Class: BSON::BSON_RUBY

Inherits:
Object
  • Object
show all
Defined in:
lib/bson/bson_ruby.rb

Overview

A BSON seralizer/deserializer in pure Ruby.

Constant Summary collapse

DEFAULT_MAX_BSON_SIZE = 
4 * 1024 * 1024
MINKEY = 
-1
EOO = 
0
NUMBER = 
1
STRING = 
2
OBJECT = 
3
ARRAY = 
4
BINARY = 
5
UNDEFINED = 
6
OID = 
7
BOOLEAN = 
8
DATE = 
9
NULL = 
10
REGEX = 
11
REF = 
12
CODE = 
13
SYMBOL = 
14
CODE_W_SCOPE = 
15
NUMBER_INT = 
16
TIMESTAMP = 
17
NUMBER_LONG = 
18
MAXKEY = 
127
NULL_BYTE = 
"\0"
UTF8_ENCODING = 
Encoding.find('utf-8')
BINARY_ENCODING = 
Encoding.find('binary')
@@max_bson_size = 
DEFAULT_MAX_BSON_SIZE

Class Method Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize ⇒ BSON_RUBY

Returns a new instance of BSON_RUBY.



49
50
51
52
 
# File 'lib/bson/bson_ruby.rb', line 49

def initialize
  @buf = ByteBuffer.new
  @encoder = BSON_RUBY
end

Class Method Details

.deserialize(buf = nil) ⇒ Object 



107
108
109
 
# File 'lib/bson/bson_ruby.rb', line 107

def self.deserialize(buf=nil)
  new.deserialize(buf)
end

.max_bson_size ⇒ Object 



79
80
81
 
# File 'lib/bson/bson_ruby.rb', line 79

def self.max_bson_size
  @@max_bson_size
end

.serialize(obj, check_keys = false, move_id = false) ⇒ Object

Serializes an object. Implemented to ensure an API compatible with BSON extension.



103
104
105
 
# File 'lib/bson/bson_ruby.rb', line 103

def self.serialize(obj, check_keys=false, move_id=false)
  new.serialize(obj, check_keys, move_id)
end

.serialize_cstr(buf, val) ⇒ Object 



83
84
85
86
 
# File 'lib/bson/bson_ruby.rb', line 83

def self.serialize_cstr(buf, val)
  buf.put_binary(to_utf8_binary(val.to_s))
  buf.put_binary(NULL_BYTE)
end

.serialize_key(buf, key) ⇒ Object

Raises:



88
89
90
91
 
# File 'lib/bson/bson_ruby.rb', line 88

def self.serialize_key(buf, key)
  raise InvalidDocument, "Key names / regex patterns must not contain the NULL byte" if key.include? "\x00"
  self.serialize_cstr(buf, key)
end

.to_utf8_binary(str) ⇒ Object 



59
60
61
 
# File 'lib/bson/bson_ruby.rb', line 59

def self.to_utf8_binary(str)
  str.encode(UTF8_ENCODING).force_encoding(BINARY_ENCODING)
end

.update_max_bson_size(connection) ⇒ Object 



75
76
77
 
# File 'lib/bson/bson_ruby.rb', line 75

def self.update_max_bson_size(connection)
  @@max_bson_size = connection.max_bson_size
end

Instance Method Details

#bson_type(o) ⇒ Object 



562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
 
# File 'lib/bson/bson_ruby.rb', line 562

def bson_type(o)
  case o
  when nil
    NULL
  when Integer
    NUMBER_INT
  when Float
    NUMBER
  when ByteBuffer
    BINARY
  when Code
    CODE_W_SCOPE
  when String
    STRING
  when Array
    ARRAY
  when Regexp
    REGEX
  when ObjectId
    OID
  when DBRef
    REF
  when true, false
    BOOLEAN
  when Time
    DATE
  when Hash
    OBJECT
  when Symbol
    SYMBOL
  when MaxKey
    MAXKEY
  when MinKey
    MINKEY
  when Numeric
    raise InvalidDocument, "Cannot serialize the Numeric type #{o.class} as BSON; only Fixum, Bignum, and Float are supported."
  when Date, DateTime
    raise InvalidDocument, "#{o.class} is not currently supported; " +
    "use a UTC Time instance instead."
  else
    if defined?(ActiveSupport::TimeWithZone) && o.is_a?(ActiveSupport::TimeWithZone)
      raise InvalidDocument, "ActiveSupport::TimeWithZone is not currently supported; " +
      "use a UTC Time instance instead."
    else
      raise InvalidDocument, "Cannot serialize #{o.class} as a BSON type; it either isn't supported or won't translate to BSON."
    end
  end
end

#deserialize(buf = nil) ⇒ Object 



195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
 
# File 'lib/bson/bson_ruby.rb', line 195

def deserialize(buf=nil)
  # If buf is nil, use @buf, assumed to contain already-serialized BSON.
  # This is only true during testing.
  if buf.is_a? String
    @buf = ByteBuffer.new(buf.unpack("C*")) if buf
  else
    @buf = ByteBuffer.new(buf.to_a) if buf
  end
  @buf.rewind
  @buf.get_int                # eat message size
  doc = BSON::OrderedHash.new
  while @buf.more?
    type = @buf.get
    case type
    when STRING, CODE
      key = deserialize_cstr(@buf)
      doc[key] = deserialize_string_data(@buf)
    when SYMBOL
      key = deserialize_cstr(@buf)
      doc[key] = deserialize_string_data(@buf).intern
    when NUMBER
      key = deserialize_cstr(@buf)
      doc[key] = deserialize_number_data(@buf)
    when NUMBER_INT
      key = deserialize_cstr(@buf)
      doc[key] = deserialize_number_int_data(@buf)
    when NUMBER_LONG
      key = deserialize_cstr(@buf)
      doc[key] = deserialize_number_long_data(@buf)
    when OID
      key = deserialize_cstr(@buf)
      doc[key] = deserialize_oid_data(@buf)
    when ARRAY
      key = deserialize_cstr(@buf)
      doc[key] = deserialize_array_data(@buf)
    when REGEX
      key = deserialize_cstr(@buf)
      doc[key] = deserialize_regex_data(@buf)
    when OBJECT
      key = deserialize_cstr(@buf)
      doc[key] = deserialize_object_data(@buf)
    when BOOLEAN
      key = deserialize_cstr(@buf)
      doc[key] = deserialize_boolean_data(@buf)
    when DATE
      key = deserialize_cstr(@buf)
      doc[key] = deserialize_date_data(@buf)
    when NULL
      key = deserialize_cstr(@buf)
      doc[key] = nil
    when UNDEFINED
      key = deserialize_cstr(@buf)
      doc[key] = nil
    when REF
      key = deserialize_cstr(@buf)
      doc[key] = deserialize_dbref_data(@buf)
    when BINARY
      key = deserialize_cstr(@buf)
      doc[key] = deserialize_binary_data(@buf)
    when CODE_W_SCOPE
      key = deserialize_cstr(@buf)
      doc[key] = deserialize_code_w_scope_data(@buf)
    when TIMESTAMP
      key = deserialize_cstr(@buf)
      doc[key] = [deserialize_number_int_data(@buf),
      deserialize_number_int_data(@buf)]
    when MAXKEY
      key = deserialize_cstr(@buf)
      doc[key] = MaxKey.new
    when MINKEY, 255 # This is currently easier than unpack the type byte as an unsigned char.
      key = deserialize_cstr(@buf)
      doc[key] = MinKey.new
    when EOO
      break
    else
      raise "Unknown type #{type}, key = #{key}"
    end
  end
  @buf.rewind
  doc
end

#deserialize_array_data(buf) ⇒ Object 



327
328
329
330
331
332
 
# File 'lib/bson/bson_ruby.rb', line 327

def deserialize_array_data(buf)
  h = deserialize_object_data(buf)
  a = []
  h.each { |k, v| a[k.to_i] = v }
  a
end

#deserialize_binary_data(buf) ⇒ Object 



389
390
391
392
393
394
 
# File 'lib/bson/bson_ruby.rb', line 389

def deserialize_binary_data(buf)
  len = buf.get_int
  type = buf.get
  len = buf.get_int if type == Binary::SUBTYPE_BYTES
  Binary.new(buf.get(len), type)
end

#deserialize_boolean_data(buf) ⇒ Object 



298
299
300
 
# File 'lib/bson/bson_ruby.rb', line 298

def deserialize_boolean_data(buf)
  buf.get == 1
end

#deserialize_code_w_scope_data(buf) ⇒ Object 



364
365
366
367
368
369
370
371
372
373
374
375
376
377
 
# File 'lib/bson/bson_ruby.rb', line 364

def deserialize_code_w_scope_data(buf)
  buf.get_int
  len = buf.get_int
  code = buf.get(len)[0..-2]
  if code.respond_to? "pack"
    code = code.pack("C*")
  end

  scope_size = buf.get_int
  buf.position -= 4
  scope = @encoder.new().deserialize(buf.get(scope_size))

  Code.new(encoded_str(code), scope)
end

#deserialize_cstr(buf) ⇒ Object 



552
553
554
555
556
557
558
559
560
 
# File 'lib/bson/bson_ruby.rb', line 552

def deserialize_cstr(buf)
  chars = ""
  while true
    b = buf.get
    break if b == 0
    chars << b.chr
  end
  encoded_str(chars)
end

#deserialize_date_data(buf) ⇒ Object 



292
293
294
295
296
 
# File 'lib/bson/bson_ruby.rb', line 292

def deserialize_date_data(buf)
  unsigned = buf.get_long()
  milliseconds = unsigned >= 2 ** 64 / 2 ? unsigned - 2**64 : unsigned
  Time.at(milliseconds.to_f / 1000.0).utc # at() takes fractional seconds
end

#deserialize_dbref_data(buf) ⇒ Object 



383
384
385
386
387
 
# File 'lib/bson/bson_ruby.rb', line 383

def deserialize_dbref_data(buf)
  ns = deserialize_string_data(buf)
  oid = deserialize_oid_data(buf)
  DBRef.new(ns, oid)
end

#deserialize_number_data(buf) ⇒ Object 



302
303
304
 
# File 'lib/bson/bson_ruby.rb', line 302

def deserialize_number_data(buf)
  buf.get_double
end

#deserialize_number_int_data(buf) ⇒ Object 



306
307
308
309
 
# File 'lib/bson/bson_ruby.rb', line 306

def deserialize_number_int_data(buf)
  unsigned = buf.get_int
  unsigned >= 2**32 / 2 ? unsigned - 2**32 : unsigned
end

#deserialize_number_long_data(buf) ⇒ Object 



311
312
313
314
 
# File 'lib/bson/bson_ruby.rb', line 311

def deserialize_number_long_data(buf)
  unsigned = buf.get_long
  unsigned >= 2 ** 64 / 2 ? unsigned - 2**64 : unsigned
end

#deserialize_object_data(buf) ⇒ Object 



316
317
318
319
320
321
322
323
324
325
 
# File 'lib/bson/bson_ruby.rb', line 316

def deserialize_object_data(buf)
  size = buf.get_int
  buf.position -= 4
  object = @encoder.new().deserialize(buf.get(size))
  if object.has_key? "$ref"
    DBRef.new(object["$ref"], object["$id"])
  else
    object
  end
end

#deserialize_oid_data(buf) ⇒ Object 



379
380
381
 
# File 'lib/bson/bson_ruby.rb', line 379

def deserialize_oid_data(buf)
  ObjectId.new(buf.get(12))
end

#deserialize_regex_data(buf) ⇒ Object 



334
335
336
337
338
339
340
341
342
 
# File 'lib/bson/bson_ruby.rb', line 334

def deserialize_regex_data(buf)
  str = deserialize_cstr(buf)
  options_str = deserialize_cstr(buf)
  opts = 0
  opts |= Regexp::IGNORECASE if options_str.include?('i')
  opts |= Regexp::MULTILINE if options_str.include?('m')
  opts |= Regexp::EXTENDED if options_str.include?('x')
  Regexp.new(str, opts)
end

#deserialize_string_data(buf) ⇒ Object 



354
355
356
357
358
359
360
361
362
 
# File 'lib/bson/bson_ruby.rb', line 354

def deserialize_string_data(buf)
  len = buf.get_int
  bytes = buf.get(len)
  str = bytes[0..-2]
  if str.respond_to? "pack"
    str = str.pack("C*")
  end
  encoded_str(str)
end

#encoded_str(str) ⇒ Object 



344
345
346
347
348
349
350
351
352
 
# File 'lib/bson/bson_ruby.rb', line 344

def encoded_str(str)
  if RUBY_VERSION >= '1.9'
    str.force_encoding("utf-8")
    if Encoding.default_internal
      str.encode!(Encoding.default_internal)
    end
  end
  str
end

#hex_dump ⇒ Object

For debugging.



278
279
280
281
282
283
284
285
286
287
288
289
290
 
# File 'lib/bson/bson_ruby.rb', line 278

def hex_dump
  str = ''
  @buf.to_a.each_with_index { |b,i|
    if (i % 8) == 0
      str << "\n" if i > 0
      str << '%4d:  ' % i
    else
      str << ' '
    end
    str << '%02X' % b
  }
  str
end

#serialize(obj, check_keys = false, move_id = false) ⇒ Object

Raises:



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
 
# File 'lib/bson/bson_ruby.rb', line 111

def serialize(obj, check_keys=false, move_id=false)
  raise(InvalidDocument, "BSON.serialize takes a Hash but got a #{obj.class}") unless obj.is_a?(Hash)
  raise "Document is null" unless obj

  @buf.rewind
  # put in a placeholder for the total size
  @buf.put_int(0)

  # Write key/value pairs. Always write _id first if it exists.
  if move_id
    if obj.has_key? '_id'
      serialize_key_value('_id', obj['_id'], false)
    elsif obj.has_key? :_id
      serialize_key_value('_id', obj[:_id], false)
    end
    obj.each {|k, v| serialize_key_value(k, v, check_keys) unless k == '_id' || k == :_id }
  else
    if obj.has_key?('_id') && obj.has_key?(:_id)
      obj['_id'] = obj.delete(:_id)
    end
    obj.each {|k, v| serialize_key_value(k, v, check_keys) }
  end

  serialize_eoo_element(@buf)
  if @buf.size > @@max_bson_size
    raise InvalidDocument, "Document is too large (#{@buf.size}). BSON documents are limited to #{@@max_bson_size} bytes."
  end
  @buf.put_int(@buf.size, 0)
  @buf
end

#serialize_array_element(buf, key, val, check_keys) ⇒ Object 



471
472
473
474
475
476
477
 
# File 'lib/bson/bson_ruby.rb', line 471

def serialize_array_element(buf, key, val, check_keys)
  # Turn array into hash with integer indices as keys
  h = BSON::OrderedHash.new
  i = 0
  val.each { |v| h[i] = v; i += 1 }
  serialize_object_element(buf, key, h, check_keys, ARRAY)
end

#serialize_binary_element(buf, key, val) ⇒ Object 



412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
 
# File 'lib/bson/bson_ruby.rb', line 412

def serialize_binary_element(buf, key, val)
  buf.put(BINARY)
  self.class.serialize_key(buf, key)

  bytes = val.to_a
  num_bytes = bytes.length
  subtype = val.respond_to?(:subtype) ? val.subtype : Binary::SUBTYPE_BYTES
  if subtype == Binary::SUBTYPE_BYTES
    buf.put_int(num_bytes + 4)
    buf.put(subtype)
    buf.put_int(num_bytes)
    buf.put_array(bytes)
  else
    buf.put_int(num_bytes)
    buf.put(subtype)
    buf.put_array(bytes)
  end
end

#serialize_boolean_element(buf, key, val) ⇒ Object 



431
432
433
434
435
 
# File 'lib/bson/bson_ruby.rb', line 431

def serialize_boolean_element(buf, key, val)
  buf.put(BOOLEAN)
  self.class.serialize_key(buf, key)
  buf.put(val ? 1 : 0)
end

#serialize_code_w_scope(buf, key, val) ⇒ Object 



535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
 
# File 'lib/bson/bson_ruby.rb', line 535

def serialize_code_w_scope(buf, key, val)
  buf.put(CODE_W_SCOPE)
  self.class.serialize_key(buf, key)

  # Make a hole for the length
  len_pos = buf.position
  buf.put_int(0)

  buf.put_int(val.code.length + 1)
  self.class.serialize_cstr(buf, val.code)
  buf.put_array(@encoder.new.serialize(val.scope).to_a)

  end_pos = buf.position
  buf.put_int(end_pos - len_pos, len_pos)
  buf.position = end_pos
end

#serialize_date_element(buf, key, val) ⇒ Object 



437
438
439
440
441
442
 
# File 'lib/bson/bson_ruby.rb', line 437

def serialize_date_element(buf, key, val)
  buf.put(DATE)
  self.class.serialize_key(buf, key)
  millisecs = (val.to_f * 1000).to_i
  buf.put_long(millisecs)
end

#serialize_dbref_element(buf, key, val) ⇒ Object 



405
406
407
408
409
410
 
# File 'lib/bson/bson_ruby.rb', line 405

def serialize_dbref_element(buf, key, val)
  oh = BSON::OrderedHash.new
  oh['$ref'] = val.namespace
  oh['$id'] = val.object_id
  serialize_object_element(buf, key, oh, false)
end

#serialize_eoo_element(buf) ⇒ Object 



396
397
398
 
# File 'lib/bson/bson_ruby.rb', line 396

def serialize_eoo_element(buf)
  buf.put(EOO)
end

#serialize_key_value(k, v, check_keys) ⇒ Object 



148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
 
# File 'lib/bson/bson_ruby.rb', line 148

def serialize_key_value(k, v, check_keys)
  k = k.to_s
  if check_keys
    if k[0] == ?$
      raise InvalidKeyName.new("key #{k} must not start with '$'")
    end
    if k.include? ?.
      raise InvalidKeyName.new("key #{k} must not contain '.'")
    end
  end
  type = bson_type(v)
  case type
  when STRING, SYMBOL
    serialize_string_element(@buf, k, v, type)
  when NUMBER, NUMBER_INT
    serialize_number_element(@buf, k, v, type)
  when OBJECT
    serialize_object_element(@buf, k, v, check_keys)
  when OID
    serialize_oid_element(@buf, k, v)
  when ARRAY
    serialize_array_element(@buf, k, v, check_keys)
  when REGEX
    serialize_regex_element(@buf, k, v)
  when BOOLEAN
    serialize_boolean_element(@buf, k, v)
  when DATE
    serialize_date_element(@buf, k, v)
  when NULL
    serialize_null_element(@buf, k)
  when REF
    serialize_dbref_element(@buf, k, v)
  when BINARY
    serialize_binary_element(@buf, k, v)
  when UNDEFINED
    serialize_null_element(@buf, k)
  when CODE_W_SCOPE
    serialize_code_w_scope(@buf, k, v)
  when MAXKEY
    serialize_max_key_element(@buf, k)
  when MINKEY
    serialize_min_key_element(@buf, k)
  else
    raise "unhandled type #{type}"
  end
end

#serialize_max_key_element(buf, key) ⇒ Object 



498
499
500
501
 
# File 'lib/bson/bson_ruby.rb', line 498

def serialize_max_key_element(buf, key)
  buf.put(MAXKEY)
  self.class.serialize_key(buf, key)
end

#serialize_min_key_element(buf, key) ⇒ Object 



503
504
505
506
 
# File 'lib/bson/bson_ruby.rb', line 503

def serialize_min_key_element(buf, key)
  buf.put(MINKEY)
  self.class.serialize_key(buf, key)
end

#serialize_null_element(buf, key) ⇒ Object 



400
401
402
403
 
# File 'lib/bson/bson_ruby.rb', line 400

def serialize_null_element(buf, key)
  buf.put(NULL)
  self.class.serialize_key(buf, key)
end

#serialize_number_element(buf, key, val, type) ⇒ Object 



444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
 
# File 'lib/bson/bson_ruby.rb', line 444

def serialize_number_element(buf, key, val, type)
  if type == NUMBER
    buf.put(type)
    self.class.serialize_key(buf, key)
    buf.put_double(val)
  else
    if val > 2**64 / 2 - 1 or val < -2**64 / 2
      raise RangeError.new("MongoDB can only handle 8-byte ints")
    end
    if val > 2**32 / 2 - 1 or val < -2**32 / 2
      buf.put(NUMBER_LONG)
      self.class.serialize_key(buf, key)
      buf.put_long(val)
    else
      buf.put(type)
      self.class.serialize_key(buf, key)
      buf.put_int(val)
    end
  end
end

#serialize_object_element(buf, key, val, check_keys, opcode = OBJECT) ⇒ Object 



465
466
467
468
469
 
# File 'lib/bson/bson_ruby.rb', line 465

def serialize_object_element(buf, key, val, check_keys, opcode=OBJECT)
  buf.put(opcode)
  self.class.serialize_key(buf, key)
  buf.put_array(@encoder.new.serialize(val, check_keys).to_a)
end

#serialize_oid_element(buf, key, val) ⇒ Object 



508
509
510
511
512
513
 
# File 'lib/bson/bson_ruby.rb', line 508

def serialize_oid_element(buf, key, val)
  buf.put(OID)
  self.class.serialize_key(buf, key)

  buf.put_array(val.to_a)
end

#serialize_regex_element(buf, key, val) ⇒ Object 



479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
 
# File 'lib/bson/bson_ruby.rb', line 479

def serialize_regex_element(buf, key, val)
  buf.put(REGEX)
  self.class.serialize_key(buf, key)

  str = val.source
  # We use serialize_key here since regex patterns aren't prefixed with
  # length (can't contain the NULL byte).
  self.class.serialize_key(buf, str)

  options = val.options
  options_str = ''
  options_str << 'i' if ((options & Regexp::IGNORECASE) != 0)
  options_str << 'm' if ((options & Regexp::MULTILINE) != 0)
  options_str << 'x' if ((options & Regexp::EXTENDED) != 0)
  options_str << val.extra_options_str if val.respond_to?(:extra_options_str)
  # Must store option chars in alphabetical order
  self.class.serialize_cstr(buf, options_str.split(//).sort.uniq.join)
end

#serialize_string_element(buf, key, val, type) ⇒ Object 



515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
 
# File 'lib/bson/bson_ruby.rb', line 515

def serialize_string_element(buf, key, val, type)
  buf.put(type)
  self.class.serialize_key(buf, key)

  # Make a hole for the length
  len_pos = buf.position
  buf.put_int(0)

  # Save the string
  start_pos = buf.position
  self.class.serialize_cstr(buf, val)
  end_pos = buf.position

  # Put the string size in front
  buf.put_int(end_pos - start_pos, len_pos)

  # Go back to where we were
  buf.position = end_pos
end

#to_a ⇒ Object 



93
94
95
 
# File 'lib/bson/bson_ruby.rb', line 93

def to_a
  @buf.to_a
end

#to_s ⇒ Object 



97
98
99
 
# File 'lib/bson/bson_ruby.rb', line 97

def to_s
  @buf.to_s
end

#unpack(arg) ⇒ Object

Returns the array stored in the buffer. Implemented to ensure an API compatible with BSON extension.



144
145
146
 
# File 'lib/bson/bson_ruby.rb', line 144

def unpack(arg)
  @buf.to_a
end