Class: Fairy::PSort::PPreSort

Inherits:
Fairy::PBasicGroupBy show all
Defined in:
lib/fairy/node/p-sort.rb

Constant Summary collapse

ST_ALL_IMPORTED = 
:ST_ALL_IMPORTED
ST_WAIT_EXPORT_FINISH = 
:ST_WAIT_EXPORT_FINISH
ST_EXPORT_FINISH = 
:ST_EXPORT_FINISH

Constants inherited from Fairy::PIOFilter

Fairy::PIOFilter::ST_WAIT_IMPORT

Constants inherited from Fairy::PFilter

Fairy::PFilter::END_OF_STREAM, Fairy::PFilter::ST_ACTIVATE, Fairy::PFilter::ST_FINISH, Fairy::PFilter::ST_INIT

Instance Attribute Summary

Attributes inherited from Fairy::PFilter

#IGNORE_EXCEPTION, #id, #log_id, #ntask

Class Method Summary collapse

Instance Method Summary collapse

Methods inherited from Fairy::PBasicGroupBy

#init_key_proc, #terminate, #wait_export_finish

Methods inherited from Fairy::PIOFilter

#input=

Methods inherited from Fairy::PFilter

#abort_running, #basic_start, #break_running, #each, #global_break, #global_break_from_other, #handle_exception, #key, #key=, #next, #no, #no=, #notice_status, #processor, #start, #start_watch_status, #status=, #terminate, #terminate_proc

Constructor Details

#initialize(id, ntask, bjob, opts, block_source) ⇒ PPreSort

Returns a new instance of PPreSort.



20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
 
# File 'lib/fairy/node/p-sort.rb', line 20

def initialize(id, ntask, bjob, opts, block_source)
  super
  @block_source = block_source

  @exports = []
  def @exports.each_pair(&block)
    each_with_index do |item, idx| 
      block.call(idx, item) if item
    end
  end
  @exports_queue = XThread::Queue.new
  
  @counter = []

  @pvs = nil
  if @opts[:pvs]
    @pvs = @opts[:pvs].dc_deep_copy
  end

  #start_watch_exports
end

Class Method Details

.each_pair(&block) ⇒ Object 



25
26
27
28
29
 
# File 'lib/fairy/node/p-sort.rb', line 25

def @exports.each_pair(&block)
  each_with_index do |item, idx| 
    block.call(idx, item) if item
  end
end

Instance Method Details

#add_export(key, export) ⇒ Object 



42
43
44
45
46
47
 
# File 'lib/fairy/node/p-sort.rb', line 42

def add_export(key, export)
  @exports[key] = export
  #      @exports_queue.push [key, export]
  # [BUG#171]同期処理でないとまずい.
  @bjob.add_exports(key, export, self)
end

#hash_key(e) ⇒ Object 



148
149
150
 
# File 'lib/fairy/node/p-sort.rb', line 148

def hash_key(e)
  @key_proc.yield(e)
end

#hashing(e) ⇒ Object 



130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
 
# File 'lib/fairy/node/p-sort.rb', line 130

def hashing(e)
  if Import::CTLTOKEN_NULLVALUE === (key = @key_proc.call(e))
    return
  end
  
  unless idx = @pvs.find_index{|pv| key < pv}
    idx = @pvs.size
  end

#Log::debug(self, "#{@pvs.inspect}")
#Log::debug(self, "#{idx}")
  

  export = @exports[idx]
  export.push e
  @counter[idx] += 1
end

#init_exports ⇒ Object 



116
117
118
119
120
121
122
123
124
125
126
127
128
 
# File 'lib/fairy/node/p-sort.rb', line 116

def init_exports
  policy = @opts[:postqueuing_policy]
  (@pvs.size+1).times do |idx|
    export = Export.new(policy)
    @exports[idx] = export
    
    export.njob_id = @id
    export.no = idx
    export.add_key(idx)
    add_export(idx, export)
    @counter[idx] = 0
  end
end

#start_export ⇒ Object 



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
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
 
# File 'lib/fairy/node/p-sort.rb', line 49

def start_export
  Log::debug(self, "START_EXPORT")

  start do
    sample_line_no = @opts[:sampling_max]
    sample_line_no ||= CONF.SORT_SAMPLING_MAX

    hash_opt = @opts[:cmp_optimize]
    hash_opt = CONF.SORT_CMP_OPTIMIZE if hash_opt.nil?
    
    if hash_opt
      @key_proc = eval("proc{#{@block_source.source}}", @context.binding)
    else
      @key_proc = BBlock.new(@block_source, @context, self)
    end
    
    buf = []
    no = 0
    begin
      if @pvs
 sampling = false
Log::debugf(self, "%s", @pvs.inspect)
 init_exports
      elsif self.no == 0
 sampling = true
      else
 sampling = false
 @pvs = @bjob.get_pvs
Log::debugf(self, "%s", @pvs.inspect)
 init_exports
      end
 
      @input.each do |e|
 if sampling
    no += 1
    buf.push e
    if no >= sample_line_no
      sampling = false
      @pvs = @bjob.get_pvs(buf)
Log::debugf(self, "%s", @pvs.inspect)
      init_exports
      buf.each{|e| hashing(e)}
    end
 else
    hashing(e)
 end
      end
      if sampling
 @pvs = @bjob.get_pvs(buf)
Log::debugf(self, "%s", @pvs.inspect)
 init_exports
 buf.each{|e| hashing(e)}
      end
    rescue
      Log::debug_exception(self)
      raise
    ensure
      @exports_queue.push nil
      @exports.each_pair do |key, export| 
 next unless export
 Log::debug(self, "G0 #{key} => #{@counter[key]}")      
 export.push END_OF_STREAM
      end
    end
  end
end