Class: Pluto::Updater

Inherits:
Object
  • Object
show all
Includes:
LogUtils::Logging, Models
Defined in:
lib/pluto/updater.rb

Instance Attribute Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize ⇒ Updater

Returns a new instance of Updater.



9
10
11
 
# File 'lib/pluto/updater.rb', line 9

def initialize
  @worker  = ::Fetcher::Worker.new
end

Instance Attribute Details

#worker ⇒ Object (readonly)

Returns the value of attribute worker.



13
14
15
 
# File 'lib/pluto/updater.rb', line 13

def worker
  @worker
end

Instance Method Details

#debug=(value) ⇒ Object 



15
16
17
18
 
# File 'lib/pluto/updater.rb', line 15

def debug=(value)
  @debug = value
  ### logger.debug "[Updater] setting debug flag - debug? #{debug?}"
end

#debug? ⇒ Boolean

Returns:

  • (Boolean) 


20
21
22
 
# File 'lib/pluto/updater.rb', line 20

def debug?
  @debug || false
end

#fetch_feed(url) ⇒ Object 



24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
 
# File 'lib/pluto/updater.rb', line 24

def fetch_feed( url )
  xml = worker.read( url )

  ###
  # NB: Net::HTTP will NOT set encoding UTF-8 etc.
  # will mostly be ASCII
  # - try to change encoding to UTF-8 ourselves
  logger.debug "xml.encoding.name (before): #{xml.encoding.name}"

  #####
  # NB: ASCII-8BIT == BINARY == Encoding Unknown; Raw Bytes Here

  ## NB:
  # for now "hardcoded" to utf8 - what else can we do?
  # - note: force_encoding will NOT change the chars only change the assumed encoding w/o translation
  xml = xml.force_encoding( Encoding::UTF_8 )
  logger.debug "xml.encoding.name (after): #{xml.encoding.name}"      
  xml
end

#handle_feed_item_atom(item) ⇒ Object 



154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
 
# File 'lib/pluto/updater.rb', line 154

def handle_feed_item_atom( item )

      ## todo: if content.content empty use summary for example
      item_attribs = {
        title:        item.title.content,
        url:          item.link.href,
        published_at: item.updated.content.utc.strftime( "%Y-%m-%d %H:%M" ),
        # content:   item.content.content,
      }

      item_attribs[ :guid ] = item.id.content

      if item.summary
        item_attribs[ :content ] = item.summary.content
      else
        if item.content
          text  = item.content.content.dup
          ## strip all html tags
          text = text.gsub( /<[^>]+>/, '' )
          text = text[ 0..400 ] # get first 400 chars
          ## todo: check for length if > 400 add ... at the end???
          item_attribs[ :content ] = text
        end
      end

      puts "- #{item.title.content}"
      puts "  link >#{item.link.href}<"
      puts "  id (~guid) >#{item.id.content}<"
      
      ### todo: use/try published first? why? why not?
      puts "  updated (~pubDate) >#{item.updated.content}< >#{item.updated.content.utc.strftime( "%Y-%m-%d %H:%M" )}< : #{item.updated.content.class.name}"
      puts
      
      # puts "*** dump item:"
      # pp item

      item_attribs
end

#handle_feed_item_rss(item) ⇒ Object 



193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
 
# File 'lib/pluto/updater.rb', line 193

def handle_feed_item_rss( item )

     item_attribs = {
        title:        item.title,
        url:          item.link,
        published_at: item.pubDate.utc.strftime( "%Y-%m-%d %H:%M" ),
        # content:  item.content_encoded,
      }
      
      # if item.content_encoded.nil?
        # puts " using description for content"
        item_attribs[ :content ] = item.description
      # end
      
      item_attribs[ :guid ] = item.guid.content
      
      puts "- #{item.title}"
      puts "  link (#{item.link})"
      puts "  guid (#{item.guid.content})"
      puts "  pubDate >#{item.pubDate}< >#{item.pubDate.utc.strftime( "%Y-%m-%d %H:%M" )}< : #{item.pubDate.class.name}"
      puts

      # puts "*** dump item:"
      # pp item

      item_attribs
end

#parse_feed(xml) ⇒ Object 



45
46
47
48
49
50
51
52
53
54
55
 
# File 'lib/pluto/updater.rb', line 45

def parse_feed( xml )
  parser = RSS::Parser.new( xml )
  parser.do_validate            = false
  parser.ignore_unknown_element = true

  puts "Parsing feed..."
  feed = parser.parse
    
  puts "  feed.class=#{feed.class.name}"
  feed
end

#update_feeds(opts = {}) ⇒ Object 



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
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
 
# File 'lib/pluto/updater.rb', line 96

def update_feeds( opts={} )

  logger.debug "using stdlib RSS::VERSION #{RSS::VERSION}"

  Feed.all.each do |feed_rec|

    feed_key = feed_rec.key
    feed_url = feed_rec.feed_url
    
    feed_xml = fetch_feed( feed_url )

    logger.debug "feed_xml:"
    logger.debug feed_xml[ 0..300 ] # get first 300 chars

    #  if opts.verbose?  # also write a copy to disk
    if debug?
      logger.debug "saving feed to >./#{feed_key}.xml<..."
      File.open( "./#{feed_key}.xml", 'w' ) do |f|
        f.write( feed_xml )
      end
    end
    
    puts "Before parsing feed >#{feed_key}<..."

    feed = parse_feed( feed_xml )

    if feed.class == RSS::Atom::Feed
      puts "== #{feed.title.content} =="
    else  ## assume RSS::Rss::Feed
      puts "==  #{feed.channel.title} =="
    end

    feed.items.each do |item|
      if feed.class == RSS::Atom::Feed
        item_attribs = handle_feed_item_atom( item )
      else  ## assume RSS::Rss::Feed
        item_attribs = handle_feed_item_rss( item )
      end

      # add feed_id fk_ref
      item_attribs[ :feed_id ] = feed_rec.id

      rec = Item.find_by_guid( item_attribs[ :guid ] )
      if rec.nil?
        rec      = Item.new
        puts "** NEW"
      else
        puts "UPDATE"
      end
              
      rec.update_attributes!( item_attribs )
    end  # each item

  end # each feed

end

#update_subscriptions(config, opts = {}) ⇒ Object 



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
87
88
89
90
91
92
93
 
# File 'lib/pluto/updater.rb', line 58

def update_subscriptions( config, opts={} )

  ## for now - use single site w/ key planet  -- fix!! allow multiple sites (planets)
  
  site_key = 'planet'
  site_rec = Site.find_by_key( site_key )
  if site_rec.nil?
    site_rec      = Site.new
    site_rec.key  = site_key
  end
  site_rec.title  = config[ 'title' ]
  site_rec.save!


  config[ 'feeds' ].each do |feed_key|
    
    feed_hash  = config[ feed_key ]
    feed_url   = feed_hash[ 'feed_url' ]
    
    puts "Updating feed subscription >#{feed_key}< - >#{feed_url}<..."

    feed_rec = Feed.find_by_key( feed_key )
    if feed_rec.nil?
      feed_rec      = Feed.new
      feed_rec.key  = feed_key
    end
    feed_rec.feed_url = feed_url
    feed_rec.url      = feed_hash[ 'url' ]
    feed_rec.title    = feed_hash[ 'title' ]    # todo: use title from feed?
    feed_rec.save!
    
    ## todo:
    #  add subscription records  (feed,site)  - how? 
  end

end