lcbo 1.0.0 → 1.0.1
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/CHANGELOG.md +5 -0
- data/lib/lcbo/crawlkit/tag_helper.rb +22 -9
- data/lib/lcbo/version.rb +1 -1
- data/spec/crawlkit/tag_helper_spec.rb +2 -1
- data/spec/pages/product_pages.yml +2 -2
- data/spec/pages/store_pages.yml +1 -1
- metadata +3 -3
data/CHANGELOG.md
CHANGED
|
@@ -1,7 +1,7 @@
|
|
|
1
1
|
module LCBO
|
|
2
2
|
module CrawlKit
|
|
3
3
|
module TagHelper
|
|
4
|
-
DELETION_RE = /\
|
|
4
|
+
DELETION_RE = /\"|\\|\/|\(|\)|\[|\]|\./
|
|
5
5
|
WHITESPACE_RE = /\*|\+|\&|\_|\,|\s/
|
|
6
6
|
|
|
7
7
|
def self.flatten(values)
|
|
@@ -21,14 +21,27 @@ module LCBO
|
|
|
21
21
|
end
|
|
22
22
|
|
|
23
23
|
def self.stem(word)
|
|
24
|
-
|
|
25
|
-
|
|
26
|
-
|
|
27
|
-
|
|
28
|
-
|
|
29
|
-
|
|
30
|
-
|
|
31
|
-
|
|
24
|
+
split = lambda { |word|
|
|
25
|
+
if word.include?('-')
|
|
26
|
+
words = word.split('-')
|
|
27
|
+
a = words.dup
|
|
28
|
+
a << word
|
|
29
|
+
a << words.join
|
|
30
|
+
a
|
|
31
|
+
else
|
|
32
|
+
[word]
|
|
33
|
+
end
|
|
34
|
+
}
|
|
35
|
+
|
|
36
|
+
tokenize = lambda { |words|
|
|
37
|
+
words.reduce([]) do |tokens, word|
|
|
38
|
+
tokens << word
|
|
39
|
+
tokens << word.gsub("'", '') if word.include?("'")
|
|
40
|
+
tokens
|
|
41
|
+
end
|
|
42
|
+
}
|
|
43
|
+
|
|
44
|
+
tokenize.(split.(word))
|
|
32
45
|
end
|
|
33
46
|
|
|
34
47
|
def self.[](*values)
|
data/lib/lcbo/version.rb
CHANGED
|
@@ -6,8 +6,9 @@ describe LCBO::CrawlKit::TagHelper do
|
|
|
6
6
|
['Hello World'] => %w[hello world],
|
|
7
7
|
['Éve Picard'] => %w[éve picard eve],
|
|
8
8
|
['Hello Hello World'] => %w[hello world],
|
|
9
|
+
['É\'ve Picard-Rowe'] => %w[é've éve picard rowe picard-rowe picardrowe e've eve],
|
|
9
10
|
['Hello', 'Éve Picard'] => %w[hello éve picard eve],
|
|
10
|
-
['Hello', 'Éve-Picard'] => %w[hello éve picard évepicard eve evepicard],
|
|
11
|
+
['Hello', 'Éve-Picard'] => %w[hello éve picard éve-picard évepicard eve eve-picard evepicard],
|
|
11
12
|
['Hello', nil, 'World'] => %w[hello world]
|
|
12
13
|
}
|
|
13
14
|
|
|
@@ -48,7 +48,7 @@
|
|
|
48
48
|
:expectation:
|
|
49
49
|
:id: 18
|
|
50
50
|
:name: Heineken Lager
|
|
51
|
-
:tags: ["heineken", "lager", "beer", "netherlands", "heinekens", "brouwerijen", "nederland", "bv", "bottle"]
|
|
51
|
+
:tags: ["heineken", "lager", "beer", "netherlands", "heineken's", "heinekens", "brouwerijen", "nederland", "bv", "bottle"]
|
|
52
52
|
:price_in_cents: 1250
|
|
53
53
|
:regular_price_in_cents: 1350
|
|
54
54
|
:limited_time_offer_savings_in_cents: 100
|
|
@@ -336,7 +336,7 @@
|
|
|
336
336
|
:expectation:
|
|
337
337
|
:id: 56796
|
|
338
338
|
:name: Heineken Draught Keg
|
|
339
|
-
:tags: ["heineken", "draught", "keg", "beer", "lager", "netherlands", "heinekens", "brouwerijen", "nederland", "bv", "can"]
|
|
339
|
+
:tags: ["heineken", "draught", "keg", "beer", "lager", "netherlands", "heineken's", "heinekens", "brouwerijen", "nederland", "bv", "can"]
|
|
340
340
|
:price_in_cents: 2995
|
|
341
341
|
:regular_price_in_cents: 2995
|
|
342
342
|
:limited_time_offer_savings_in_cents: 0
|
data/spec/pages/store_pages.yml
CHANGED
|
@@ -7,7 +7,7 @@
|
|
|
7
7
|
:expectation:
|
|
8
8
|
:id: 444
|
|
9
9
|
:name: Kennedy & Sheppard
|
|
10
|
-
:tags: ["kennedy", "sheppard", "2356", "road", "agincourt", "mall", "toronto", "scarborough", "torontoscarborough", "m1t3h1"]
|
|
10
|
+
:tags: ["kennedy", "sheppard", "2356", "road", "agincourt", "mall", "toronto", "scarborough", "toronto-scarborough", "torontoscarborough", "m1t3h1"]
|
|
11
11
|
:address_line_1: 2356 Kennedy Road
|
|
12
12
|
:address_line_2: Agincourt Mall
|
|
13
13
|
:city: Toronto-Scarborough
|
metadata
CHANGED
|
@@ -5,8 +5,8 @@ version: !ruby/object:Gem::Version
|
|
|
5
5
|
segments:
|
|
6
6
|
- 1
|
|
7
7
|
- 0
|
|
8
|
-
-
|
|
9
|
-
version: 1.0.
|
|
8
|
+
- 1
|
|
9
|
+
version: 1.0.1
|
|
10
10
|
platform: ruby
|
|
11
11
|
authors:
|
|
12
12
|
- Carsten Nielsen
|
|
@@ -14,7 +14,7 @@ autorequire:
|
|
|
14
14
|
bindir: bin
|
|
15
15
|
cert_chain: []
|
|
16
16
|
|
|
17
|
-
date: 2011-01-
|
|
17
|
+
date: 2011-01-20 00:00:00 -05:00
|
|
18
18
|
default_executable:
|
|
19
19
|
dependencies:
|
|
20
20
|
- !ruby/object:Gem::Dependency
|