Class: ScrapKit::Recipe

Inherits:
Object
  • Object
show all
Defined in:
lib/scrap_kit/recipe.rb

Instance Attribute Summary collapse

Class Method Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize(url: nil, steps: [], attributes: {}) ⇒ Recipe

Returns a new instance of Recipe.


23
24
25
26
27
 
# File 'lib/scrap_kit/recipe.rb', line 23

def initialize(url: nil, steps: [], attributes: {})
  @url = url
  @steps = steps
  @attributes = attributes
end

Instance Attribute Details

#browser ⇒ Object

Returns the value of attribute browser.


7
8
9
 
# File 'lib/scrap_kit/recipe.rb', line 7

def browser
  @browser
end

#user_agent ⇒ Object

Returns the value of attribute user_agent.


7
8
9
 
# File 'lib/scrap_kit/recipe.rb', line 7

def user_agent
  @user_agent
end

Class Method Details

.load(source) ⇒ Object 


10
11
12
13
14
15
16
17
18
19
20
 
# File 'lib/scrap_kit/recipe.rb', line 10

def load(source)
  input = if source.is_a?(Hash)
    source
  elsif source.is_a?(IO)
    JSON.parse(source.read)
  else
    JSON.parse(File.read(source))
  end

  new(input.deep_symbolize_keys)
end

Instance Method Details

#elements_from_selector(browser_or_element, selector) ⇒ Object 


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
 
# File 'lib/scrap_kit/recipe.rb', line 67

def elements_from_selector(browser_or_element, selector)
  if selector.is_a?(String)
    browser_or_element.elements(css: selector)
  elsif selector.is_a?(Hash)
    browser_or_element.elements(selector)
  elsif selector.is_a?(Array)
    *remainder, condition = selector
    condition_key, condition_value = condition.first
    elements = browser_or_element

    if remainder.empty?
      elements = elements.elements(css: condition_key.to_s)
    else
      remainder.each do |item|
        elements = elements.elements(css: item)
      end
    end

    elements.filter do |element|
      found_element = element.element(css: condition_key.to_s)
      extracted_value = extract_value_from_element(found_element)
      extracted_value.match(condition_value) || extracted_value == condition_value
    end
  end
end

#extract_attribute(browser_or_element, selector_or_object) ⇒ Object 


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
 
# File 'lib/scrap_kit/recipe.rb', line 107

def extract_attribute(browser_or_element, selector_or_object)
  if selector_or_object.is_a?(String)
    extract_value_from_element(browser_or_element.element(css: selector_or_object))
  elsif selector_or_object.is_a?(Array)
    found_elements = elements_from_selector(browser_or_element, selector_or_object)

    if found_elements.size === 1
      extract_value_from_element(found_elements.first)
    else
      found_elements.map do |element|
        extract_value_from_element(element)
      end
    end
  elsif selector_or_object.is_a?(Hash)
    if selector_or_object[:selector] && selector_or_object[:children_attributes]
      selector = selector_or_object[:selector]
      selector_for_children_attributes = selector_or_object[:children_attributes]

      elements_from_selector(browser_or_element, selector).map do |element|
        output = {}

        selector_for_children_attributes.each do |child_attribute_name, child_selector|
          output[child_attribute_name] = extract_attribute(element, child_selector)
        end

        output
      end
    elsif selector_or_object[:javascript]
      @browser.execute_script(selector_or_object[:javascript])
    else
      found_elements = elements_from_selector(browser_or_element, selector_or_object)

      if found_elements.size === 1
        extract_value_from_element(found_elements.first)
      else
        found_elements.map do |element|
          extract_value_from_element(element)
        end
      end
    end
  end
rescue
  nil
end

#extract_value_from_element(element) ⇒ Object 


93
94
95
96
97
98
99
100
101
102
103
104
105
 
# File 'lib/scrap_kit/recipe.rb', line 93

def extract_value_from_element(element)
  return nil unless element.exists?

  if element&.respond_to?(:tag_name)
    if element.tag_name.downcase == "input"
      return element.attribute_value(:value)
    elsif element.tag_name.downcase == "img"
      return element.attribute_value(:src)
    end
  end

  element&.text_content
end

#find_element_by_name_or_selector(browser_or_element, name_or_selector) ⇒ Object 


57
58
59
60
61
62
63
64
65
 
# File 'lib/scrap_kit/recipe.rb', line 57

def find_element_by_name_or_selector(browser_or_element, name_or_selector)
  element = browser_or_element.element(name: name_or_selector.to_s)
  return element if element.exists?

  element = browser_or_element.element(css: name_or_selector.to_s)
  return element if element.exists?

  nil
end

#run ⇒ Object 


29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
 
# File 'lib/scrap_kit/recipe.rb', line 29

def run
  output = {}

  @browser = create_browser
  @browser.goto @url

  @steps.each do |step|
    run_step(step)
  end

  @attributes.each do |attribute_name, selector|
    output[attribute_name] = extract_attribute(@browser, selector)
  end

  @browser.close
  @browser = nil

  output
end

#run_step(step) ⇒ Object 


49
50
51
52
53
54
55
 
# File 'lib/scrap_kit/recipe.rb', line 49

def run_step(step)
  return goto(step[:goto]) if step[:goto]
  return click(step[:click]) if step[:click]
  return fill_form(step[:fill_form]) if step[:fill_form]

  nil
end