govspeak 5.3.0 → 5.4.0
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.
- checksums.yaml +4 -4
- data/CHANGELOG.md +3 -0
- data/lib/govspeak.rb +2 -2
- data/lib/govspeak/link_extractor.rb +12 -3
- data/lib/govspeak/version.rb +1 -1
- data/test/govspeak_link_extractor_test.rb +8 -1
- metadata +12 -12
checksums.yaml
CHANGED
@@ -1,7 +1,7 @@
|
|
1
1
|
---
|
2
2
|
SHA1:
|
3
|
-
metadata.gz:
|
4
|
-
data.tar.gz:
|
3
|
+
metadata.gz: 40e1c3a28d1f53eab54a104139116e0405c4450a
|
4
|
+
data.tar.gz: 89520dfd370e31fd803fa8d127fb546ab35521e3
|
5
5
|
SHA512:
|
6
|
-
metadata.gz:
|
7
|
-
data.tar.gz:
|
6
|
+
metadata.gz: 0afe56bcf4cd1160adb293500706b3ee3d2198b53792d18b7e8ad263ba7b81c3a770bfd63bd25e41b7ea9063d9243e31679e2fafa0b70e9a06e74f46eeb740f9
|
7
|
+
data.tar.gz: ff7509b60e910c01f83a8fbb7a0326b008898903186167609221c9f38fa805624f214139cea19afd2feb8181043aaf11faf455db51f5ca975078d7946c8b4bea
|
data/CHANGELOG.md
CHANGED
@@ -1,3 +1,6 @@
|
|
1
|
+
## 5.4.0
|
2
|
+
* Add an optional `website_root` argument to `Govspeak::Document#extracted_links` in order to get all links as fully qualified URLs [#122](https://github.com/alphagov/govspeak/pull/122)
|
3
|
+
|
1
4
|
## 5.3.0
|
2
5
|
* Add a link extraction class for finding links in documents [#120](https://github.com/alphagov/govspeak/pull/120)
|
3
6
|
|
data/lib/govspeak.rb
CHANGED
@@ -95,8 +95,8 @@ module Govspeak
|
|
95
95
|
Govspeak::StructuredHeaderExtractor.new(self).call
|
96
96
|
end
|
97
97
|
|
98
|
-
def extracted_links
|
99
|
-
Govspeak::LinkExtractor.new(self).call
|
98
|
+
def extracted_links(website_root: nil)
|
99
|
+
Govspeak::LinkExtractor.new(self, website_root: website_root).call
|
100
100
|
end
|
101
101
|
|
102
102
|
def preprocess(source)
|
@@ -1,7 +1,8 @@
|
|
1
1
|
module Govspeak
|
2
2
|
class LinkExtractor
|
3
|
-
def initialize(document)
|
3
|
+
def initialize(document, website_root: nil)
|
4
4
|
@document = document
|
5
|
+
@website_root = website_root
|
5
6
|
end
|
6
7
|
|
7
8
|
def call
|
@@ -10,8 +11,16 @@ module Govspeak
|
|
10
11
|
|
11
12
|
private
|
12
13
|
|
14
|
+
attr_reader :document, :website_root
|
15
|
+
|
13
16
|
def extract_links
|
14
|
-
document_anchors.map
|
17
|
+
document_anchors.map do |link|
|
18
|
+
if website_root && link['href'].start_with?('/')
|
19
|
+
"#{website_root}#{link['href']}"
|
20
|
+
else
|
21
|
+
link['href']
|
22
|
+
end
|
23
|
+
end
|
15
24
|
end
|
16
25
|
|
17
26
|
def document_anchors
|
@@ -22,7 +31,7 @@ module Govspeak
|
|
22
31
|
doc = Nokogiri::HTML::Document.new
|
23
32
|
doc.encoding = "UTF-8"
|
24
33
|
|
25
|
-
doc.fragment(
|
34
|
+
doc.fragment(document.to_html)
|
26
35
|
end
|
27
36
|
end
|
28
37
|
end
|
data/lib/govspeak/version.rb
CHANGED
@@ -12,6 +12,8 @@ class GovspeakLinkExtractorTest < Minitest::Test
|
|
12
12
|
[not_a_link](#somepage)
|
13
13
|
|
14
14
|
[mailto:](mailto:someone@www.example.com)
|
15
|
+
|
16
|
+
[absolute_path](/cais-trwydded-yrru-dros-dro)
|
15
17
|
}
|
16
18
|
end
|
17
19
|
|
@@ -24,7 +26,7 @@ class GovspeakLinkExtractorTest < Minitest::Test
|
|
24
26
|
end
|
25
27
|
|
26
28
|
test "Links are extracted from the body" do
|
27
|
-
expected_links =
|
29
|
+
expected_links = %w{http://www.example.com http://www.gov.com /cais-trwydded-yrru-dros-dro}
|
28
30
|
assert_equal expected_links, links
|
29
31
|
end
|
30
32
|
|
@@ -39,4 +41,9 @@ class GovspeakLinkExtractorTest < Minitest::Test
|
|
39
41
|
test "Links are not extracted if they begin with mailto:" do
|
40
42
|
refute_includes ["mailto:someone@www.example.com"], links
|
41
43
|
end
|
44
|
+
|
45
|
+
test "Absolute links are transformed to a url when website_root passed in" do
|
46
|
+
urls = doc.extracted_links(website_root: "http://www.example.com")
|
47
|
+
assert urls.include?("http://www.example.com/cais-trwydded-yrru-dros-dro")
|
48
|
+
end
|
42
49
|
end
|
metadata
CHANGED
@@ -1,14 +1,14 @@
|
|
1
1
|
--- !ruby/object:Gem::Specification
|
2
2
|
name: govspeak
|
3
3
|
version: !ruby/object:Gem::Version
|
4
|
-
version: 5.
|
4
|
+
version: 5.4.0
|
5
5
|
platform: ruby
|
6
6
|
authors:
|
7
7
|
- GOV.UK Dev
|
8
8
|
autorequire:
|
9
9
|
bindir: bin
|
10
10
|
cert_chain: []
|
11
|
-
date: 2018-01-
|
11
|
+
date: 2018-01-12 00:00:00.000000000 Z
|
12
12
|
dependencies:
|
13
13
|
- !ruby/object:Gem::Dependency
|
14
14
|
name: kramdown
|
@@ -320,19 +320,19 @@ signing_key:
|
|
320
320
|
specification_version: 4
|
321
321
|
summary: Markup language for single domain
|
322
322
|
test_files:
|
323
|
-
- test/govspeak_attachments_test.rb
|
324
|
-
- test/govspeak_attachments_inline_test.rb
|
325
|
-
- test/govspeak_link_test.rb
|
326
323
|
- test/govspeak_link_extractor_test.rb
|
324
|
+
- test/govspeak_structured_headers_test.rb
|
325
|
+
- test/govspeak_button_test.rb
|
326
|
+
- test/with_deep_merge_test.rb
|
327
|
+
- test/blockquote_extra_quote_remover_test.rb
|
328
|
+
- test/govspeak_test_helper.rb
|
329
|
+
- test/govspeak_link_test.rb
|
327
330
|
- test/govspeak_contacts_test.rb
|
328
331
|
- test/test_helper.rb
|
332
|
+
- test/html_validator_test.rb
|
333
|
+
- test/html_sanitizer_test.rb
|
334
|
+
- test/govspeak_attachments_inline_test.rb
|
329
335
|
- test/govspeak_test.rb
|
336
|
+
- test/govspeak_attachments_test.rb
|
330
337
|
- test/govspeak_attachments_image_test.rb
|
331
|
-
- test/with_deep_merge_test.rb
|
332
|
-
- test/html_validator_test.rb
|
333
|
-
- test/govspeak_structured_headers_test.rb
|
334
|
-
- test/blockquote_extra_quote_remover_test.rb
|
335
338
|
- test/presenters/h_card_presenter_test.rb
|
336
|
-
- test/govspeak_button_test.rb
|
337
|
-
- test/govspeak_test_helper.rb
|
338
|
-
- test/html_sanitizer_test.rb
|