Class: TreasureData::FileReader

Inherits:
Object
  • Object
show all
Defined in:
lib/td/file_reader.rb

Overview

json and msgpack format supports array types with columns

  • when –column-header option

“a”, “b”, “c”

# first line is header

“v”, 10, true

# array types, e.g. generate => “v”, “b” => 10, “c” => true

…

  • when –columns a,b,c

“v”, 10, true

# array types

…

Defined Under Namespace

Modules: RegexpParserMixin Classes: ApacheParser, AutoTypeConvertParserFilter, DecompressIOFilter, DelimiterParser, HashBuilder, JSONParser, LineReader, MessagePackParsingReader, SeparatedValueParsingReader, SetTimeParserFilter, SyslogParser, TimeParserFilter

Instance Attribute Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize ⇒ FileReader

Returns a new instance of FileReader.



302
303
304
305
306
307
308
309
310
311
312
313
314
 
# File 'lib/td/file_reader.rb', line 302

def initialize
  @format = "text"
  @default_opts = {
    :delimiter_expr => /\t|,/,
    #:line_delimiter_expr => /\r?\n/,
    :null_expr => /\A(?:null||\-|\\N)\z/i,
    :true_expr => /\A(?:true)\z/i,
    :false_expr => /\A(?:false)\z/i,
    #:quote_char => "\"",
  }
  @opts = {}
  @parser_class = nil
end

Instance Attribute Details

#default_opts ⇒ Object (readonly)

Returns the value of attribute default_opts.



316
317
318
 
# File 'lib/td/file_reader.rb', line 316

def default_opts
  @default_opts
end

#opts ⇒ Object (readonly)

Returns the value of attribute opts.



316
317
318
 
# File 'lib/td/file_reader.rb', line 316

def opts
  @opts
end

#parser_class ⇒ Object

Returns the value of attribute parser_class.



317
318
319
 
# File 'lib/td/file_reader.rb', line 317

def parser_class
  @parser_class
end

Instance Method Details

#compose_factory ⇒ Object 



404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
 
# File 'lib/td/file_reader.rb', line 404

def compose_factory
  opts = @default_opts.merge(@opts)
  case @format
  when 'text'
    Proc.new {|io,error|
      io = DecompressIOFilter.filter(io, error, opts)
      reader = LineReader.new(io, error, opts)
      parser = DelimiterParser.new(reader, error, opts)
      if opts[:column_header]
        column_names = parser.forward
      elsif opts[:column_names]
        column_names = opts[:column_names]
      else
        raise "--column-header or --columns option is required"
      end
      unless opts[:all_string]
        parser = AutoTypeConvertParserFilter.new(parser, error, opts)
      end
      parser = HashBuilder.new(parser, error, column_names)
      if opts[:time_value]
        parser = SetTimeParserFilter.new(parser, error, opts)
      else
        parser = TimeParserFilter.new(parser, error, opts)
      end
    }

  when 'apache', 'syslog'
    Proc.new {|io,error|
      io = DecompressIOFilter.filter(io, error, opts)
      reader = LineReader.new(io, error, opts)
      parser = if @format == 'apache'
                 ApacheParser.new(reader, error, opts)
               else
                 SyslogParser.new(reader, error, opts)
               end
      if opts[:column_names]
        column_names = opts[:column_names]
      else
        raise "--columns option is required"
      end
      unless opts[:all_string]
        parser = AutoTypeConvertParserFilter.new(parser, error, opts)
      end
      parser = HashBuilder.new(parser, error, column_names)
      if opts[:time_value]
        parser = SetTimeParserFilter.new(parser, error, opts)
      else
        parser = TimeParserFilter.new(parser, error, opts)
      end
    }

  when 'json'
    Proc.new {|io,error|
      io = DecompressIOFilter.filter(io, error, opts)
      reader = LineReader.new(io, error, opts)
      parser = JSONParser.new(reader, error, opts)
      if opts[:column_header]
        column_names = parser.forward
      elsif opts[:column_names]
        column_names = opts[:column_names]
      end
      if column_names
        parser = HashBuilder.new(parser, error, column_names)
      end
      if opts[:time_value]
        parser = SetTimeParserFilter.new(parser, error, opts)
      else
        parser = TimeParserFilter.new(parser, error, opts)
      end
    }

  when 'msgpack'
    Proc.new {|io,error|
      io = DecompressIOFilter.filter(io, error, opts)
      parser = MessagePackParsingReader.new(io, error, opts)
      if opts[:column_header]
        column_names = parser.forward
      elsif opts[:column_names]
        column_names = opts[:column_names]
      end
      if column_names
        parser = HashBuilder.new(parser, error, column_names)
      end
      if opts[:time_value]
        parser = SetTimeParserFilter.new(parser, error, opts)
      else
        parser = TimeParserFilter.new(parser, error, opts)
      end
    }
  end
end

#init_optparse(op) ⇒ Object 



319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
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
 
# File 'lib/td/file_reader.rb', line 319

def init_optparse(op)
  op.on('-f', '--format NAME', "source file format [csv, tsv, msgpack, json]") {|s|
    set_format_template(s)
  }
  op.on('-h', '--columns NAME,NAME,...', 'column names (use --column-header instead if the first line has column names)') {|s|
    @opts[:column_names] = s.split(',')
  }
  op.on('-H', '--column-header', 'first line includes column names', TrueClass) {|b|
    @opts[:column_header] = b
  }
  op.on('-d', '--delimiter REGEX', "delimiter between columns (default: #{@default_opts[:delimiter_expr].to_s})") {|s|
    @opts[:delimiter_expr] = Regexp.new(s)
  }
  #op.on('-D', '--line-delimiter REGEX', "delimiter between rows (default: #{@default_opts[:line_delimiter_expr].inspect[1..-2]})") {|s|
  #  @opts[:line_delimiter_expr] = Regexp.new(s)
  #}
  op.on('--null REGEX', "null expression for the automatic type conversion (default: #{@default_opts[:null_expr].to_s})") {|s|
    @opts[:null_expr] = Regexp.new(s)
  }
  op.on('--true REGEX', "true expression for the automatic type conversion (default: #{@default_opts[:true_expr].to_s})") {|s|
    @opts[:true_expr] = Regexp.new(s)
  }
  op.on('--false REGEX', "false expression for the automatic type conversion (default: #{@default_opts[:false_expr].to_s})") {|s|
    @opts[:false_expr] = Regexp.new(s)
  }
  # TODO
  #op.on('-E', '--escape CHAR', "escape character (default: no escape character)") {|s|
  #  @opts[:escape_char] = s
  #}
  #op.on('-Q', '--quote CHAR', "quote character (default: #{@default_opts[:quote_char]}") {|s|
  #  @opts[:quote_char] = s
  #}
  op.on('-S', '--all-string', 'disable automatic type conversion', TrueClass) {|b|
    @opts[:all_string] = b
  }
  op.on('-t', '--time-column NAME', 'name of the time column') {|s|
    @opts[:time_column] = s
  }
  op.on('-T', '--time-format FORMAT', 'strftime(3) format of the time column') {|s|
    @opts[:time_format] = s
  }
  op.on('--time-value TIME', 'value of the time column') {|s|
    if s.to_i.to_s == s
      @opts[:time_value] = s.to_i
    else
      require 'time'
      @opts[:time_value] = Time.parse(s).to_i
    end
  }
  op.on('-e', '--encoding NAME', "text encoding") {|s|
    @opts[:encoding] = s
  }
  op.on('-C', '--compress NAME', 'compression format name [plain, gzip] (default: auto detect)') {|s|
    @opts[:compress] = s
  }
end

#parse(io, error, &block) ⇒ Object 



496
497
498
499
500
501
502
503
504
505
 
# File 'lib/td/file_reader.rb', line 496

def parse(io, error, &block)
  factory = compose_factory
  parser = factory.call(io, error)
  begin
    while record = parser.forward
      block.call(record)
    end
  rescue EOFError
  end
end

#set_format_template(name) ⇒ Object 



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
 
# File 'lib/td/file_reader.rb', line 376

def set_format_template(name)
  case name
  when 'csv'
    @format = 'text'
    @opts[:delimiter_expr] = /,/
  when 'tsv'
    @format = 'text'
    @opts[:delimiter_expr] = /\t/
  when 'apache'
    @format = name
    @opts[:column_names] = ApacheParser::COLUMNS
    @opts[:null_expr] = /\A(?:\-|)\z/
    @opts[:time_column] = 'time'
    @opts[:time_format] = ApacheParser::TIME_FORMAT
  when 'syslog'
    @format = name
    @opts[:column_names] = SyslogParser::COLUMNS
    @opts[:time_column] = 'time'
    @opts[:time_format] = SyslogParser::TIME_FORMAT
  when 'msgpack'
    @format = 'msgpack'
  when 'json'
    @format = 'json'
  else
    raise "Unknown format: #{name}"
  end
end