otg-microformat 0.0.2 → 0.0.3
Sign up to get free protection for your applications and to get access to all the features.
- data/.gitignore +1 -0
- data/lib/microformat.rb +25 -8
- data/lib/microformat/version.rb +1 -1
- data/microformat.gemspec +1 -0
- data/spec/schema_org_spec.rb +99 -0
- metadata +21 -9
data/.gitignore
CHANGED
data/lib/microformat.rb
CHANGED
@@ -4,15 +4,26 @@ require 'andand'
|
|
4
4
|
|
5
5
|
module Microformat
|
6
6
|
class ItemProp
|
7
|
-
def self.parse(node)
|
7
|
+
def self.parse(node, strict)
|
8
8
|
# If the element has no itemprop attribute
|
9
9
|
# The attribute must return null on getting and must throw an INVALID_ACCESS_ERR exception on setting.
|
10
10
|
return nil unless node.attribute('itemprop')
|
11
11
|
|
12
12
|
# If the element has an itemscope attribute
|
13
13
|
# The attribute must return the element itself on getting and must throw an INVALID_ACCESS_ERR exception on setting.
|
14
|
-
return ItemScope.new(node) if node.attribute('itemscope')
|
14
|
+
return ItemScope.new(node, strict) if node.attribute('itemscope')
|
15
15
|
|
16
|
+
if strict
|
17
|
+
parse_strict node
|
18
|
+
else
|
19
|
+
parse_weak node
|
20
|
+
end
|
21
|
+
end
|
22
|
+
|
23
|
+
private
|
24
|
+
ATTRIBUTES = ['content', 'src', 'href', 'data', 'datetime']
|
25
|
+
|
26
|
+
def self.parse_strict(node)
|
16
27
|
# If the element is a meta element
|
17
28
|
# The attribute must act as it would if it was reflecting the element's content content attribute
|
18
29
|
return node.attribute('content').andand.value if node.name == 'meta'
|
@@ -37,15 +48,20 @@ module Microformat
|
|
37
48
|
# The attribute must act the same as the element's textContent attribute.
|
38
49
|
return node.text.chomp.strip
|
39
50
|
end
|
51
|
+
|
52
|
+
def self.parse_weak(node)
|
53
|
+
ATTRIBUTES.map { |attr| node.attribute(attr).andand.value }.compact.first || node.text.chomp.strip
|
54
|
+
end
|
40
55
|
end
|
41
56
|
|
42
57
|
class ItemScope
|
43
58
|
attr_reader :type, :id
|
44
59
|
|
45
|
-
def initialize(node)
|
60
|
+
def initialize(node, strict)
|
46
61
|
@type = attr 'itemtype', node
|
47
62
|
@id = attr 'itemid', node
|
48
63
|
@properties = {}
|
64
|
+
@strict = strict
|
49
65
|
|
50
66
|
parse_elements node.search './*'
|
51
67
|
end
|
@@ -63,14 +79,14 @@ module Microformat
|
|
63
79
|
def parse_elements(elements)
|
64
80
|
elements.each do |el|
|
65
81
|
itemprop = attr('itemprop', el)
|
66
|
-
prop = ItemProp.parse el
|
82
|
+
prop = ItemProp.parse el, @strict
|
67
83
|
|
68
84
|
if prop
|
69
85
|
@properties[itemprop] ||= []
|
70
86
|
@properties[itemprop] << prop
|
71
|
-
else
|
72
|
-
parse_elements el.search('./*')
|
73
87
|
end
|
88
|
+
|
89
|
+
parse_elements el.search('./*')
|
74
90
|
end
|
75
91
|
end
|
76
92
|
end
|
@@ -88,10 +104,11 @@ module Microformat
|
|
88
104
|
end
|
89
105
|
|
90
106
|
class SchemaOrg
|
91
|
-
def self.parse(html)
|
107
|
+
def self.parse(html, opts={})
|
108
|
+
strict = opts.has_key? :strict ? opts[:strict] : true
|
92
109
|
html = Nokogiri::HTML.parse html unless html.respond_to? :search
|
93
110
|
scopes = html.search('//*[@itemscope and not(@itemprop)]')
|
94
|
-
.map { |node| ItemScope.new node }
|
111
|
+
.map { |node| ItemScope.new node, strict }
|
95
112
|
|
96
113
|
if scopes.any?
|
97
114
|
ItemDocument.new scopes
|
data/lib/microformat/version.rb
CHANGED
data/microformat.gemspec
CHANGED
@@ -0,0 +1,99 @@
|
|
1
|
+
require 'rspec'
|
2
|
+
require 'microformat'
|
3
|
+
|
4
|
+
describe Microformat::SchemaOrg do
|
5
|
+
describe 'a strict implementation' do
|
6
|
+
let(:html) {
|
7
|
+
<<-HTML
|
8
|
+
<div itemscope itemtype="http://schema.org/Product">
|
9
|
+
<h1 itemprop="name">Product</h1>
|
10
|
+
<div itemprop="offers" itemscope itemtype="http://schema.org/Offer">
|
11
|
+
<meta itemprop="currency" content="AUD" />
|
12
|
+
<meta itemprop="availability" content="OutOfStock" />
|
13
|
+
Price: $<span itemprop="price">12.99</span>
|
14
|
+
</div>
|
15
|
+
</div>
|
16
|
+
HTML
|
17
|
+
}
|
18
|
+
let(:result) { Microformat::SchemaOrg.parse html, strict: true }
|
19
|
+
let(:product) { result.products.first }
|
20
|
+
|
21
|
+
it 'should find a product' do
|
22
|
+
product.should_not be_nil
|
23
|
+
end
|
24
|
+
|
25
|
+
it 'should parse name' do
|
26
|
+
product['name'].should eq ['Product']
|
27
|
+
end
|
28
|
+
|
29
|
+
it 'should parse offer' do
|
30
|
+
product['offers'].first['currency'].should eq ['AUD']
|
31
|
+
product['offers'].first['availability'].should eq ['OutOfStock']
|
32
|
+
product['offers'].first['price'].should eq ['12.99']
|
33
|
+
end
|
34
|
+
end
|
35
|
+
|
36
|
+
describe 'a weak implementation' do
|
37
|
+
let(:html) {
|
38
|
+
<<-HTML
|
39
|
+
<div itemscope itemtype="http://schema.org/Product">
|
40
|
+
<h1 itemprop="name">Product</h1>
|
41
|
+
<div itemprop="offers" itemscope itemtype="http://schema.org/Offer">
|
42
|
+
<span itemprop="currency" content="AUD" />
|
43
|
+
<span itemprop="availability" href="OutOfStock" />
|
44
|
+
Price: $<span itemprop="price">12.99</span>
|
45
|
+
</div>
|
46
|
+
</div>
|
47
|
+
HTML
|
48
|
+
}
|
49
|
+
let(:result) { Microformat::SchemaOrg.parse html, strict: false }
|
50
|
+
let(:product) { result.products.first }
|
51
|
+
|
52
|
+
it 'should find a product' do
|
53
|
+
product.should_not be_nil
|
54
|
+
end
|
55
|
+
|
56
|
+
it 'should parse name' do
|
57
|
+
product['name'].should eq ['Product']
|
58
|
+
end
|
59
|
+
|
60
|
+
it 'should parse offer' do
|
61
|
+
product['offers'].first['currency'].should eq ['AUD']
|
62
|
+
product['offers'].first['availability'].should eq ['OutOfStock']
|
63
|
+
product['offers'].first['price'].should eq ['12.99']
|
64
|
+
end
|
65
|
+
end
|
66
|
+
|
67
|
+
describe 'a weird weak implementation' do
|
68
|
+
let(:html) {
|
69
|
+
<<-HTML
|
70
|
+
<div itemscope itemtype="http://schema.org/Product">
|
71
|
+
<h1 itemprop="name">Product</h1>
|
72
|
+
<div itemprop="offers" itemscope itemtype="http://schema.org/Offer">
|
73
|
+
<span itemprop="currency" content="AUD">
|
74
|
+
<span itemprop="availability" href="OutOfStock">
|
75
|
+
Price: $<span itemprop="price">12.99</span>
|
76
|
+
</span>
|
77
|
+
</span>
|
78
|
+
</div>
|
79
|
+
</div>
|
80
|
+
HTML
|
81
|
+
}
|
82
|
+
let(:result) { Microformat::SchemaOrg.parse html, strict: false }
|
83
|
+
let(:product) { result.products.first }
|
84
|
+
|
85
|
+
it 'should find a product' do
|
86
|
+
product.should_not be_nil
|
87
|
+
end
|
88
|
+
|
89
|
+
it 'should parse name' do
|
90
|
+
product['name'].should eq ['Product']
|
91
|
+
end
|
92
|
+
|
93
|
+
it 'should parse offer' do
|
94
|
+
product['offers'].first['currency'].should eq ['AUD']
|
95
|
+
product['offers'].first['availability'].should eq ['OutOfStock']
|
96
|
+
product['offers'].first['price'].should eq ['12.99']
|
97
|
+
end
|
98
|
+
end
|
99
|
+
end
|
metadata
CHANGED
@@ -1,7 +1,7 @@
|
|
1
1
|
--- !ruby/object:Gem::Specification
|
2
2
|
name: otg-microformat
|
3
3
|
version: !ruby/object:Gem::Version
|
4
|
-
version: 0.0.
|
4
|
+
version: 0.0.3
|
5
5
|
prerelease:
|
6
6
|
platform: ruby
|
7
7
|
authors:
|
@@ -9,7 +9,7 @@ authors:
|
|
9
9
|
autorequire:
|
10
10
|
bindir: bin
|
11
11
|
cert_chain: []
|
12
|
-
date: 2013-05-
|
12
|
+
date: 2013-05-12 00:00:00.000000000 Z
|
13
13
|
dependencies:
|
14
14
|
- !ruby/object:Gem::Dependency
|
15
15
|
name: nokogiri
|
@@ -75,6 +75,22 @@ dependencies:
|
|
75
75
|
- - ! '>='
|
76
76
|
- !ruby/object:Gem::Version
|
77
77
|
version: '0'
|
78
|
+
- !ruby/object:Gem::Dependency
|
79
|
+
name: rspec
|
80
|
+
requirement: !ruby/object:Gem::Requirement
|
81
|
+
none: false
|
82
|
+
requirements:
|
83
|
+
- - ! '>='
|
84
|
+
- !ruby/object:Gem::Version
|
85
|
+
version: '0'
|
86
|
+
type: :development
|
87
|
+
prerelease: false
|
88
|
+
version_requirements: !ruby/object:Gem::Requirement
|
89
|
+
none: false
|
90
|
+
requirements:
|
91
|
+
- - ! '>='
|
92
|
+
- !ruby/object:Gem::Version
|
93
|
+
version: '0'
|
78
94
|
description: A basic microformat parser for Ruby.
|
79
95
|
email:
|
80
96
|
- james@onthegame.com.au
|
@@ -90,6 +106,7 @@ files:
|
|
90
106
|
- lib/microformat.rb
|
91
107
|
- lib/microformat/version.rb
|
92
108
|
- microformat.gemspec
|
109
|
+
- spec/schema_org_spec.rb
|
93
110
|
homepage: http://www.onthegame.com.au/about/opensource
|
94
111
|
licenses:
|
95
112
|
- MIT
|
@@ -103,22 +120,17 @@ required_ruby_version: !ruby/object:Gem::Requirement
|
|
103
120
|
- - ! '>='
|
104
121
|
- !ruby/object:Gem::Version
|
105
122
|
version: '0'
|
106
|
-
segments:
|
107
|
-
- 0
|
108
|
-
hash: 640169864049438702
|
109
123
|
required_rubygems_version: !ruby/object:Gem::Requirement
|
110
124
|
none: false
|
111
125
|
requirements:
|
112
126
|
- - ! '>='
|
113
127
|
- !ruby/object:Gem::Version
|
114
128
|
version: '0'
|
115
|
-
segments:
|
116
|
-
- 0
|
117
|
-
hash: 640169864049438702
|
118
129
|
requirements: []
|
119
130
|
rubyforge_project:
|
120
131
|
rubygems_version: 1.8.25
|
121
132
|
signing_key:
|
122
133
|
specification_version: 3
|
123
134
|
summary: Simple Microformat parser
|
124
|
-
test_files:
|
135
|
+
test_files:
|
136
|
+
- spec/schema_org_spec.rb
|