Class: Zhangmen::Client

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

Overview

Wraps a client session for accessing Baidu’s streaming music service.

Instance Attribute Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize(options = {}) ⇒ Client

New client session.

Parameters:

  • options (Hash) (defaults to: {}) —

    a customizable set of options

Options Hash (options):

  • proxy (String) —

    “host:port” string, “auto” to have a proxy discovered automatically, or nil / false to use direct requests

  • cache_ttl (Integer) —

    validity of cached requests, in seconds

  • log_level (Integer) —

    severity treshold (e.g., Logger::ERROR)

  • logger (Logger) —

    receiver of logging info



21
22
23
24
25
26
27
28
29
30
31
32
33
34
 
# File 'lib/zhangmen/client.rb', line 21

def initialize(options = {})
  if options[:proxy] == 'auto'
    options[:proxy] = Zhangmen::Proxy.fetch
  end
  
  @mech = mechanizer options
  @curb = curber options
  @cache = {}
  @cache_ttl = options[:cache_ttl] || (24 * 60 * 60)  # 1 day
  log_level = options[:log_level] || Logger::WARN
  @logger = options[:logger] || Logger.new(STDERR)
  @logger.level = log_level
  @parser = options[:use_hpricot] ? :hpricot : :nokogiri
end

Instance Attribute Details

#cache ⇒ Object

Cache of HTTP requests / responses.

The cache covers all the song metadata, but not actual song data.



39
40
41
 
# File 'lib/zhangmen/client.rb', line 39

def cache
  @cache
end

Instance Method Details

#category(category_id) ⇒ Object

Fetches a collection of playlists by the category ID.

Args:

category_id:: right now, categories seem to be numbers, so enumerating
              from 1 onwards should be good

Returns an array of playlists.



48
49
50
51
52
53
54
55
56
57
 
# File 'lib/zhangmen/client.rb', line 48

def category(category_id)
  result = op 3, :list_cat => category_id
  result.search('data').map do |playlist_node|
    {
      :id => playlist_node.search('id').inner_text,
      :name => playlist_node.search('name').inner_text.encode('UTF-8'),
      :song_count => playlist_node.search('tcount').inner_text.to_i
    }
  end
end

#curber(options = {}) ⇒ Object

Curl::Easy instance customized to maximize download success.



245
246
247
248
249
250
251
252
253
254
255
256
 
# File 'lib/zhangmen/client.rb', line 245

def curber(options = {})
  curb = Curl::Easy.new
  curb.enable_cookies = true
  curb.follow_location = true
  curb.useragent = 'Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/534.30 (KHTML, like Gecko) Chrome/12.0.742.124 Safari/534.30'
  if options[:proxy]
    curb.proxy_url = options[:proxy]
  else
    curb.proxy_url = nil
  end
  curb
end

#hostname ⇒ Object

The service hostname.



228
229
230
 
# File 'lib/zhangmen/client.rb', line 228

def hostname
  'box.zhangmen.baidu.com'
end

#mechanizer(options = {}) ⇒ Object

Mechanize instance customized to maximize fetch success.



233
234
235
236
237
238
239
240
241
242
 
# File 'lib/zhangmen/client.rb', line 233

def mechanizer(options = {})
  mech = Mechanize.new
  mech.user_agent = 'Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/534.30 (KHTML, like Gecko) Chrome/12.0.742.124 Safari/534.30'
  if options[:proxy]
    host, _, port_str = *options[:proxy].rpartition(':')
    port_str ||= 80
    mech.set_proxy host, port_str.to_i
  end
  mech
end

#op(opcode, args) ⇒ Object

Performs a numbered operation.

Args:

opcode:: operation number (e.g. 22 for playlist fetch)
args:: operation arguments (e.g. :listid => number for playlist ID)

Returns a Nokogiri root node.



180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
 
# File 'lib/zhangmen/client.rb', line 180

def op(opcode, args)
  @logger.debug { "XML op #{opcode} with #{args.inspect}" }
  cache_key = op_cache_key opcode, args
  if @cache[cache_key] && Time.now.to_f - @cache[cache_key][:at] < @cache_ttl
    xml = @cache[cache_key][:xml]
    @logger.debug { "Cached response\n#{xml}" }
  else
    xml = op_xml_without_cache opcode, args
    @cache[cache_key] = { :at => Time.now.to_f, :xml => xml }
    @logger.debug { "Live response\n#{xml}" }
  end
  
  if @parser == :nokogiri
    Nokogiri.XML(xml).root
  else
    Hpricot(xml)
  end
end

#op_cache_key(opcode, args) ⇒ Object

A string suitable as a key for caching a numbered operation’s result.

Accepts the same arguments as Client#op.



211
212
213
214
215
 
# File 'lib/zhangmen/client.rb', line 211

def op_cache_key(opcode, args)
  { :op => opcode }.merge(args).
      map { |k, v| "#{CGI.escape(k.to_s)}=#{CGI.escape(v.to_s)}" }.
      sort.join('&')
end

#op_url(opcode, args) ⇒ Object

The fetch URL for an XML opcode.

Accepts the same arguments as Client#op.



220
221
222
223
224
225
 
# File 'lib/zhangmen/client.rb', line 220

def op_url(opcode, args)
  query = { :op => opcode }.merge(args).
      map { |k, v| "#{CGI.escape(k.to_s)}=#{CGI.escape(v.to_s)}" }.
      join('&') + '&.r=' + ('%.16f' % Kernel.rand)
  URI.parse "http://#{hostname}/x?#{query}"
end

#op_xml_without_cache(opcode, args) ⇒ Object

Performs a numbered operation, returning the raw XML.

Accepts the same arguments as Client#op.

Does not perform any caching.



204
205
206
 
# File 'lib/zhangmen/client.rb', line 204

def op_xml_without_cache(opcode, args)
  @mech.get(op_url(opcode, args)).body
end

#playlist(list) ⇒ Object

Fetches a playlist.

Args:

list:: a playlist obtained by calling category

Returns an array of songs.



65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
 
# File 'lib/zhangmen/client.rb', line 65

def playlist(list)
  result = op 22, :listid => list[:id]
  
  count = result.search('count').inner_text.to_i
  result.search('data').map do |song_node|
    raw_name = song_node.search('name').inner_text
    if match = /^(.*)\$\$(.*)\$\$\$\$/.match(raw_name)
      title = match[1].encode('UTF-8')
      author = match[2].encode('UTF-8')
    else
      author = title = raw_name.encode('UTF-8')
    end
    
    if @parser == :nokogiri
      native_encoding = result.document.encoding
    else
      native_encoding = raw_name.encoding
    end
    {
      :raw_name => raw_name.encode('UTF-8'),
      :raw_encoding => native_encoding,
      :title => title, :author => author,
      :id => song_node.search('id').inner_text
    }
  end
end

#song(entry) ⇒ Object

Fetches the MP3 contents of a song.

Args:

song:: a song obtained by calling playlist

Returns the MP3 bits.



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
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
 
# File 'lib/zhangmen/client.rb', line 98

def song(entry)
  song_sources(entry).each do |src|
    3.times do
      begin
        @curb.url = src[:url]
        begin
          @curb.perform
        rescue Curl::Err::PartialFileError
          got = @curb.body_str.length
          expected = @curb.downloaded_content_length.to_i
          if got < expected
            @logger.warn do
              "Server hangup fetching #{src[:url]}; got #{got} bytes, " +
              "expected #{expected}"
            end
            # Server gave us fewer bytes than promised in Content-Length.
            # Try again in case the error is temporary.
            sleep 1
            next
          end
        end
        next unless @curb.response_code >= 200 && @curb.response_code < 300
        bits = @curb.body_str
        if bits[-256, 3] == 'TAG' || bits[0, 3] == 'ID3'
          return bits
        else
          break
        end
      rescue Curl::Err::GotNothingError
        @logger.warn do
          "Server hangup fetching #{src[:url]}; got no HTTP response"
        end
        # Try again in case the error is temporary.
        sleep 1
      rescue Curl::Err::RecvError
        @logger.warn do
          "TCP error fetching #{src[:url]}"
        end
        # Try again in case the error is temporary.
        sleep 1
      rescue Timeout::Error
        @logger.warn do
          "Timeout while downloading #{src[:url]}"
        end
        # Try again in case the error is temporary.
        sleep 1
      end
    end
  end
  nil
end

#song_sources(entry) ⇒ Object

Fetches the MP3 download locations for a song.

Args:

song:: a song obtained by calling playlist

Returns



156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
 
# File 'lib/zhangmen/client.rb', line 156

def song_sources(entry)
  title = entry[:raw_name].encode(entry[:raw_encoding])
  result = op 12, :count => 1, :mtype => 1, :title => title, :url => '',
                  :listenreelect => 0
  result.search('url').map do |url_node|
    filename = url_node.search('decode').inner_text
    encoded_url = url_node.search('encode').inner_text
    url = File.join File.dirname(encoded_url), filename
    {
      :url => url,
      :type => url_node.search('type').inner_text.to_i,
      :lyrics_id => url_node.search('lrid').inner_text.to_i,
      :flag => url_node.search('flag').inner_text
    }
  end
end