Class: SpiderBot::CLI

Inherits:
Thor
  • Object
show all
Defined in:
lib/spider_bot/cli.rb

Instance Method Summary collapse

Instance Method Details

#crawl ⇒ Object 



46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
 
# File 'lib/spider_bot/cli.rb', line 46

def crawl
  $expire_num = options[:expire].to_i if options[:expire]
  require File.join(File.expand_path('../..',__FILE__), "spider_bot/load")

  if options[:bot]
    bot_file = File.expand_path(options[:bot]) 
    return raise "Bot file not found" if !File.exists?(bot_file)
    load bot_file 
  end

  if options[:dir]
    bot_dir = File.expand_path(options[:dir]) 
    return raise "Dir is not found" if !Dir.exists?(bot_dir)
    
    threads = []
    Dir.glob("#{bot_dir}/*_bot.rb").each do |file|
      threads << Thread.new do
        begin
          SpiderBot.logger.info "loading bot file with #{file}."
          load file
        rescue Exception => e
          SpiderBot.logger.error "has errors with loading bot file #{ file }"
          SpiderBot.logger.error e.to_s
        end
      end
    end
    threads.each { |t| t.join }
  end
end

#start ⇒ Object 



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
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
 
# File 'lib/spider_bot/cli.rb', line 99

def start
  puts "start....."
  
  $expire_num = options[:expire].to_i if options[:expire]
  
  if options[:env]
    ENV['RACK_ENV'] = options[:env]
  else
    ENV['RACK_ENV']= 'development'
  end
  
  require File.join(File.expand_path('../..',__FILE__), "spider_bot/load")

  FileUtils.mkdir_p("tmp/pids") if !File.exists?("tmp/pids")
  
  daemon_options = {
    app_name: 'spider',
    ontop: true,
    dir: 'tmp/pids',
  }

  sleep_time = 10
  
  if options[:daemon]
    daemon_options[:ontop] = false 
  else
    puts "press ctrl-c exit"
  end

  stop if File.exists?("tmp/spider.pid")

  if option_time = options[:time]
    parse_time = option_time.match(/[d|h|m]/)
    sleep_time = if parse_time
      case parse_time[0]
      when "d"
        option_time.to_i * 60 * 60 * 24
      when "h"
        option_time.to_i * 60 * 60
      when "m"
        option_time.to_i * 60
      end
    else
      option_time.to_i
    end
  end

  Daemons.daemonize(daemon_options)
  
  loop do
    threads = []
    
    BOTDIR.each do |file|
      threads << Thread.new do
        begin
          SpiderBot.logger.info "loading bot file with #{file}."
          load file
        rescue Exception => e
          SpiderBot.logger.error "has errors with loading bot file #{ file }"
          SpiderBot.logger.error e.to_s
        end
        sleep(10)
      end
    end

    threads.each { |t| t.join }
    
    if options[:random]
      random_time = Random.new.rand(sleep_time)
      sleep(random_time.to_i)
    else
      sleep(sleep_time.to_i)
    end
  end
end

#stop ⇒ Object 



177
178
179
180
181
 
# File 'lib/spider_bot/cli.rb', line 177

def stop
  pid = File.read("tmp/pids/spider.pid").to_i
  Process.kill(9, pid)
  File.delete("tmp/pids/spider.pid")
end

#url(arg) ⇒ Object 



26
27
28
29
30
 
# File 'lib/spider_bot/cli.rb', line 26

def url(arg)
  data = Crawl.new(arg, options).crawl_data
  return File.open(options[:out], "w"){ file.puts data } if options[:out]
  return puts data 
end