - /, '')
  # Matches 'Movie Name (2010)' or 'Movie Name (2010/I)' or 'Lost (TV Series 2004–2010)'
  if title_text && title_text =~ /(.*) \([^\)0-9]*(\d{4})((\/\w*)|(.\d{4})|([^\)]*))?\)/
    movie_title = clean_title($1)
    movie_year = $2.to_i
  end
  return movie_title, movie_year
end

.movie_properties(doc) ⇒ Object 



146
147
148
149
150
151
152
153
154
155
156
 
# File 'lib/yay_imdbs.rb', line 146

def movie_properties(doc)
  doc.css("div h4").each do |h4|
    div = h4.parent
    raw_key = h4.inner_text
    key = raw_key.sub(':', '').strip.downcase
    value = div.inner_text[((div.inner_text =~ /#{Regexp.escape(raw_key)}/) + raw_key.length).. -1]
    value = value.gsub(/\302\240\302\273/u, '').strip.gsub(/(#{MORE_INFO_LINKS.join(')|(')})$/i, '').strip
    symbol_key = key.downcase.gsub(/[^a-zA-Z0-9 ]/, '').gsub(/\s/, '_').to_sym
    yield symbol_key, value
  end
end

.scrap_episodes(info_hash) ⇒ Object 



184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
 
# File 'lib/yay_imdbs.rb', line 184

def scrap_episodes(info_hash)
  episodes = []
  doc = get_episodes_page(info_hash[:imdb_id])

  doc.css(".filter-all").each do |e_div|
    next unless e_div.at_css('h3').inner_text =~ /Season (\d+), Episode (\d+):/
      episode = {"series" => $1.to_i, "episode" => $2.to_i, "title" => $'.strip}

    raw_date = e_div.at_css('strong').inner_text.strip
    episode['date'] = Date.parse(raw_date) rescue nil

    # Seems that the day can sometimes be ???? which doesnt play will with regex
    episode['plot'] = $'.strip if e_div.inner_text =~ /#{raw_date}/ rescue nil

    episodes << episode
  end
  info_hash['episodes'] = episodes
end

.scrap_images(doc, info_hash) ⇒ Object 



167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
 
# File 'lib/yay_imdbs.rb', line 167

def scrap_images(doc, info_hash)
  #scrap poster image urls
  thumbnail_url = doc.at_css("td[id=img_primary] a img").try(:[], 'src')
  return if thumbnail_url.nil? || thumbnail_url =~ /\/nopicture\//

  info_hash['medium_image'] = thumbnail_url
  # Small thumbnail image, gotten by hacking medium url
  info_hash['small_image'] = thumbnail_url.sub(/@@.*$/, '@@._V1._SX120_120,160_.jpg')

  #Try to scrap a larger version of the image url
  large_img_page_link = doc.at_css("td[id=img_primary] a").try(:[], 'href')
  return unless large_img_page_link
  large_img_doc = get_media_page(large_img_page_link) 
  large_img_url = large_img_doc.at_css("img[id=primary-img]").try(:[], 'src')
  info_hash['large_image'] = large_img_url
end

.scrap_movie_info(imdb_id) ⇒ Object 



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
 
# File 'lib/yay_imdbs.rb', line 88

def scrap_movie_info(imdb_id)
  info_hash = {:imdb_id => imdb_id}.with_indifferent_access

  doc = get_movie_page(imdb_id)
  title, year = get_title_and_year_from_meta(doc)
  info_hash[:title], info_hash[:year] = title, year
  if info_hash['title'].nil?
    #If we cant get title and year something is wrong
    raise "Unable to find title or year for imdb id #{imdb_id}"
  end
  info_hash[:video_type] = video_type_from_meta(doc)
  
  info_hash[:plot] = doc.xpath("//td[@id='overview-top']/p[2]").inner_text.strip
  info_hash[:rating] = doc.at_css('.star-box-giga-star').inner_text.gsub(/[^0-9.]/, '').to_f rescue nil

  found_info_divs = false
  movie_properties(doc) do |key, value|
    found_info_divs = true
    info_hash["raw_#{key}"] = value
    info_hash[key] = clean_movie_property(key, value, imdb_id)
    info_hash[PROPERTY_ALIAS[key]] = info_hash[key] if PROPERTY_ALIAS[key]
  end

  unless found_info_divs
    #If we don't find any info divs assume parsing failed
    raise "No info divs found for imdb id #{imdb_id}"
  end

  # Hack: tv shows can have a year property, which is a list, fixing ...
  info_hash[:year] = year

  scrap_images(doc, info_hash)

  #scrap episodes if tv series
  scrap_episodes(info_hash) if info_hash.has_key?('season')

  return info_hash
end

.search_for_imdb_id(name, year = nil, type = nil) ⇒ Object 



42
43
44
45
46
47
48
49
50
51
52
53
 
# File 'lib/yay_imdbs.rb', line 42

def search_for_imdb_id(name, year=nil, type=nil)
  search_results = search_imdb(name)

  search_results.each do |result|
    # Ensure result is the correct video type
    next if type && (result[:video_type].to_s != type.to_s)

    # If no year provided just return first result
    return result[:imdb_id] if year.nil? || result[:year] == year
  end
  return nil
end

.search_imdb(search_term) ⇒ 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
85
86
 
# File 'lib/yay_imdbs.rb', line 55

def search_imdb(search_term)
  search_results = []

  doc = get_search_page(search_term)

  # If the search is an exact match imdb will redirect to the movie page not search results page
  # we uses the title meta element to determine if we got an exact match
  movie_title, movie_year = get_title_and_year_from_meta(doc)
  if movie_title
    canonical_link = doc.at_css("link[rel='canonical']").try(:[], 'href')
    if canonical_link && canonical_link =~ /tt(\d+)\//
      return [:name => movie_title, :year => movie_year, :imdb_id => $1, :video_type => video_type_from_meta(doc)]
    else
      raise "Unable to extract imdb id from exact search result"
    end
  end

  doc.css("td").each do |td| 
    td.css("a").each do |link|
      href = link['href']
      current_name = link.content

      # Ignore links with no text (e.g. image links) or links that don't link to movie pages
      next unless current_name.present? && href =~ /^\/title\/tt(\d+)/
      imdb_id = $1
      current_year = $1.gsub(/\(\)/, '').to_i if td.inner_text =~ /\((\d{4}\/?\w*)\)/
      search_results << {:imdb_id => imdb_id, :name => clean_title(current_name), :year => current_year, :video_type => video_type(td)}
    end
  end

  return search_results
end

.strip_whitespace(s) ⇒ Object

Hackyness to get around ruby 1.9 encoding issue



241
242
243
 
# File 'lib/yay_imdbs.rb', line 241

def strip_whitespace(s)
  s.encode('UTF-8').gsub(STRIP_WHITESPACE, '').strip
end

.video_type(td) ⇒ Object 



245
246
247
248
 
# File 'lib/yay_imdbs.rb', line 245

def video_type(td)
  return :tv_show if td.content =~ /\((TV series|TV)\)/
  return :movie
end

.video_type_from_meta(doc) ⇒ Object 



250
251
252
253
 
# File 'lib/yay_imdbs.rb', line 250

def video_type_from_meta(doc)
  type_text = doc.at_css("meta[property='og:type']").try(:[], 'content')
  type_text =~ /tv_show/ ? :tv_show : :movie
end