webpage 0.0.8 → 0.0.9
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- data/lib/webpage.rb +2 -5
- data/spec/webpage_spec.rb +3 -0
- data/webpage.gemspec +1 -1
- metadata +1 -1
data/lib/webpage.rb
CHANGED
|
@@ -1,12 +1,12 @@
|
|
|
1
1
|
#coding:UTF-8
|
|
2
2
|
require 'webpage/common'
|
|
3
3
|
class Webpage
|
|
4
|
+
attr_reader :nokogiri
|
|
4
5
|
def initialize(body,options={})
|
|
5
6
|
raise ArgumentError 'body cannot be empty' unless body
|
|
6
|
-
@body = body
|
|
7
7
|
@options = options
|
|
8
8
|
#@body = @body.force_encoding(@options[:encoding]).encode("UTF-8", :invalid => :replace, :undef => :replace, :replace => "") if @options.has_key?:encoding
|
|
9
|
-
@nokogiri = Nokogiri::HTML(
|
|
9
|
+
@nokogiri = Nokogiri::HTML(body)
|
|
10
10
|
if options.has_key?:uri
|
|
11
11
|
@uri = fuzzy_uri(@options[:uri])
|
|
12
12
|
raise '@uri should be absolute' unless @uri.absolute?
|
|
@@ -37,9 +37,6 @@ class Webpage
|
|
|
37
37
|
def nodes_with(key)
|
|
38
38
|
@nokogiri.xpath("//@#{key}")
|
|
39
39
|
end
|
|
40
|
-
def xpath(xstring)
|
|
41
|
-
@nokogiri.xpath(xstring)
|
|
42
|
-
end
|
|
43
40
|
|
|
44
41
|
def keywords
|
|
45
42
|
@keywords ||= @nokogiri.xpath("//meta[@name='keywords']").map{|meta|meta['content']}.flatten.join.split(',')
|
data/spec/webpage_spec.rb
CHANGED
|
@@ -28,6 +28,9 @@ describe Webpage do
|
|
|
28
28
|
it "links should be an array" do
|
|
29
29
|
page.links.class.should == Array
|
|
30
30
|
end
|
|
31
|
+
it "['h1'].to_a should return Array" do
|
|
32
|
+
page['h1'].to_a.class.should == Array
|
|
33
|
+
end
|
|
31
34
|
|
|
32
35
|
it "[] should return class Nokogiri::XML::NodeSet" do
|
|
33
36
|
%w(canonical keywords description).each do |tag|
|
data/webpage.gemspec
CHANGED