Class: TSV::Parser

Inherits:
Object
  • Object
show all
Defined in:
lib/scout/tsv/parser.rb

Instance Attribute Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize(file, fix: true, header_hash: "#", sep: "\t", type: :double) ⇒ Parser

Returns a new instance of Parser.



311
312
313
314
315
316
317
318
319
320
321
322
323
 
# File 'lib/scout/tsv/parser.rb', line 311

def initialize(file, fix: true, header_hash: "#", sep: "\t", type: :double)
  if IO === file
    @stream = file
  else
    @stream = Open.open(file)
  end
  @fix = fix
  @source_options, @key_field, @fields, @first_line, @preamble = TSV.parse_header(@stream, fix:fix, header_hash:header_hash, sep:sep)
  @source_options[:filename] = file if Path.is_filename?(file)
  @source_options[:sep] = sep if @source_options[:sep].nil?
  @source_options.merge!(:key_field => @key_field, :fields => @fields)
  @type = @source_options[:type] || type
end

Instance Attribute Details

#fields ⇒ Object

Returns the value of attribute fields.



310
311
312
 
# File 'lib/scout/tsv/parser.rb', line 310

def fields
  @fields
end

#first_line ⇒ Object

Returns the value of attribute first_line.



310
311
312
 
# File 'lib/scout/tsv/parser.rb', line 310

def first_line
  @first_line
end

#key_field ⇒ Object

Returns the value of attribute key_field.



310
311
312
 
# File 'lib/scout/tsv/parser.rb', line 310

def key_field
  @key_field
end

#preamble ⇒ Object

Returns the value of attribute preamble.



310
311
312
 
# File 'lib/scout/tsv/parser.rb', line 310

def preamble
  @preamble
end

#source_options ⇒ Object

Returns the value of attribute source_options.



310
311
312
 
# File 'lib/scout/tsv/parser.rb', line 310

def source_options
  @source_options
end

#stream ⇒ Object

Returns the value of attribute stream.



310
311
312
 
# File 'lib/scout/tsv/parser.rb', line 310

def stream
  @stream
end

#type ⇒ Object

Returns the value of attribute type.



310
311
312
 
# File 'lib/scout/tsv/parser.rb', line 310

def type
  @type
end

Instance Method Details

#all_fields ⇒ Object 



329
330
331
332
 
# File 'lib/scout/tsv/parser.rb', line 329

def all_fields
  return nil if @fields.nil?
  [@key_field] + @fields
end

#digest_str ⇒ Object 



421
422
423
 
# File 'lib/scout/tsv/parser.rb', line 421

def digest_str
  fingerprint
end

#fingerprint ⇒ Object 



417
418
419
 
# File 'lib/scout/tsv/parser.rb', line 417

def fingerprint
  "Parser:{"<< Log.fingerprint(self.all_fields|| []) << "}"
end

#identify_field(name) ⇒ Object 



342
343
344
 
# File 'lib/scout/tsv/parser.rb', line 342

def identify_field(name)
  TSV.identify_field(@key_field, @fields, name)
end

#inspect ⇒ Object 



425
426
427
 
# File 'lib/scout/tsv/parser.rb', line 425

def inspect
  fingerprint
end

#options ⇒ Object 



325
326
327
 
# File 'lib/scout/tsv/parser.rb', line 325

def options
  IndiferentHash.add_defaults @source_options.dup, type: type, key_field: key_field, fields: fields
end

#traverse(key_field: nil, fields: nil, filename: nil, namespace: nil, **kwargs, &block) ⇒ Object 



346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
 
# File 'lib/scout/tsv/parser.rb', line 346

def traverse(key_field: nil, fields: nil, filename: nil, namespace: nil,  **kwargs, &block)
  kwargs[:type] ||=  self.source_options[:type] ||= @type
  kwargs[:type] = kwargs[:type].to_sym if kwargs[:type]

  if fields
    if @fields
      all_field_names ||= [@key_field] + @fields
      fields = all_field_names if fields == :all
      positions = NamedArray.identify_name(all_field_names, fields)

      raise "Not all fields (#{Log.fingerprint fields}) identified in #{Log.fingerprint all_field_names}" if positions.include?(nil)
      kwargs[:positions] = positions
      field_names = all_field_names.values_at *positions
    elsif fields.reject{|f| Numeric === f}.empty?
      positions = fields
      kwargs[:positions] = positions
    else
      raise "Non-numeric fields specified, but no field names available"
    end
  else
    field_names = @fields
  end

  kwargs[:positions] = nil if @type == :flat

  if key_field
    if @fields
      all_field_names ||= [@key_field] + @fields
      key = NamedArray.identify_name(all_field_names, key_field)
      kwargs[:key] = key == :key ? 0 : key
      key_field_name = (key.nil? || key == :key) ? @key_field : all_field_names[key]
      if fields.nil?
        field_names = all_field_names - [key_field_name]
      end
    else
      kwargs[:key] = key_field == :key ? 0 : key_field
      key = key_field
    end
  else
    key_field_name = @key_field
  end

  if field_names && (kwargs[:type] == :single || kwargs[:type] == :flat)
    field_names = field_names.slice(0,1)
  end

  @source_options.each do |option,value|
    option = option.to_sym
    next unless KEY_PARAMETERS.include? option
    kwargs[option] = value unless kwargs.include?(option)
  end

  kwargs[:source_type] = @source_options[:type]
  kwargs[:data] = false if kwargs[:data].nil?

  if kwargs[:tsv_grep]
    data = with_stream do |stream|
      grep_stream = Open.grep(stream, kwargs.delete(:tsv_grep), kwargs.delete(:tsv_invert_grep))
      TSV.parse_stream(grep_stream, first_line: nil, fix: @fix, field_names: @fields, **kwargs, &block)
    end
  else
    data = TSV.parse_stream(@stream, first_line: @first_line, fix: @fix, field_names: @fields, **kwargs, &block)
  end

  if data
    TSV.setup(data, :key_field => key_field_name, :fields => field_names, :type => @type)
  else
    [key_field || self.key_field, fields || self.fields]
  end
end

#with_stream {|sout| ... } ⇒ Object

Yields:

  • (sout) 


429
430
431
432
433
434
435
 
# File 'lib/scout/tsv/parser.rb', line 429

def with_stream
  sout = Open.open_pipe do |sin|
    sin.puts @first_line
    Open.consume_stream(@stream, false, sin)
  end
  yield sout
end