cyx-scraper 0.3.1 → 0.4.0
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/VERSION +1 -1
- data/lib/scraper.rb +2 -1
- data/lib/scraper/flickr.rb +70 -0
- data/lib/scraper/modules/video.rb +0 -31
- data/lib/scraper/modules/web.rb +39 -0
- data/lib/scraper/vimeo.rb +1 -0
- data/lib/scraper/youtube.rb +2 -1
- data/scraper.gemspec +7 -2
- data/test/fixtures/photostream.html +834 -0
- data/test/fixtures/show_photo.html +1329 -0
- data/test/flickr_test.rb +55 -0
- data/test/scraper_test.rb +23 -0
- metadata +7 -2
data/test/flickr_test.rb
ADDED
|
@@ -0,0 +1,55 @@
|
|
|
1
|
+
require 'test_helper'
|
|
2
|
+
|
|
3
|
+
class Scraper::FlickrTest < Test::Unit::TestCase
|
|
4
|
+
context "given a photostream url" do
|
|
5
|
+
setup do
|
|
6
|
+
@url = "http://www.flickr.com/photos/80186783@N00/"
|
|
7
|
+
@flickr = Scraper::Flickr.new(:url => @url)
|
|
8
|
+
Scraper::Modules::Web.stubs(:open).returns(
|
|
9
|
+
File.open(@@fixture_path + '/photostream.html', 'r')
|
|
10
|
+
)
|
|
11
|
+
end
|
|
12
|
+
|
|
13
|
+
should "return the photo stream's title" do
|
|
14
|
+
assert_equal "David Lazar's photostream", @flickr.title
|
|
15
|
+
end
|
|
16
|
+
|
|
17
|
+
should "say that it's a photostream" do
|
|
18
|
+
assert @flickr.photostream?
|
|
19
|
+
end
|
|
20
|
+
|
|
21
|
+
should "return the first photo as the thumbnail" do
|
|
22
|
+
@t = 'http://farm1.static.flickr.com/38/124484929_ed8c345cb9_m.jpg'
|
|
23
|
+
assert_equal @t, @flickr.thumbnail
|
|
24
|
+
end
|
|
25
|
+
|
|
26
|
+
should "have no description" do
|
|
27
|
+
assert_equal '', @flickr.description
|
|
28
|
+
end
|
|
29
|
+
end
|
|
30
|
+
|
|
31
|
+
context "given a photo url" do
|
|
32
|
+
setup do
|
|
33
|
+
@url = "http://www.flickr.com/photos/80186783@N00/124484929/"
|
|
34
|
+
Scraper::Modules::Web.stubs(:open).returns(
|
|
35
|
+
File.open(@@fixture_path + '/show_photo.html', 'r')
|
|
36
|
+
)
|
|
37
|
+
@flickr = Scraper::Flickr.new(:url => @url)
|
|
38
|
+
end
|
|
39
|
+
|
|
40
|
+
should "return the photo's title" do
|
|
41
|
+
assert_equal 'Debian Box', @flickr.title
|
|
42
|
+
end
|
|
43
|
+
|
|
44
|
+
should "return the photo's description" do
|
|
45
|
+
assert_equal 'An empty desktop.', @flickr.description
|
|
46
|
+
end
|
|
47
|
+
|
|
48
|
+
should "return the photo's thumbnail" do
|
|
49
|
+
@t = "http://farm1.static.flickr.com/38/124484929_ed8c345cb9_m.jpg"
|
|
50
|
+
assert_equal @t, @flickr.thumbnail
|
|
51
|
+
end
|
|
52
|
+
end
|
|
53
|
+
|
|
54
|
+
#
|
|
55
|
+
end
|
data/test/scraper_test.rb
CHANGED
|
@@ -35,6 +35,29 @@ class ScraperTest < Test::Unit::TestCase
|
|
|
35
35
|
end
|
|
36
36
|
end
|
|
37
37
|
|
|
38
|
+
context "Scraper( <flickr url> )" do
|
|
39
|
+
setup do
|
|
40
|
+
@url = "http://www.flickr.com/photos/80186783@N00/124484929/"
|
|
41
|
+
@scraper = Scraper(:url => @url)
|
|
42
|
+
end
|
|
43
|
+
|
|
44
|
+
should "return a Scraper::Flickr object" do
|
|
45
|
+
assert_instance_of Scraper::Flickr, @scraper
|
|
46
|
+
end
|
|
47
|
+
|
|
48
|
+
context "when the url doesn't have the photo part at the end" do
|
|
49
|
+
setup do
|
|
50
|
+
@url = "http://www.flickr.com/photos/80186783@N00"
|
|
51
|
+
@scraper = Scraper(:url => @url)
|
|
52
|
+
end
|
|
53
|
+
|
|
54
|
+
should "return a Scraper::Flickr object still" do
|
|
55
|
+
assert_instance_of Scraper::Flickr, @scraper
|
|
56
|
+
end
|
|
57
|
+
end
|
|
58
|
+
end
|
|
59
|
+
|
|
60
|
+
|
|
38
61
|
context "given an article from A-List-Apart" do
|
|
39
62
|
setup do
|
|
40
63
|
@article = fixture_file('unwebbable.html')
|
metadata
CHANGED
|
@@ -1,7 +1,7 @@
|
|
|
1
1
|
--- !ruby/object:Gem::Specification
|
|
2
2
|
name: cyx-scraper
|
|
3
3
|
version: !ruby/object:Gem::Version
|
|
4
|
-
version: 0.
|
|
4
|
+
version: 0.4.0
|
|
5
5
|
platform: ruby
|
|
6
6
|
authors:
|
|
7
7
|
- Cyril David
|
|
@@ -9,7 +9,7 @@ autorequire:
|
|
|
9
9
|
bindir: bin
|
|
10
10
|
cert_chain: []
|
|
11
11
|
|
|
12
|
-
date: 2009-
|
|
12
|
+
date: 2009-08-01 00:00:00 -07:00
|
|
13
13
|
default_executable:
|
|
14
14
|
dependencies: []
|
|
15
15
|
|
|
@@ -31,6 +31,7 @@ files:
|
|
|
31
31
|
- VERSION
|
|
32
32
|
- lib/scraper.rb
|
|
33
33
|
- lib/scraper/article.rb
|
|
34
|
+
- lib/scraper/flickr.rb
|
|
34
35
|
- lib/scraper/modules.rb
|
|
35
36
|
- lib/scraper/modules/video.rb
|
|
36
37
|
- lib/scraper/modules/web.rb
|
|
@@ -41,8 +42,11 @@ files:
|
|
|
41
42
|
- test/fixtures/5826468.html
|
|
42
43
|
- test/fixtures/dLO2s7SDHJo.html
|
|
43
44
|
- test/fixtures/non-article.html
|
|
45
|
+
- test/fixtures/photostream.html
|
|
44
46
|
- test/fixtures/scraped.html
|
|
47
|
+
- test/fixtures/show_photo.html
|
|
45
48
|
- test/fixtures/unwebbable.html
|
|
49
|
+
- test/flickr_test.rb
|
|
46
50
|
- test/scraper_test.rb
|
|
47
51
|
- test/test_helper.rb
|
|
48
52
|
- test/vimeo_test.rb
|
|
@@ -77,6 +81,7 @@ specification_version: 3
|
|
|
77
81
|
summary: TODO
|
|
78
82
|
test_files:
|
|
79
83
|
- test/article_test.rb
|
|
84
|
+
- test/flickr_test.rb
|
|
80
85
|
- test/scraper_test.rb
|
|
81
86
|
- test/test_helper.rb
|
|
82
87
|
- test/vimeo_test.rb
|