96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
|
# File 'lib/bibletools.rb', line 96
def initialize(book: 'Exodus', url: '', debug: false)
raise ScraperErr, 'Please specify the URL' if url.empty?
baseurl = url
id = if book.is_a? String then
i = BOOKS.index(book.gsub(/\b./){|x| x.upcase})
puts 'book *' + book + '* not found' unless i
(i + 1).to_s
else
book.to_s
end
puts 'id: ' + id.inspect if @debug
url2 = baseurl + '?id=' + id
doc = Nokorexi.new(url2).to_doc
e = doc.root.at_css('.pagination')
endchapter = e.xpath('li')[-2].text('a')
book = Rexle.new('<book/>')
(1..endchapter.to_i).each do |chp|
c = chp.to_s
puts 'reading chapter ' + c
sleep 1
url2 = baseurl + ("?id=%s&bible_chapter=%s" % [id, c])
doc = Nokorexi.new(url2).to_doc
a = doc.root.xpath('//div/p')
a2 = a.select {|e| e2 = e.element('a'); e2.attributes[:name] if e2}
chapter = Rexle::Element.new('chapter', attributes: {no: c})
a2.each.with_index do |e, i|
puts 'processing verse ' + i.to_s
txt = e.plaintext
n = txt.slice!(0,2).to_i.to_s
chapter.add Rexle::Element.new('verse', attributes: {no: n}).add_text(txt)
end
book.root.add chapter
end
@doc = book
end
|