Class: Crabfarm::BaseState

Inherits:
Object
  • Object
show all
Extended by:
Forwardable
Includes:
Assertion::Context
Defined in:
lib/crabfarm/base_state.rb

Constant Summary collapse

PARSE_METHOD_RX = 
/^parse_(.*)$/

Instance Attribute Summary collapse

Class Method Summary collapse

Instance Method Summary collapse

Methods included from Assertion::Context

#assert

Constructor Details

#initialize(_context, _params) ⇒ BaseState

Returns a new instance of BaseState.



26
27
28
29
30
31
32
 
# File 'lib/crabfarm/base_state.rb', line 26

def initialize(_context, _params)
  @context = _context
  @params = _params

  @dsl = Strategies.load(:browser_dsl, class_browser_dsl || Crabfarm.config.browser_dsl)
  @builder = Strategies.load(:output_builder, class_output_builder || Crabfarm.config.output_builder)
end

Dynamic Method Handling

This class handles dynamic methods through the method_missing method

#method_missing(_method, *_args, &_block) ⇒ Object 



76
77
78
79
80
81
82
83
 
# File 'lib/crabfarm/base_state.rb', line 76

def method_missing(_method, *_args, &_block)
  m = PARSE_METHOD_RX.match(_method)
  if m
    options = _args[1] || {}
    options[:using] = (m[1].camelize + 'Parser').constantize
    parse _args[0], options
  else super end
end

Instance Attribute Details

#output ⇒ Object (readonly)

Returns the value of attribute output.



12
13
14
 
# File 'lib/crabfarm/base_state.rb', line 12

def output
  @output
end

#params ⇒ Object (readonly)

Returns the value of attribute params.



12
13
14
 
# File 'lib/crabfarm/base_state.rb', line 12

def params
  @params
end

Class Method Details

.browser_dsl(_dsl) ⇒ Object 



18
19
20
 
# File 'lib/crabfarm/base_state.rb', line 18

def self.browser_dsl(_dsl)
  @class_browser_dsl = _dsl
end

.output_builder(_builder) ⇒ Object 



22
23
24
 
# File 'lib/crabfarm/base_state.rb', line 22

def self.output_builder(_builder)
  @class_output_builder = _builder
end

Instance Method Details

#browser(_name = nil) ⇒ Object 



34
35
36
 
# File 'lib/crabfarm/base_state.rb', line 34

def browser(_name=nil)
  @dsl.wrap driver(_name)
end

#crawl ⇒ Object

Raises:

  • (NotImplementedError) 


50
51
52
 
# File 'lib/crabfarm/base_state.rb', line 50

def crawl
  raise NotImplementedError.new
end

#download(_url) ⇒ Object 



38
39
40
 
# File 'lib/crabfarm/base_state.rb', line 38

def download(_url)
  @context.http.get(_url).body
end

#fork_each(_enumerator, &_block) ⇒ Object 



66
67
68
69
70
71
72
73
74
 
# File 'lib/crabfarm/base_state.rb', line 66

def fork_each(_enumerator, &_block)
  session_id = 0
  mutex = Mutex.new
  ths = _enumerator.map do |value|
    session_id += 1
    start_forked_state("th_session_#{session_id}", value, _block, mutex)
  end
  ThreadsWait.all_waits(*ths)
end

#output_as_json ⇒ Object 



46
47
48
 
# File 'lib/crabfarm/base_state.rb', line 46

def output_as_json
  @builder.serialize @output
end

#parse(_target = nil, _options = {}) ⇒ Object 



54
55
56
57
58
59
60
61
62
63
64
 
# File 'lib/crabfarm/base_state.rb', line 54

def parse(_target=nil, _options={})
  parser_class = _options.delete :using

  if parser_class.nil?
    parser_class = (self.class.name + 'Parser').constantize
  end

  parser = parser_class.new _target, @params.merge(_options)
  parser.parse
  return parser
end

#respond_to?(_method, _include_all = false) ⇒ Boolean

Returns:

  • (Boolean) 


85
86
87
88
 
# File 'lib/crabfarm/base_state.rb', line 85

def respond_to?(_method, _include_all=false)
  return true if PARSE_METHOD_RX === _method
  super
end