Class: FideXmlParser::Processor

Inherits:
Nokogiri::XML::SAX::Document
  • Object
show all
Defined in:
lib/fide_xml_parser/processor.rb

Direct Known Subclasses

FideProcessor

Constant Summary collapse

ANSI_GO_TO_LINE_START = 
"\033[1G"

Instance Attribute Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize(array_name, record_name, numeric_fields) ⇒ Processor

Returns a new instance of Processor.



45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
 
# File 'lib/fide_xml_parser/processor.rb', line 45

def initialize(array_name, record_name, numeric_fields)
  @array_name = array_name
  @record_name = record_name
  @numeric_fields = numeric_fields
  @key_filter = nil
  @record_filter = nil
  @field_name_renames = nil
  @current_property_name = nil
  @record = {}
  @records = []
  @start_time = current_time
  @keys_to_exclude = []
  @input_record_count = 0
  @output_record_count = 0
end

Instance Attribute Details

#array_name ⇒ Object

Constructor parameters:



35
36
37
 
# File 'lib/fide_xml_parser/processor.rb', line 35

def array_name
  @array_name
end

#current_property_name ⇒ Object

For internal use:



41
42
43
 
# File 'lib/fide_xml_parser/processor.rb', line 41

def current_property_name
  @current_property_name
end

#field_name_renames ⇒ Object

User-provided callbacks:



38
39
40
 
# File 'lib/fide_xml_parser/processor.rb', line 38

def field_name_renames
  @field_name_renames
end

#input_record_count ⇒ Object

For internal use:



41
42
43
 
# File 'lib/fide_xml_parser/processor.rb', line 41

def input_record_count
  @input_record_count
end

#key_filter ⇒ Object

User-provided callbacks:



38
39
40
 
# File 'lib/fide_xml_parser/processor.rb', line 38

def key_filter
  @key_filter
end

#numeric_fields ⇒ Object

Constructor parameters:



35
36
37
 
# File 'lib/fide_xml_parser/processor.rb', line 35

def numeric_fields
  @numeric_fields
end

#output_record_count ⇒ Object

For internal use:



41
42
43
 
# File 'lib/fide_xml_parser/processor.rb', line 41

def output_record_count
  @output_record_count
end

#record ⇒ Object

For internal use:



41
42
43
 
# File 'lib/fide_xml_parser/processor.rb', line 41

def record
  @record
end

#record_filter ⇒ Object

User-provided callbacks:



38
39
40
 
# File 'lib/fide_xml_parser/processor.rb', line 38

def record_filter
  @record_filter
end

#record_name ⇒ Object

Constructor parameters:



35
36
37
 
# File 'lib/fide_xml_parser/processor.rb', line 35

def record_name
  @record_name
end

#records ⇒ Object

For internal use:



41
42
43
 
# File 'lib/fide_xml_parser/processor.rb', line 41

def records
  @records
end

#start_time ⇒ Object (readonly)

Returns the value of attribute start_time.



32
33
34
 
# File 'lib/fide_xml_parser/processor.rb', line 32

def start_time
  @start_time
end

Instance Method Details

#characters(string) ⇒ Object 



113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
 
# File 'lib/fide_xml_parser/processor.rb', line 113

def characters(string)
  if current_property_name
    if key_filter.nil? || key_filter.(current_property_name)
      value = numeric_fields.include?(current_property_name) ? Integer(string) : string
      key = current_property_name
      if field_name_renames
        new_field_name = field_name_renames[key]
        if new_field_name
          key = new_field_name
        end
      end
      record[key] = value
    end
  end
end

#current_time ⇒ Object 



69
70
71
 
# File 'lib/fide_xml_parser/processor.rb', line 69

def current_time
  Process.clock_gettime(Process::CLOCK_MONOTONIC)
end

#end_element(name) ⇒ Object 



97
98
99
100
101
102
103
104
105
106
107
108
109
110
 
# File 'lib/fide_xml_parser/processor.rb', line 97

def end_element(name)
  case name
  when array_name  # end of data, write JSON file
    finish
  when record_name
    if record_filter.nil? || record_filter.(record)
      self.output_record_count += 1
      records << record
    end
    self.record = {}
  else
    self.current_property_name = nil
  end
end

#finish ⇒ Object 



130
131
132
133
 
# File 'lib/fide_xml_parser/processor.rb', line 130

def finish
  output_status
  puts
end

#output_status ⇒ Object 



74
75
76
77
78
79
80
81
 
# File 'lib/fide_xml_parser/processor.rb', line 74

def output_status
  print ANSI_GO_TO_LINE_START
  print "Records processed: %9d   kept: %9d    Seconds elapsed: %11.2f" % [
      input_record_count,
      output_record_count,
      current_time - start_time
  ]
end

#parse(data_source) ⇒ Object 



62
63
64
65
66
 
# File 'lib/fide_xml_parser/processor.rb', line 62

def parse(data_source)
  parser = Nokogiri::XML::SAX::Parser.new(self)
  parser.parse(data_source)
  records
end

#start_element(name, _attrs) ⇒ Object 



84
85
86
87
88
89
90
91
92
93
94
 
# File 'lib/fide_xml_parser/processor.rb', line 84

def start_element(name, _attrs)
  case name
  when array_name
    # ignore
  when record_name
    self.input_record_count += 1
    output_status if input_record_count % 1000 == 0
  else # this is a field in the players record; process it as such
    self.current_property_name = name
  end
end