factbook 1.2.0 → 1.2.1
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +4 -4
- data/Manifest.txt +1 -0
- data/README.md +6 -38
- data/Rakefile +1 -1
- data/lib/factbook/sanitizer.rb +19 -0
- data/lib/factbook/version.rb +1 -1
- data/test/data/src/ag.html +716 -0
- data/test/test_json.rb +7 -6
- data/test/test_sanitizer.rb +4 -1
- data/test/test_sanitizer_regex.rb +26 -1
- metadata +4 -3
data/test/test_json.rb
CHANGED
|
@@ -13,8 +13,9 @@ class TestJson < MiniTest::Test
|
|
|
13
13
|
|
|
14
14
|
def test_json
|
|
15
15
|
Dir.mkdir( 'tmp' ) unless Dir.exists?( 'tmp' )
|
|
16
|
-
|
|
16
|
+
|
|
17
17
|
codes = [ 'au',
|
|
18
|
+
'ag',
|
|
18
19
|
'be',
|
|
19
20
|
#'br',
|
|
20
21
|
#'mx',
|
|
@@ -25,13 +26,13 @@ class TestJson < MiniTest::Test
|
|
|
25
26
|
]
|
|
26
27
|
|
|
27
28
|
codes.each do |code|
|
|
28
|
-
|
|
29
|
-
html = File.read( "#{Factbook.root}/test/data/src/#{code}.html" )
|
|
30
|
-
page = Factbook::Page.new( code, html: html )
|
|
31
|
-
|
|
29
|
+
|
|
30
|
+
html = File.read( "#{Factbook.root}/test/data/src/#{code}.html" )
|
|
31
|
+
page = Factbook::Page.new( code, html: html )
|
|
32
|
+
|
|
32
33
|
h = page.data
|
|
33
34
|
pp h
|
|
34
|
-
|
|
35
|
+
|
|
35
36
|
### save to json
|
|
36
37
|
puts "saving a copy to #{code}.json for debugging"
|
|
37
38
|
File.open( "tmp/#{code}.json", 'w' ) do |f|
|
data/test/test_sanitizer.rb
CHANGED
|
@@ -12,7 +12,10 @@ class TestSanitizer < MiniTest::Test
|
|
|
12
12
|
|
|
13
13
|
def test_sanitize
|
|
14
14
|
|
|
15
|
-
|
|
15
|
+
## austria (au)
|
|
16
|
+
## algeria (ag)
|
|
17
|
+
## belgium (be)
|
|
18
|
+
['au','ag','be'].each do |cnty|
|
|
16
19
|
|
|
17
20
|
## use/fix: ASCII-8BIT (e.g.keep as is) -???
|
|
18
21
|
html_ascii = File.read( "#{Factbook.root}/test/data/src/#{cnty}.html" ) ## fix/todo: use ASCII8BIT/binary reader ??
|
|
@@ -59,6 +59,31 @@ HTML
|
|
|
59
59
|
pp m
|
|
60
60
|
|
|
61
61
|
assert m.nil? == false
|
|
62
|
-
|
|
62
|
+
end # method test_pop_pyramid
|
|
63
|
+
|
|
64
|
+
|
|
65
|
+
def test_rel_affiliation
|
|
66
|
+
|
|
67
|
+
html =<<HTML
|
|
68
|
+
<div class='disTable relAffiliation'><span class='category tCell' style='margin-bottom:0px; vertical-align:bottom;'>religious affiliation:</span>
|
|
69
|
+
<span class="tCell"><a data-toggle="modal" href="#relAffiliationModal"><img title="" src="../graphics/middle-east-religion-icon.jpg" style="cursor:pointer; border: 0px solid #CCC;"></span></a></div>
|
|
70
|
+
|
|
71
|
+
<div class="modal fade" id="relAffiliationModal" role="dialog">
|
|
72
|
+
<div class="wfb-modal-dialog">
|
|
73
|
+
<div class="modal-content" >
|
|
74
|
+
<div class="wfb-modal-header" style="border-radius: 4px; font-family: Verdana,Arial,sans-serif; font-size: 14px !important; font-weight: bold; padding: 0.4em 16px 0.4em 1em; background: #cccccc url("..images/ui-bg_highlight-soft_75_cccccc_1x100.png") repeat-x scroll 50% 50%;" >
|
|
75
|
+
<span style="font-size: 14px !important; margin: 0.1em 16px 0.1em 0;" class="modal-title wfb-title">The World Factbook</span><span style="float: right; margin-top: -4px;">
|
|
76
|
+
<button type="button" class="close" title="close" data-dismiss="modal">×</button></span>
|
|
77
|
+
</div>
|
|
78
|
+
<div class="wfb-modal-body">
|
|
79
|
+
...
|
|
80
|
+
<div id='field'
|
|
81
|
+
HTML
|
|
82
|
+
|
|
83
|
+
m = Factbook::Sanitizer::REL_AFFILIATION_CATEGORY_REGEX.match( html )
|
|
84
|
+
pp m
|
|
85
|
+
|
|
86
|
+
assert m.nil? == false
|
|
87
|
+
end # method test_rel_affiliation
|
|
63
88
|
|
|
64
89
|
end # class TestSanitizerRegex
|
metadata
CHANGED
|
@@ -1,14 +1,14 @@
|
|
|
1
1
|
--- !ruby/object:Gem::Specification
|
|
2
2
|
name: factbook
|
|
3
3
|
version: !ruby/object:Gem::Version
|
|
4
|
-
version: 1.2.
|
|
4
|
+
version: 1.2.1
|
|
5
5
|
platform: ruby
|
|
6
6
|
authors:
|
|
7
7
|
- Gerald Bauer
|
|
8
8
|
autorequire:
|
|
9
9
|
bindir: bin
|
|
10
10
|
cert_chain: []
|
|
11
|
-
date: 2016-11-
|
|
11
|
+
date: 2016-11-06 00:00:00.000000000 Z
|
|
12
12
|
dependencies:
|
|
13
13
|
- !ruby/object:Gem::Dependency
|
|
14
14
|
name: logutils
|
|
@@ -148,6 +148,7 @@ files:
|
|
|
148
148
|
- test/data/be.html
|
|
149
149
|
- test/data/be.yml
|
|
150
150
|
- test/data/json/au.json
|
|
151
|
+
- test/data/src/ag.html
|
|
151
152
|
- test/data/src/au-2015-09-24.html
|
|
152
153
|
- test/data/src/au.html
|
|
153
154
|
- test/data/src/be-2015-09-24.html
|
|
@@ -169,7 +170,7 @@ files:
|
|
|
169
170
|
- test/test_page.rb
|
|
170
171
|
- test/test_sanitizer.rb
|
|
171
172
|
- test/test_sanitizer_regex.rb
|
|
172
|
-
homepage: https://github.com/
|
|
173
|
+
homepage: https://github.com/factbook/factbook
|
|
173
174
|
licenses:
|
|
174
175
|
- Public Domain
|
|
175
176
|
metadata: {}
|