web-page-parser 0.21 → 0.22
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.
|
@@ -1,3 +1,4 @@
|
|
|
1
|
+
# -*- coding: utf-8 -*-
|
|
1
2
|
$:.unshift File.join(File.dirname(__FILE__), '../../lib')
|
|
2
3
|
require 'spec/base_parser_spec'
|
|
3
4
|
require 'web-page-parser'
|
|
@@ -69,6 +70,18 @@ describe BbcNewsPageParserV4 do
|
|
|
69
70
|
@pa.content.size.should == 18
|
|
70
71
|
end
|
|
71
72
|
|
|
73
|
+
it "should parse the content of an article with market data" do
|
|
74
|
+
@pa = BbcNewsPageParserV4.new(:page => File.read('spec/fixtures/bbc_news/13293006.html'))
|
|
75
|
+
@pa.content.to_s.should_not =~ /Market Data/
|
|
76
|
+
@pa.content.to_s.should_not =~ /Last updated at/
|
|
77
|
+
@pa.content.size.should == 13
|
|
78
|
+
end
|
|
79
|
+
|
|
80
|
+
it "should ignore embedded-hyper content" do
|
|
81
|
+
@pa = BbcNewsPageParserV4.new(:page => File.read('spec/fixtures/bbc_news/12921632.html'))
|
|
82
|
+
@pa.content.to_s.should_not =~ /Fake and real quotes/
|
|
83
|
+
end
|
|
84
|
+
|
|
72
85
|
end
|
|
73
86
|
|
|
74
87
|
|
|
@@ -98,8 +111,6 @@ describe BbcNewsPageParserV3 do
|
|
|
98
111
|
@pa.content[1].should == "These include an £80m loan to Sheffield Forgemasters and new programmes for the young unemployed, Chief Secretary to the Treasury Danny Alexander told MPs."
|
|
99
112
|
@pa.content[2].should == 'Mr Alexander said the cuts were necessary to tackle the budget deficit and would be done in a "fair" way.'
|
|
100
113
|
end
|
|
101
|
-
|
|
102
|
-
|
|
103
114
|
|
|
104
115
|
end
|
|
105
116
|
|
data/spec/spec.opts
CHANGED
metadata
CHANGED
|
@@ -1,12 +1,12 @@
|
|
|
1
1
|
--- !ruby/object:Gem::Specification
|
|
2
2
|
name: web-page-parser
|
|
3
3
|
version: !ruby/object:Gem::Version
|
|
4
|
-
hash:
|
|
4
|
+
hash: 39
|
|
5
5
|
prerelease: false
|
|
6
6
|
segments:
|
|
7
7
|
- 0
|
|
8
|
-
-
|
|
9
|
-
version: "0.
|
|
8
|
+
- 22
|
|
9
|
+
version: "0.22"
|
|
10
10
|
platform: ruby
|
|
11
11
|
authors:
|
|
12
12
|
- John Leach
|
|
@@ -14,7 +14,7 @@ autorequire:
|
|
|
14
14
|
bindir: bin
|
|
15
15
|
cert_chain: []
|
|
16
16
|
|
|
17
|
-
date:
|
|
17
|
+
date: 2011-05-08 00:00:00 +01:00
|
|
18
18
|
default_executable:
|
|
19
19
|
dependencies:
|
|
20
20
|
- !ruby/object:Gem::Dependency
|
|
@@ -75,6 +75,8 @@ files:
|
|
|
75
75
|
- spec/fixtures/bbc_news/8063681.stm.html
|
|
76
76
|
- spec/fixtures/bbc_news/10249066.stm.html
|
|
77
77
|
- spec/fixtures/bbc_news/8011268.stm.html
|
|
78
|
+
- spec/fixtures/bbc_news/12921632.html
|
|
79
|
+
- spec/fixtures/bbc_news/13293006.html
|
|
78
80
|
- spec/fixtures/bbc_news/11125504.html
|
|
79
81
|
- spec/fixtures/bbc_news/6072486.stm.html
|
|
80
82
|
- spec/spec.opts
|
|
@@ -122,6 +124,8 @@ test_files:
|
|
|
122
124
|
- spec/fixtures/bbc_news/8063681.stm.html
|
|
123
125
|
- spec/fixtures/bbc_news/10249066.stm.html
|
|
124
126
|
- spec/fixtures/bbc_news/8011268.stm.html
|
|
127
|
+
- spec/fixtures/bbc_news/12921632.html
|
|
128
|
+
- spec/fixtures/bbc_news/13293006.html
|
|
125
129
|
- spec/fixtures/bbc_news/11125504.html
|
|
126
130
|
- spec/fixtures/bbc_news/6072486.stm.html
|
|
127
131
|
- spec/spec.opts
|