onebox 1.7.2 → 1.7.3
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.
- checksums.yaml +4 -4
- data/lib/onebox/engine.rb +2 -0
- data/lib/onebox/engine/bandcamp_onebox.rb +1 -1
- data/lib/onebox/engine/five_hundred_px_onebox.rb +17 -0
- data/lib/onebox/engine/flickr_onebox.rb +17 -0
- data/lib/onebox/engine/soundcloud_onebox.rb +1 -1
- data/lib/onebox/engine/standard_embed.rb +5 -3
- data/lib/onebox/engine/wikipedia_onebox.rb +11 -11
- data/lib/onebox/helpers.rb +2 -8
- data/lib/onebox/version.rb +1 -1
- metadata +4 -2
checksums.yaml
CHANGED
@@ -1,7 +1,7 @@
|
|
1
1
|
---
|
2
2
|
SHA1:
|
3
|
-
metadata.gz:
|
4
|
-
data.tar.gz:
|
3
|
+
metadata.gz: 67ec2675eb6687bee752534aef6d1e610545717b
|
4
|
+
data.tar.gz: cd82452275130c85076cb9078f25650ac9c73444
|
5
5
|
SHA512:
|
6
|
-
metadata.gz:
|
7
|
-
data.tar.gz:
|
6
|
+
metadata.gz: 557f8fbdfd7e33e247b5cd82ffc89f1484b57ba075e6c1ea412382be1e846c9aaa08e14777dec3799b5ce76dc61881a01909665087b21c4bd1775aa84f18f65b
|
7
|
+
data.tar.gz: 77c61683b419f729ba6d3b506b8c3bd960ae40f5b52662fb8d6e150605caea7f33e83c8aac7a3fdf9f82f4d1c92c045a571e7df2d5633c4457691e0067b735ae
|
data/lib/onebox/engine.rb
CHANGED
@@ -171,3 +171,5 @@ require_relative "engine/asciinema_onebox"
|
|
171
171
|
require_relative "engine/mixcloud_onebox"
|
172
172
|
require_relative "engine/bandcamp_onebox"
|
173
173
|
require_relative "engine/coub_onebox"
|
174
|
+
require_relative "engine/flickr_onebox"
|
175
|
+
require_relative "engine/five_hundred_px_onebox"
|
@@ -0,0 +1,17 @@
|
|
1
|
+
module Onebox
|
2
|
+
module Engine
|
3
|
+
class FiveHundredPxOnebox
|
4
|
+
include Engine
|
5
|
+
include StandardEmbed
|
6
|
+
|
7
|
+
matches_regexp(/^https?:\/\/500px\.com\/photo\/\d+\//)
|
8
|
+
always_https
|
9
|
+
|
10
|
+
def to_html
|
11
|
+
og = get_opengraph
|
12
|
+
"<img src='#{og[:image]}' width='#{og[:image_width]}' height='#{og[:image_height]}' #{Helpers.title_attr(og)}>"
|
13
|
+
end
|
14
|
+
|
15
|
+
end
|
16
|
+
end
|
17
|
+
end
|
@@ -0,0 +1,17 @@
|
|
1
|
+
module Onebox
|
2
|
+
module Engine
|
3
|
+
class FlickrOnebox
|
4
|
+
include Engine
|
5
|
+
include StandardEmbed
|
6
|
+
|
7
|
+
matches_regexp(/^https?:\/\/www\.flickr\.com\/photos\//)
|
8
|
+
always_https
|
9
|
+
|
10
|
+
def to_html
|
11
|
+
og = get_opengraph
|
12
|
+
"<img src='#{og[:image]}' width='#{og[:image_width]}' height='#{og[:image_height]}' #{Helpers.title_attr(og)}>"
|
13
|
+
end
|
14
|
+
|
15
|
+
end
|
16
|
+
end
|
17
|
+
end
|
@@ -1,3 +1,5 @@
|
|
1
|
+
require "cgi"
|
2
|
+
|
1
3
|
module Onebox
|
2
4
|
module Engine
|
3
5
|
module StandardEmbed
|
@@ -93,14 +95,14 @@ module Onebox
|
|
93
95
|
html_doc.css('meta').each do |m|
|
94
96
|
if (m["property"] && m["property"][/^og:(.+)$/i]) || (m["name"] && m["name"][/^og:(.+)$/i])
|
95
97
|
value = (m["content"] || m["value"]).to_s
|
96
|
-
og[$1.tr('-:','_').to_sym] ||= value unless Onebox::Helpers::blank?(value)
|
98
|
+
og[$1.tr('-:','_').to_sym] ||= CGI.escapeHTML(value) unless Onebox::Helpers::blank?(value)
|
97
99
|
end
|
98
100
|
end
|
99
101
|
|
100
102
|
# Attempt to retrieve the title from the meta tag
|
101
103
|
title_element = html_doc.at_css('title')
|
102
104
|
if title_element && title_element.text
|
103
|
-
og[:title] ||= title_element.text unless Onebox::Helpers.blank?(title_element.text)
|
105
|
+
og[:title] ||= CGI.escapeHTML(title_element.text) unless Onebox::Helpers.blank?(title_element.text)
|
104
106
|
end
|
105
107
|
|
106
108
|
og
|
@@ -114,7 +116,7 @@ module Onebox
|
|
114
116
|
html_doc.css('meta').each do |m|
|
115
117
|
if (m["property"] && m["property"][/^twitter:(.+)$/i]) || (m["name"] && m["name"][/^twitter:(.+)$/i])
|
116
118
|
value = (m["content"] || m["value"]).to_s
|
117
|
-
twitter[$1.tr('-:','_').to_sym] ||= value unless Onebox::Helpers::blank?(value)
|
119
|
+
twitter[$1.tr('-:','_').to_sym] ||= CGI.escapeHTML(value) unless Onebox::Helpers::blank?(value)
|
118
120
|
end
|
119
121
|
end
|
120
122
|
|
@@ -5,7 +5,7 @@ module Onebox
|
|
5
5
|
include LayoutSupport
|
6
6
|
include HTML
|
7
7
|
|
8
|
-
matches_regexp(/^https
|
8
|
+
matches_regexp(/^https?:\/\/.*\.wikipedia\.(com|org)/)
|
9
9
|
always_https
|
10
10
|
|
11
11
|
private
|
@@ -22,23 +22,23 @@ module Onebox
|
|
22
22
|
end
|
23
23
|
|
24
24
|
unless m_url_hash.nil?
|
25
|
-
section_header_title = raw.xpath("//span[@id='#{m_url_hash_name}']")
|
25
|
+
section_header_title = raw.xpath("//span[@id='#{m_url_hash_name}']")
|
26
26
|
|
27
|
-
if section_header_title.empty?
|
27
|
+
if section_header_title.empty?
|
28
28
|
paras = raw.search("p") #default get all the paras
|
29
|
-
else
|
29
|
+
else
|
30
30
|
section_title_text = section_header_title.inner_text
|
31
31
|
section_header = section_header_title[0].parent #parent element of the section span element should be an <h3> node
|
32
32
|
cur_element = section_header
|
33
|
-
|
33
|
+
|
34
34
|
# p|text|div covers the general case. We assume presence of atleast 1 P node. if section has no P node we may end up with a P node from the next section.
|
35
35
|
# div tag is commonly used as an assets wraper in an article section. often as the first element holding an image.
|
36
36
|
# ul support will imporve the output generated for a section with a list as the main content (for example: an Author Bibliography, A musician Discography, etc)
|
37
|
-
first_p_found = nil
|
37
|
+
first_p_found = nil
|
38
38
|
while ( ((next_sibling = cur_element.next_sibling).name =~ /p|text|div|ul/) || first_p_found.nil? ) do #from section header get the next sibling until it is a breaker tag
|
39
39
|
cur_element = next_sibling
|
40
|
-
if (cur_element.name == "p" || cur_element.name == "ul") #we treat a list as we detect a p to avoid showing
|
41
|
-
first_p_found = true
|
40
|
+
if (cur_element.name == "p" || cur_element.name == "ul") #we treat a list as we detect a p to avoid showing
|
41
|
+
first_p_found = true
|
42
42
|
paras.push(cur_element)
|
43
43
|
end
|
44
44
|
end
|
@@ -52,16 +52,16 @@ module Onebox
|
|
52
52
|
while text.length < Onebox::LayoutSupport.max_text && cnt <= 3
|
53
53
|
break if cnt >= paras.size
|
54
54
|
text << " " unless cnt == 0
|
55
|
-
|
55
|
+
|
56
56
|
if paras[cnt].name =="ul" #Handle UL tag. Generate a textual ordered list (1.item | 2.item | 3.item). Unfourtently no newline allowed in output
|
57
57
|
li_index=1
|
58
58
|
list_items = []
|
59
59
|
paras[cnt].children.css("li").each {|li| list_items.push "#{li_index}." + li.inner_text ; li_index+=1}
|
60
60
|
paragraph = (list_items.join " |\n ")[0..Onebox::LayoutSupport.max_text]
|
61
61
|
else
|
62
|
-
paragraph = paras[cnt].inner_text[0..Onebox::LayoutSupport.max_text]
|
62
|
+
paragraph = paras[cnt].inner_text[0..Onebox::LayoutSupport.max_text]
|
63
63
|
end
|
64
|
-
|
64
|
+
|
65
65
|
paragraph.gsub!(/\[\d+\]/mi, "")
|
66
66
|
text << paragraph
|
67
67
|
cnt += 1
|
data/lib/onebox/helpers.rb
CHANGED
@@ -4,14 +4,8 @@ module Onebox
|
|
4
4
|
return {} if hash.nil?
|
5
5
|
|
6
6
|
hash.inject({}){|result, (key, value)|
|
7
|
-
new_key =
|
8
|
-
|
9
|
-
else key
|
10
|
-
end
|
11
|
-
new_value = case value
|
12
|
-
when Hash then symbolize_keys(value)
|
13
|
-
else value
|
14
|
-
end
|
7
|
+
new_key = key.is_a?(String) ? key.to_sym : key
|
8
|
+
new_value = value.is_a?(Hash) ? symbolize_keys(value) : value
|
15
9
|
result[new_key] = new_value
|
16
10
|
result
|
17
11
|
}
|
data/lib/onebox/version.rb
CHANGED
metadata
CHANGED
@@ -1,7 +1,7 @@
|
|
1
1
|
--- !ruby/object:Gem::Specification
|
2
2
|
name: onebox
|
3
3
|
version: !ruby/object:Gem::Version
|
4
|
-
version: 1.7.
|
4
|
+
version: 1.7.3
|
5
5
|
platform: ruby
|
6
6
|
authors:
|
7
7
|
- Joanna Zeta
|
@@ -10,7 +10,7 @@ authors:
|
|
10
10
|
autorequire:
|
11
11
|
bindir: bin
|
12
12
|
cert_chain: []
|
13
|
-
date: 2017-01-
|
13
|
+
date: 2017-01-11 00:00:00.000000000 Z
|
14
14
|
dependencies:
|
15
15
|
- !ruby/object:Gem::Dependency
|
16
16
|
name: multi_json
|
@@ -321,6 +321,8 @@ files:
|
|
321
321
|
- lib/onebox/engine/bandcamp_onebox.rb
|
322
322
|
- lib/onebox/engine/coub_onebox.rb
|
323
323
|
- lib/onebox/engine/douban_onebox.rb
|
324
|
+
- lib/onebox/engine/five_hundred_px_onebox.rb
|
325
|
+
- lib/onebox/engine/flickr_onebox.rb
|
324
326
|
- lib/onebox/engine/gfycat_onebox.rb
|
325
327
|
- lib/onebox/engine/giphy_onebox.rb
|
326
328
|
- lib/onebox/engine/github_blob_onebox.rb
|