omnivore 0.0.2 → 0.0.3
Sign up to get free protection for your applications and to get access to all the features.
- data/lib/omnivore/document.rb +76 -9
- data/lib/omnivore/version.rb +1 -1
- data/spec/document_spec.rb +24 -3
- metadata +2 -4
- data/lib/omnivore/html_helper.rb +0 -52
- data/spec/html_helper_spec.rb +0 -42
data/lib/omnivore/document.rb
CHANGED
@@ -1,33 +1,100 @@
|
|
1
|
+
require "nokogiri"
|
2
|
+
require "omnivore/http_client"
|
3
|
+
|
1
4
|
module Omnivore
|
2
|
-
require "omnivore/http_client"
|
3
|
-
require "omnivore/html_helper"
|
4
5
|
|
5
6
|
class Document
|
6
|
-
attr_reader :
|
7
|
+
attr_reader :model
|
8
|
+
CONTAINER_TAGS = %w[div p]
|
9
|
+
Paragraph = Struct.new("Block", :path, :html, :text)
|
10
|
+
|
7
11
|
|
8
12
|
def self.from_url(url)
|
9
13
|
Document.new(HttpClient.get(url))
|
10
14
|
end
|
11
15
|
|
16
|
+
|
12
17
|
def self.from_html(html)
|
13
18
|
Document.new(html)
|
14
19
|
end
|
15
20
|
|
16
21
|
|
17
22
|
def initialize(html)
|
18
|
-
@
|
23
|
+
@model = Nokogiri::HTML.parse(html) { |config|
|
24
|
+
config.options = Nokogiri::XML::ParseOptions::NOBLANKS
|
25
|
+
}
|
26
|
+
end
|
27
|
+
|
28
|
+
|
29
|
+
def to_html
|
30
|
+
self.model.to_html
|
19
31
|
end
|
20
32
|
|
21
33
|
|
22
34
|
def title
|
23
|
-
|
24
|
-
|
25
|
-
|
26
|
-
|
27
|
-
|
35
|
+
@title ||= self.model.xpath("/html/head/title").text.gsub(/\s+/, " ").strip
|
36
|
+
end
|
37
|
+
|
38
|
+
|
39
|
+
def metadata
|
40
|
+
@metadata ||= self.model.xpath("//meta").inject({ }) { |memo, el|
|
41
|
+
memo[el.attr("name")] = el.attr("content") || "" if el.attr("name")
|
42
|
+
memo
|
43
|
+
}
|
28
44
|
end
|
29
45
|
|
30
46
|
|
47
|
+
def to_text
|
48
|
+
paragraphs = self.to_paragraphs.keep_if { |p| (p.text.size / p.html.size.to_f) > 0.1 }
|
49
|
+
paragraphs.map { |p| p.text }.join("\n")
|
50
|
+
end
|
51
|
+
|
52
|
+
|
53
|
+
def to_paragraphs
|
54
|
+
filter(self.model.xpath("/html/body")).map { |block|
|
55
|
+
html = block.to_html.gsub(/\s+/, " ").strip
|
56
|
+
text = flatten(block).inject([ ]) { |memo, node|
|
57
|
+
memo << node.text.gsub(/\s+/, " ").strip if node.kind_of?(Nokogiri::XML::Text)
|
58
|
+
memo
|
59
|
+
}.join(" ")
|
60
|
+
Paragraph.new(block.path.to_s, html, text)
|
61
|
+
}
|
62
|
+
end
|
63
|
+
|
64
|
+
|
65
|
+
private
|
66
|
+
|
67
|
+
def filter(container)
|
68
|
+
elements = [ ]
|
69
|
+
container.children.each { |child|
|
70
|
+
if CONTAINER_TAGS.include?(child.name)
|
71
|
+
unless child.attr("class") =~ /comment/i
|
72
|
+
elements << child
|
73
|
+
elements += filter(child)
|
74
|
+
end
|
75
|
+
end
|
76
|
+
}
|
77
|
+
elements
|
78
|
+
end
|
79
|
+
|
80
|
+
|
81
|
+
def flatten(block)
|
82
|
+
elements = [ ]
|
83
|
+
return elements if block.nil?
|
84
|
+
return elements if block.respond_to?('cdata?') and block.cdata?
|
85
|
+
return elements if block.respond_to?('comment?') and block.comment?
|
86
|
+
if block.children.empty?
|
87
|
+
elements << block
|
88
|
+
else
|
89
|
+
block.children.each { |child|
|
90
|
+
unless %w[div p].include?(child.name)
|
91
|
+
elements += flatten(child)
|
92
|
+
end
|
93
|
+
}
|
94
|
+
end
|
95
|
+
elements
|
96
|
+
end
|
97
|
+
|
31
98
|
end
|
32
99
|
|
33
100
|
end
|
data/lib/omnivore/version.rb
CHANGED
data/spec/document_spec.rb
CHANGED
@@ -7,6 +7,8 @@ describe Document do
|
|
7
7
|
<html>
|
8
8
|
<head>
|
9
9
|
<title>Nothing To See</title>
|
10
|
+
<meta name="description" content="This is a test page.">
|
11
|
+
<meta name="keywords" content="metadata, testing, kayne west">
|
10
12
|
</head>
|
11
13
|
<body>
|
12
14
|
<p>Nothing to see here, move along.</p>
|
@@ -14,17 +16,36 @@ describe Document do
|
|
14
16
|
</html>
|
15
17
|
}
|
16
18
|
|
19
|
+
|
17
20
|
it "should fetch the content of the provided url" do
|
18
21
|
document = Document.from_url("http://www.google.com")
|
19
|
-
document.
|
20
|
-
document.html.should_not be_empty
|
22
|
+
document.to_html.should_not be_empty
|
21
23
|
end
|
22
24
|
|
23
|
-
|
25
|
+
|
26
|
+
it "should contain the document title" do
|
24
27
|
document = Document.from_html(STATIC_HTML)
|
25
28
|
document.title.should_not be_nil
|
26
29
|
document.title.should_not be_empty
|
27
30
|
document.title.should == "Nothing To See"
|
28
31
|
end
|
29
32
|
|
33
|
+
|
34
|
+
it "should contain the document metadata" do
|
35
|
+
document = Document.from_html(STATIC_HTML)
|
36
|
+
document.metadata.should_not be_nil
|
37
|
+
document.metadata.should_not be_empty
|
38
|
+
document.metadata["keywords"].split(",").first.strip.should == "metadata"
|
39
|
+
end
|
40
|
+
|
41
|
+
|
42
|
+
it "should be able to extract the main content and ignore navigation and ads." do
|
43
|
+
#document = Document.from_url("http://www.marieclaire.com/career-money/jobs/thia-breen-interview")
|
44
|
+
document = Document.from_html(STATIC_HTML)
|
45
|
+
text = document.to_text
|
46
|
+
text.should_not be_nil
|
47
|
+
text.should_not be_empty
|
48
|
+
text.should == "Nothing to see here, move along."
|
49
|
+
end
|
50
|
+
|
30
51
|
end
|
metadata
CHANGED
@@ -2,7 +2,7 @@
|
|
2
2
|
name: omnivore
|
3
3
|
version: !ruby/object:Gem::Version
|
4
4
|
prerelease:
|
5
|
-
version: 0.0.
|
5
|
+
version: 0.0.3
|
6
6
|
platform: ruby
|
7
7
|
authors:
|
8
8
|
- Matthias Eder
|
@@ -10,7 +10,7 @@ autorequire:
|
|
10
10
|
bindir: bin
|
11
11
|
cert_chain: []
|
12
12
|
|
13
|
-
date: 2012-01-
|
13
|
+
date: 2012-01-10 00:00:00 -07:00
|
14
14
|
default_executable:
|
15
15
|
dependencies:
|
16
16
|
- !ruby/object:Gem::Dependency
|
@@ -52,12 +52,10 @@ files:
|
|
52
52
|
- Rakefile
|
53
53
|
- lib/omnivore.rb
|
54
54
|
- lib/omnivore/document.rb
|
55
|
-
- lib/omnivore/html_helper.rb
|
56
55
|
- lib/omnivore/http_client.rb
|
57
56
|
- lib/omnivore/version.rb
|
58
57
|
- omnivore.gemspec
|
59
58
|
- spec/document_spec.rb
|
60
|
-
- spec/html_helper_spec.rb
|
61
59
|
- spec/http_client_spec.rb
|
62
60
|
has_rdoc: true
|
63
61
|
homepage: ""
|
data/lib/omnivore/html_helper.rb
DELETED
@@ -1,52 +0,0 @@
|
|
1
|
-
require "nokogiri"
|
2
|
-
|
3
|
-
module Omnivore
|
4
|
-
module HtmlHelper
|
5
|
-
|
6
|
-
class HtmlTransformer
|
7
|
-
|
8
|
-
def initialize(html)
|
9
|
-
@html = html
|
10
|
-
end
|
11
|
-
|
12
|
-
def to_text
|
13
|
-
document = Nokogiri::HTML.parse(@html)
|
14
|
-
partition(document, 'style', 'script').values.join(' ').strip.gsub(/\s+/, ' ')
|
15
|
-
end
|
16
|
-
|
17
|
-
def partition(node, *ignore_tags)
|
18
|
-
elements = { }
|
19
|
-
return elements if node.nil?
|
20
|
-
return elements if node.respond_to?('cdata?') and node.cdata?
|
21
|
-
return elements if node.respond_to?('comment?') and node.comment?
|
22
|
-
|
23
|
-
if node.kind_of?(Nokogiri::XML::Element) and ignore_tags and ignore_tags.size > 0
|
24
|
-
return elements if node.name =~ %r[#{ignore_tags.join('|')}]i
|
25
|
-
end
|
26
|
-
|
27
|
-
elements = { }
|
28
|
-
if node.kind_of?(Nokogiri::XML::Text)
|
29
|
-
elements[node.path.to_s] = node.text
|
30
|
-
return elements
|
31
|
-
end
|
32
|
-
node.children.each do |child|
|
33
|
-
elements.merge!(partition(child, *ignore_tags))
|
34
|
-
end
|
35
|
-
elements
|
36
|
-
end
|
37
|
-
|
38
|
-
end
|
39
|
-
|
40
|
-
def HtmlHelper.to_text(html)
|
41
|
-
transformer = HtmlTransformer.new(html)
|
42
|
-
transformer.to_text
|
43
|
-
end
|
44
|
-
|
45
|
-
|
46
|
-
def HtmlHelper.xpath(html, xpath)
|
47
|
-
document = Nokogiri::HTML.parse(html)
|
48
|
-
document.xpath(xpath).map { |m| m.to_html }
|
49
|
-
end
|
50
|
-
|
51
|
-
end
|
52
|
-
end
|
data/spec/html_helper_spec.rb
DELETED
@@ -1,42 +0,0 @@
|
|
1
|
-
require "omnivore/html_helper"
|
2
|
-
include Omnivore
|
3
|
-
|
4
|
-
|
5
|
-
|
6
|
-
describe HtmlHelper do
|
7
|
-
|
8
|
-
it "should match the correct xpath" do
|
9
|
-
content = %{
|
10
|
-
<html>
|
11
|
-
<head></head>
|
12
|
-
<body>
|
13
|
-
<div class="banner">
|
14
|
-
I don't want to see this.
|
15
|
-
</div>
|
16
|
-
<div class="content">
|
17
|
-
This is what I want to see.
|
18
|
-
</div>
|
19
|
-
</body>
|
20
|
-
</html>
|
21
|
-
}
|
22
|
-
matches = HtmlHelper.xpath(content, "//div[@class=\"content\"]")
|
23
|
-
matches.size.should == 1
|
24
|
-
end
|
25
|
-
|
26
|
-
|
27
|
-
it "should be able to extract text from markup" do
|
28
|
-
html = %{
|
29
|
-
<p>
|
30
|
-
Content may contain some additional markup, such as:
|
31
|
-
<ul>
|
32
|
-
<li>Ordered or unordered lists,</li>
|
33
|
-
<li><a href="#">Hyperlinks,</a></li>
|
34
|
-
<li>and Images
|
35
|
-
</ul>
|
36
|
-
</p>
|
37
|
-
}
|
38
|
-
text = HtmlHelper.to_text(html)
|
39
|
-
text.should == "Content may contain some additional markup, such as: Ordered or unordered lists, Hyperlinks, and Images"
|
40
|
-
end
|
41
|
-
|
42
|
-
end
|