# Magic RSS
# Helps getting around with RSS while ignoring standards and tolerating much
$KCODE = "UTF-8"
require 'rexml/document'
require 'cgi'
require 'iconv'
# Search macros
class REXML::Element
def s(expr)
each_element(expr) { |e| return e }
nil
end
def s_text(expr)
if (e = s(expr))
(e.cdatas.size > 0) ? e.cdatas.to_s : e.text
else
nil
end
end
def s_attr(expr, attr)
(e = s(expr)) ? e.attributes[attr] : nil
end
def atom_content(expr)
e = s(expr)
if e
type = e.attributes['type']
mode = e.attributes['mode']
if type == 'xhtml' or mode == 'xml'
e.children.to_s
elsif type == 'html' or mode == 'escaped'
if e.cdatas.size > 0
e.cdatas.to_s
else
e.text
end
else # 'text'
if e.cdatas.size > 0
CGI::escapeHTML(e.cdatas.to_s)
else
CGI::escapeHTML(e.text.to_s)
end
end
else
nil
end
end
def atom_content_text(expr)
c = atom_content(expr)
if c
c.gsub!(/<.+?>/, '')
CGI::unescapeHTML(c)
else
nil
end
end
def atom_link
link, rank = 0, 0
each_element('link') do |e|
rel = e.attributes['rel']
type = e.attributes['type']
e_rank = (rel == 'alternate' ? 2 : 1) *
(rel == 'enclosure' ? 0 : 1) *
((type == 'text/html' ? 2 : 0) +
(type == 'application/xhtml+xml' ? 1 : 0) +
1)
if e_rank > rank
link = e.attributes['href']
rank = e_rank
end
end
link
end
end
class AbstractMethod < RuntimeError; end
class MRSS
def initialize(root)
@root = root
end
# Common properties
def items
res = []
@root.each_element(items_expr) do |i|
item = self.class::Item.new(i)
res << item unless item.title.to_s == '' and item.link.to_s == ''
end
res
end
# Abstract stuff
def title
raise AbstractMethod.new
end
def type
raise AbstractMethod.new
end
def link
raise AbstractMethod.new
end
def description
raise AbstractMethod.new
end
def feed_expr
raise AbstractMethod.new
end
def items_expr
raise AbstractMethod.new
end
# Item
class Item
def initialize(e)
@e = e
end
def title
raise AbstractMethod.new
end
def link
raise AbstractMethod.new
end
def description
raise AbstractMethod.new
end
def date
raise AbstractMethod.new
end
def enclosures
raise AbstractMethod.new
end
end
# Flavours
# RSS 0.9 & 2.0
class RSS < MRSS
def title
@root.s_text "#{feed_expr}/title"
end
def feed_expr
"/rss/channel"
end
def items_expr
"/rss/channel/item"
end
def link
@root.s_text "#{feed_expr}/link"
end
def description
@root.s_text "#{feed_expr}/description"
end
class Item < ::MRSS::Item
def title
@e.s_text "title"
end
def link
@e.s_text "link"
end
def description
@e.s_text("content:encoded") ||
@e.s_text("encoded") ||
@e.s_text("description")
end
def date
d = @e.s_text('date') || @e.s_text('pubDate') || @e.s_text('dc:date')
d.to_s.detect_time.localtime
end
def enclosures
r = []
@e.each_element('enclosure') do |e|
h = {
'href' => e.attributes['url'],
'type' => e.attributes['type'],
'title' => e.attributes['title'],
'length' => e.attributes['length']
}
r << h
end
r
end
end
end
# RSS 1.0
class RDF < RSS
def feed_expr
"/*/channel"
end
def items_expr
"/*/item"
end
end
# ATOM 0.3 & 1.0
class ATOM < MRSS
def feed_expr
"/feed"
end
def items_expr
"/feed/entry"
end
def title
@root.atom_content_text "title"
end
def link
@root.atom_link
end
def description
@root.atom_content_text('tagline') ||
@root.atom_content_text('subtitle')
end
class Item < ::MRSS::Item
def title
@e.atom_content_text "title"
end
def link
@e.atom_link
end
def description
@e.atom_content("content") ||
@e.atom_content("summary")
end
def date
d = @e.s_text('published') || @e.s_text('issued') || @e.s_text('updated')
d.to_s.detect_time.localtime
end
def enclosures
r = []
@e.each_element("link") do |e|
if e.attributes['rel'] == 'enclosure'
h = {
'href' => e.attributes['href'],
'type' => e.attributes['type'],
'title' => e.attributes['title'],
'length' => e.attributes['length']
}
r << h
end
end
r
end
end
end
def self.parse(body)
begin
xml = REXML::Document.new(body).root
rescue REXML::ParseException
xml = REXML::Document.new(Iconv::iconv('UTF-8', 'ISO-8859-15', body).to_s).root
end
begin
# Try to validate UTF-8
Iconv::iconv('UTF-8', 'UTF-8', xml.to_s).to_s
rescue Iconv::IllegalSequence, Iconv::InvalidCharacter
xml = REXML::Document.new(Iconv::iconv('UTF-8', 'ISO-8859-15', body).to_s).root
end
k = case xml.name
when 'rss'
RSS
when 'rdf'
RDF
when 'RDF'
RDF
when 'feed'
ATOM
else
raise "Unknwon feed format: #{xml.name} (#{xml.namespace})"
end
k.new(xml)
end
end
class String
def detect_time
s = self
tz_offset = 0
# Fix it up
s.strip!
s.scan(/([\+\-])(\d\d):?(\d\d)$/).each do |plus_minus,hours,minutes|
tz_offset = ((hours.to_i * 60) + minutes.to_i) * 60
tz_offset *= (plus_minus == '+') ? -1 : 1
end
# 2004-04-01T21:23+00:00
# 2004-09-05T21:23Z
s.scan(/^(\d+)-(\d+)-(\d+)T(\d+):(\d+)/).each do |y,mo,d,h,m|
return Time.gm(y.to_i, mo.to_i, d.to_i, h.to_i, m.to_i) + tz_offset
end
# 2004-04-01T21:23:23+00:00
s.scan(/^(\d+)-(\d+)-(\d+)T(\d+):(\d+):(\d+)\+(\d+):(\d+)$/).each do |y,mo,d,h,m,s,tz_h,tz_m|
return Time.gm(y.to_i, mo.to_i, d.to_i, h.to_i, m.to_i, s.to_i) + tz_offset
end
# Wed, 20 Apr 2005 19:38:15 +0200
# Fri, 22 Apr 2005 10:31:12 GMT
months = {"Jan" => 1, "Feb" => 2, "Mar" => 3, "Mär" => 3, "Apr" => 4, "May" => 5, "Mai" => 5, "Jun" => 6, "Jul" => 7, "Aug" => 8, "Sep" => 9, "Oct" => 10, "Okt" => 10, "Nov" => 11, "Dec" => 12, "Dez" => 12 }
s.scan(/^(.+?), +(\d+) (.+?) (\d+) (\d+):(\d+):(\d+) /).each do |wday,d,mo,y,h,m,s|
return Time.gm(y.to_i, months[mo], d.to_i, h.to_i, m.to_i, s.to_i) + tz_offset
end
# 06 May 2007 02:20:00
s.scan(/^(\d+) (.+?) (\d{4}) (\d\d):(\d\d):(\d\d)/).each do |d,mo,y,h,m,s|
return Time.gm(y.to_i, months[mo], d.to_i, h.to_i, m.to_i, s.to_i) + tz_offset
end
Time.new
end
end
