Class: AikatsuCalendar::Scraper

Inherits:
Object
  • Object
show all
Defined in:
lib/aikatsu_calendar/scraper.rb

Instance Attribute Summary collapse

Class Method Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize ⇒ Scraper

Returns a new instance of Scraper.



22
23
24
 
# File 'lib/aikatsu_calendar/scraper.rb', line 22

def initialize
  @schedules = []
end

Instance Attribute Details

#day ⇒ Object

Returns the value of attribute day.



11
12
13
 
# File 'lib/aikatsu_calendar/scraper.rb', line 11

def day
  @day
end

#month ⇒ Object

Returns the value of attribute month.



11
12
13
 
# File 'lib/aikatsu_calendar/scraper.rb', line 11

def month
  @month
end

#schedules ⇒ Object

Returns the value of attribute schedules.



11
12
13
 
# File 'lib/aikatsu_calendar/scraper.rb', line 11

def schedules
  @schedules
end

#year ⇒ Object

Returns the value of attribute year.



11
12
13
 
# File 'lib/aikatsu_calendar/scraper.rb', line 11

def year
  @year
end

Class Method Details

.scrape(path = AikatsuCalendar::URL) ⇒ Object 



13
14
15
16
17
18
19
20
 
# File 'lib/aikatsu_calendar/scraper.rb', line 13

def self.scrape(path=AikatsuCalendar::URL)
  scraper = new()
  doc = open(path) {|f| Nokogiri::HTML.parse(f) }
  scraper.feed(doc)
  scraper.schedules.uniq do |x|
    [x[:type], x[:content], x[:date_from], x[:date_until]]
  end
end

Instance Method Details

#class_to_type(s) ⇒ Object 



94
95
96
97
98
99
100
 
# File 'lib/aikatsu_calendar/scraper.rb', line 94

def class_to_type(s)
  if s =~ /schedule-(\w+)/
    $1
  else
    nil
  end
end

#feed(doc) ⇒ Object 



26
27
28
29
30
31
 
# File 'lib/aikatsu_calendar/scraper.rb', line 26

def feed(doc)
  container = doc.at_css(".info-schedule")
  container.css('table').each do |table|
    feed_table(table)
  end
end

#feed_item(p) ⇒ Object 



56
57
58
 
# File 'lib/aikatsu_calendar/scraper.rb', line 56

def feed_item(p)
  @schedules << parse_item(p)
end

#feed_row(tr) ⇒ Object 



45
46
47
48
49
50
51
52
53
54
 
# File 'lib/aikatsu_calendar/scraper.rb', line 45

def feed_row(tr)
  # 日付
  text = tr.at_css('td').text
  m = text.match(/(\d+)日/) or raise ValueError, text
  @day = m[1].to_i

  tr.css('p').each do |p|
    feed_item(p)
  end
end

#feed_table(table) ⇒ Object 



33
34
35
36
37
38
39
40
41
42
43
 
# File 'lib/aikatsu_calendar/scraper.rb', line 33

def feed_table(table)
  # 年と月
  text = table.at_css('th').text
  m = text.match(/(\d+)年(\d+)月/) or raise ValueError, text
  @year = m[1].to_i
  @month = m[2].to_i

  table.css('tr')[1..-1].each do |tr|
    feed_row(tr)
  end
end

#parse_item(p) ⇒ Object 



60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
 
# File 'lib/aikatsu_calendar/scraper.rb', line 60

def parse_item(p)
  # 日付
  text = p.text
  re = /(?:　※)?(\d+)年(\d+)月(\d+)日～(?:(\d+)年)?(\d+)月(\d+)日/
  m = text.match(re)
  if m
    year_until = (m[4] || m[1]).to_i
    date_from = Time.local(m[1].to_i, m[2].to_i, m[3].to_i)
    date_until = Time.local(year_until, m[5].to_i, m[6].to_i)
  else
    date_from = date_until = Time.local(@year, @month, @day)
  end
  # 日付をとっぱらう
  text = text.sub(re, '')

  # URL
  url = nil
  if (a = p.at_css('a[href]'))
    base = 'http://www.aikatsu.com/calender/'
    url = URI.join(base, a.attr(:href)).to_s
  end

  # type
  type = class_to_type(p.attr(:class))

  {
    type: type,
    date_from: date_from,
    date_until: date_until,
    content: text.strip,
    link: url,
  }
end

#to_json(pretty = false) ⇒ Object 



102
103
104
105
106
107
108
 
# File 'lib/aikatsu_calendar/scraper.rb', line 102

def to_json(pretty=false)
  if pretty
    JSON.pretty_generate(@schedules)
  else
    JSON.dump(@schedules)
  end
end