sosowa 0.3 → 0.4
Sign up to get free protection for your applications and to get access to all the features.
- data/CHANGELOG.md +3 -0
- data/lib/sosowa/parser.rb +4 -1
- data/lib/sosowa/scheme.rb +18 -14
- data/lib/sosowa/version.rb +1 -1
- data/lib/sosowa.rb +2 -0
- metadata +4 -4
data/CHANGELOG.md
CHANGED
data/lib/sosowa/parser.rb
CHANGED
@@ -1,3 +1,5 @@
|
|
1
|
+
# coding: utf-8
|
2
|
+
|
1
3
|
module Sosowa
|
2
4
|
class Parser
|
3
5
|
|
@@ -42,7 +44,8 @@ module Sosowa
|
|
42
44
|
else
|
43
45
|
title = tr.search(%{td[@class="title cell_title"] > a}).inner_html.to_s.toutf8.strip
|
44
46
|
tags = tr.search(%{td[@class="title cell_title"] > a})[0].attributes["title"].value.split(" / ")
|
45
|
-
log = tr.search(%{td[@class="title cell_title"] > a})[0].attributes["href"].value.gsub(/log=(\d+)$/, '\1').to_i
|
47
|
+
#log = tr.search(%{td[@class="title cell_title"] > a})[0].attributes["href"].value.gsub(/log=(\d+)$/, '\1').to_i
|
48
|
+
log = parse_absolute_log_number(page)
|
46
49
|
key = tr.search(%{td[@class="title cell_title"] > a})[0].attributes["href"].value.gsub(/^.+key=(.+?)&.+$/, '\1').to_i
|
47
50
|
author = tr.search(%{td[@class="cell_author"]}).inner_html.to_s.toutf8.strip
|
48
51
|
created_at = Time.parse(tr.search(%{td[@class="cell_created"]}).inner_html.to_s.toutf8.strip)
|
data/lib/sosowa/scheme.rb
CHANGED
@@ -1,19 +1,23 @@
|
|
1
|
+
# coding: utf-8
|
2
|
+
|
1
3
|
module Sosowa
|
2
4
|
class Scheme
|
3
|
-
|
5
|
+
attr_reader :element
|
4
6
|
|
5
7
|
def initialize(element)
|
6
|
-
@element = element
|
8
|
+
@element = element
|
7
9
|
end
|
8
10
|
|
9
11
|
def method_missing(action, *args)
|
10
12
|
return @element[action.to_s.to_sym] rescue nil
|
11
13
|
end
|
12
14
|
|
13
|
-
public
|
14
|
-
|
15
15
|
def params() @element.keys.map{|k|k.to_sym} ; end
|
16
16
|
alias_method :available_methods, :params
|
17
|
+
|
18
|
+
def to_hash
|
19
|
+
@element
|
20
|
+
end
|
17
21
|
end
|
18
22
|
|
19
23
|
class Novel < Scheme
|
@@ -22,7 +26,7 @@ module Sosowa
|
|
22
26
|
@key = args[:key]
|
23
27
|
@agent = Mechanize.new
|
24
28
|
@page = nil
|
25
|
-
|
29
|
+
super(fetch(@log, @key))
|
26
30
|
end
|
27
31
|
|
28
32
|
def fetch(log, key)
|
@@ -31,7 +35,8 @@ module Sosowa
|
|
31
35
|
title = (@page/%{div[@class="header"] > h1})[0].inner_html.to_s.toutf8.strip
|
32
36
|
tags = (@page/%{dl[@class="info"][1] > dd > a}).map{|t| t.inner_html.to_s.toutf8 }
|
33
37
|
text = (@page/%{div[@class="contents ss"]})[0].inner_html.to_s.toutf8
|
34
|
-
|
38
|
+
aft = @page/%{div[@class="aft"]}
|
39
|
+
ps = (aft.size > 0) ? aft[0].inner_html.to_s.toutf8 : ""
|
35
40
|
author_name = (@page/%{div[@class="author"] b})[0].inner_html.to_s.toutf8
|
36
41
|
author_email, author_website = (@page/%{div[@class="author"] a}).map{|e| e.attributes["href"].value}
|
37
42
|
author = Author.new(:name => author_name, :email => (author_email ? author_email.gsub(/^mailto:/, "") : nil), :website => author_website)
|
@@ -50,7 +55,7 @@ module Sosowa
|
|
50
55
|
:name => bobj[0],
|
51
56
|
:created_at => Time.parse(bobj[1].gsub(/[^\/\d\s:]/, "")),
|
52
57
|
:text => element[1].inner_html.to_s.toutf8.strip
|
53
|
-
|
58
|
+
)
|
54
59
|
comments << comment
|
55
60
|
end
|
56
61
|
end
|
@@ -94,11 +99,11 @@ module Sosowa
|
|
94
99
|
end
|
95
100
|
|
96
101
|
class Comment < Scheme
|
97
|
-
|
102
|
+
|
98
103
|
end
|
99
104
|
|
100
105
|
class Author < Scheme
|
101
|
-
|
106
|
+
|
102
107
|
end
|
103
108
|
|
104
109
|
class Index < Scheme
|
@@ -108,13 +113,12 @@ module Sosowa
|
|
108
113
|
alias_method :get, :fetch
|
109
114
|
end
|
110
115
|
|
111
|
-
class Log <
|
116
|
+
class Log < Scheme
|
112
117
|
attr_reader :log
|
113
|
-
|
114
|
-
def initialize(
|
115
|
-
|
118
|
+
|
119
|
+
def initialize(element, log=0)
|
120
|
+
super(element)
|
116
121
|
@log = log
|
117
|
-
super(page)
|
118
122
|
end
|
119
123
|
|
120
124
|
def next_page
|
data/lib/sosowa/version.rb
CHANGED
data/lib/sosowa.rb
CHANGED
metadata
CHANGED
@@ -1,7 +1,7 @@
|
|
1
1
|
--- !ruby/object:Gem::Specification
|
2
2
|
name: sosowa
|
3
3
|
version: !ruby/object:Gem::Version
|
4
|
-
version: '0.
|
4
|
+
version: '0.4'
|
5
5
|
prerelease:
|
6
6
|
platform: ruby
|
7
7
|
authors:
|
@@ -9,11 +9,11 @@ authors:
|
|
9
9
|
autorequire:
|
10
10
|
bindir: bin
|
11
11
|
cert_chain: []
|
12
|
-
date: 2012-07-
|
12
|
+
date: 2012-07-22 00:00:00.000000000Z
|
13
13
|
dependencies:
|
14
14
|
- !ruby/object:Gem::Dependency
|
15
15
|
name: mechanize
|
16
|
-
requirement: &
|
16
|
+
requirement: &70270040436420 !ruby/object:Gem::Requirement
|
17
17
|
none: false
|
18
18
|
requirements:
|
19
19
|
- - ! '>='
|
@@ -21,7 +21,7 @@ dependencies:
|
|
21
21
|
version: '0'
|
22
22
|
type: :runtime
|
23
23
|
prerelease: false
|
24
|
-
version_requirements: *
|
24
|
+
version_requirements: *70270040436420
|
25
25
|
description: Sosowa Parser for Ruby
|
26
26
|
email:
|
27
27
|
- oame@oameya.com
|