oald_parser 0.1.1 → 0.1.2
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/lib/oald_parser/facade.rb +15 -3
- data/lib/oald_parser/word_extractor.rb +19 -0
- data/lib/oald_parser.rb +5 -0
- metadata +4 -2
data/lib/oald_parser/facade.rb
CHANGED
@@ -2,24 +2,27 @@ require_relative 'oald_parser_exception'
|
|
2
2
|
require_relative 'page_parser'
|
3
3
|
require_relative 'formatter'
|
4
4
|
require_relative 'page_downloader'
|
5
|
+
require_relative 'word_extractor'
|
5
6
|
|
6
7
|
module OaldParser
|
7
8
|
class Facade
|
8
|
-
def initialize(downloader, parser, formatter)
|
9
|
+
def initialize(downloader, parser, formatter, extractor)
|
9
10
|
@downloader = downloader
|
10
11
|
@parser = parser
|
11
12
|
@formatter = formatter
|
13
|
+
@extractor = extractor
|
12
14
|
end
|
13
15
|
|
14
16
|
def self.create_facade
|
15
17
|
downloader = PageDownloader.new('http://www.oup.com/oald-bin/web_getald7index1a.pl')
|
16
18
|
parser = PageParser.new
|
17
19
|
formatter = Formatter.new(lines: 15)
|
18
|
-
|
20
|
+
extractor = WordExtractor.new
|
21
|
+
Facade.new(downloader, parser, formatter, extractor)
|
19
22
|
end
|
20
23
|
|
21
24
|
def describe(args)
|
22
|
-
word = args
|
25
|
+
word = get_word(args)
|
23
26
|
raise OaldParserException.new(OaldParserException::INTERNAL) unless word
|
24
27
|
|
25
28
|
page = @downloader.download(word)
|
@@ -33,5 +36,14 @@ module OaldParser
|
|
33
36
|
|
34
37
|
formatted
|
35
38
|
end
|
39
|
+
|
40
|
+
private
|
41
|
+
def get_word(args)
|
42
|
+
if args[:word]
|
43
|
+
args[:word]
|
44
|
+
elsif args[:str]
|
45
|
+
@extractor.extract(args[:str])
|
46
|
+
end
|
47
|
+
end
|
36
48
|
end
|
37
49
|
end
|
@@ -0,0 +1,19 @@
|
|
1
|
+
module OaldParser
|
2
|
+
class WordExtractor
|
3
|
+
def extract(str)
|
4
|
+
res = remove_unused_words(str)
|
5
|
+
find_first_big_word(res)
|
6
|
+
end
|
7
|
+
|
8
|
+
private
|
9
|
+
def remove_unused_words(str)
|
10
|
+
str.gsub(' ', ' ').
|
11
|
+
gsub(/\Aa | a |\Aan | an |\Athe | the |\[.*\]|\(.*\)| adj | adj\z| adv | adv\z/i, '')
|
12
|
+
end
|
13
|
+
|
14
|
+
def find_first_big_word(str)
|
15
|
+
parts = str.split(' ')
|
16
|
+
parts.size > 1 ? parts.find{|w| w.size > 2} : str.strip
|
17
|
+
end
|
18
|
+
end
|
19
|
+
end
|
data/lib/oald_parser.rb
ADDED
metadata
CHANGED
@@ -5,8 +5,8 @@ version: !ruby/object:Gem::Version
|
|
5
5
|
segments:
|
6
6
|
- 0
|
7
7
|
- 1
|
8
|
-
-
|
9
|
-
version: 0.1.
|
8
|
+
- 2
|
9
|
+
version: 0.1.2
|
10
10
|
platform: ruby
|
11
11
|
authors:
|
12
12
|
- Victor Savkin
|
@@ -56,6 +56,8 @@ files:
|
|
56
56
|
- lib/oald_parser/oald_parser_exception.rb
|
57
57
|
- lib/oald_parser/page_parser.rb
|
58
58
|
- lib/oald_parser/facade.rb
|
59
|
+
- lib/oald_parser/word_extractor.rb
|
60
|
+
- lib/oald_parser.rb
|
59
61
|
has_rdoc: true
|
60
62
|
homepage:
|
61
63
|
licenses: []
|