Module: Feedbag

Defined in:
lib/feedbag.rb

Class Method Summary collapse

Class Method Details

._is_http_valid(uri, orig_url) ⇒ Object

not used. yet.



171
172
173
174
175
176
177
178
179
180
 
# File 'lib/feedbag.rb', line 171

def self._is_http_valid(uri, orig_url)
	req = Net::HTTP.get_response(uri)
	orig_uri = URI.parse(orig_url)
	case req
		when Net::HTTPSuccess then
			return true
		else
			return false
	end
end

.add_feed(feed_url, orig_url, base_uri = nil) ⇒ Object 



146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
 
# File 'lib/feedbag.rb', line 146

def self.add_feed(feed_url, orig_url, base_uri = nil)
	# puts "#{feed_url} - #{orig_url}"
	url = feed_url.sub(/^feed:/, '').strip

	if base_uri
		#	url = base_uri + feed_url
		url = URI.parse(base_uri).merge(feed_url).to_s
	end

	begin
		uri = URI.parse(url)
	rescue
		puts "Error with `#{url}'"
		exit 1
	end
	unless uri.absolute?
		orig = URI.parse(orig_url)
		url = orig.merge(url).to_s
	end

	# verify url is really valid
	$feeds.push(url) unless $feeds.include?(url)# if self._is_http_valid(URI.parse(url), orig_url)
end

.feed?(url) ⇒ Boolean

Returns:

  • (Boolean) 


38
39
40
41
42
43
44
45
46
47
48
49
50
51
 
# File 'lib/feedbag.rb', line 38

def self.feed?(url)
	# use LWR::Simple.normalize some time
	url_uri = URI.parse(url)
	url = "#{url_uri.scheme or 'http'}://#{url_uri.host}#{url_uri.path}"
	# hack:
	url.sub!(/^feed:\/\//, 'http://')

	res = self.find(url)
	if res.size == 1 and res.first == url
		return true
	else
		return false
	end
end

.find(url, args = {}, debug = false) ⇒ Object 



53
54
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
87
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
126
127
128
129
130
131
132
133
134
135
136
 
# File 'lib/feedbag.rb', line 53

def self.find(url, args = {}, debug = false)
	$feeds = []

	url_uri = URI.parse(url)
	url = nil
	if url_uri.scheme.nil?
	  url = "http://#{url_uri.to_s}"
	elsif url_uri.scheme == "feed"
	  return self.add_feed(url_uri.to_s.sub(/^feed:\/\//, 'http://'), nil)
	else
	  url = url_uri.to_s
	end
	#url = "#{url_uri.scheme or 'http'}://#{url_uri.host}#{url_uri.path}"

	# check if feed_valid is avail
   unless args[:narrow]
     begin
  		require "feed_validator"
	  	v = W3C::FeedValidator.new
		  v.validate_url(url)
 			return self.add_feed(url, nil) if v.valid?
  	rescue LoadError
	  	# scoo
 		rescue REXML::ParseException
  	  # usually indicates timeout
	    # TODO: actually find out timeout. use Terminator?
	    $stderr.puts "Feed looked like feed but might not have passed validation or timed out" if debug
     rescue => ex
  		$stderr.puts "#{ex.class} error ocurred with: `#{url}': #{ex.message}" if debug
	  end
   end

	begin
		html = open(url) do |f|
			if @content_types.include?(f.content_type.downcase)
				return self.add_feed(url, nil)
			end

			doc = Hpricot(f.read)

			if doc.at("base") and doc.at("base")["href"]
				$base_uri = doc.at("base")["href"]
			else
				$base_uri = nil
			end

			# first with links
			(doc/"link").each do |l|
				next unless l["rel"]
				if l["type"] and @content_types.include?(l["type"].downcase.strip) and (l["rel"].downcase =~ /alternate/i or l["rel"] == "service.feed")
					self.add_feed(l["href"], url, $base_uri)
				end
			end

       unless args[:narrow]
 				(doc/"a").each do |a|
  				next unless a["href"]
	  			if self.looks_like_feed?(a["href"]) and (a["href"] =~ /\// or a["href"] =~ /#{url_uri.host}/)
		  			self.add_feed(a["href"], url, $base_uri)
			  	end
 				end

  			(doc/"a").each do |a|
	  			next unless a["href"]
		  		if self.looks_like_feed?(a["href"])
			  		self.add_feed(a["href"], url, $base_uri)
				  end
 				end
       end

		end
	rescue Timeout::Error => err
		$stderr.puts "Timeout error ocurred with `#{url}: #{err}'" if debug
	rescue OpenURI::HTTPError => the_error
		$stderr.puts "Error ocurred with `#{url}': #{the_error}" if debug
	rescue SocketError => err
		$stderr.puts "Socket error ocurred with: `#{url}': #{err}" if debug
	rescue => ex
		$stderr.puts "#{ex.class} error ocurred with: `#{url}': #{ex.message}" if debug
	ensure
		return $feeds
	end
	
end

.looks_like_feed?(url) ⇒ Boolean

Returns:

  • (Boolean) 


138
139
140
141
142
143
144
 
# File 'lib/feedbag.rb', line 138

def self.looks_like_feed?(url)
	if url =~ /(\.(rdf|xml|rdf|rss)$|feed=(rss|atom)|(atom|feed)\/?$)/i
		true
	else
		false
	end
end