aboutsummaryrefslogtreecommitdiffstats
path: root/src/3rdparty/python/lib/python2.7/site-packages/biplist/__init__.py
blob: f9d5836dd2469fa05176b57d4d11e9fb0124f3fa (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
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
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
141
142
143
144
145
146
147
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
194
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
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
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
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
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
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
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
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
"""biplist -- a library for reading and writing binary property list files.

Binary Property List (plist) files provide a faster and smaller serialization
format for property lists on OS X. This is a library for generating binary
plists which can be read by OS X, iOS, or other clients.

The API models the plistlib API, and will call through to plistlib when
XML serialization or deserialization is required.

To generate plists with UID values, wrap the values with the Uid object. The
value must be an int.

To generate plists with NSData/CFData values, wrap the values with the
Data object. The value must be a string.

Date values can only be datetime.datetime objects.

The exceptions InvalidPlistException and NotBinaryPlistException may be 
thrown to indicate that the data cannot be serialized or deserialized as
a binary plist.

Plist generation example:
    
    from biplist import *
    from datetime import datetime
    plist = {'aKey':'aValue',
             '0':1.322,
             'now':datetime.now(),
             'list':[1,2,3],
             'tuple':('a','b','c')
             }
    try:
        writePlist(plist, "example.plist")
    except (InvalidPlistException, NotBinaryPlistException), e:
        print "Something bad happened:", e

Plist parsing example:

    from biplist import *
    try:
        plist = readPlist("example.plist")
        print plist
    except (InvalidPlistException, NotBinaryPlistException), e:
        print "Not a plist:", e
"""

from collections import namedtuple
import datetime
import io
import math
import plistlib
from struct import pack, unpack, unpack_from
from struct import error as struct_error
import sys
import time

try:
    unicode
    unicodeEmpty = r''
except NameError:
    unicode = str
    unicodeEmpty = ''
try:
    long
except NameError:
    long = int
try:
    {}.iteritems
    iteritems = lambda x: x.iteritems()
except AttributeError:
    iteritems = lambda x: x.items()

__all__ = [
    'Uid', 'Data', 'readPlist', 'writePlist', 'readPlistFromString',
    'writePlistToString', 'InvalidPlistException', 'NotBinaryPlistException'
]

# Apple uses Jan 1, 2001 as a base for all plist date/times.
apple_reference_date = datetime.datetime.utcfromtimestamp(978307200)

class Uid(object):
    """Wrapper around integers for representing UID values. This
       is used in keyed archiving."""
    integer = 0
    def __init__(self, integer):
        self.integer = integer
    
    def __repr__(self):
        return "Uid(%d)" % self.integer
    
    def __eq__(self, other):
        if isinstance(self, Uid) and isinstance(other, Uid):
            return self.integer == other.integer
        return False
    
    def __cmp__(self, other):
        return self.integer - other.integer
    
    def __lt__(self, other):
        return self.integer < other.integer
    
    def __hash__(self):
        return self.integer
    
    def __int__(self):
        return int(self.integer)

class Data(bytes):
    """Wrapper around bytes to distinguish Data values."""

class InvalidPlistException(Exception):
    """Raised when the plist is incorrectly formatted."""

class NotBinaryPlistException(Exception):
    """Raised when a binary plist was expected but not encountered."""

def readPlist(pathOrFile):
    """Raises NotBinaryPlistException, InvalidPlistException"""
    didOpen = False
    result = None
    if isinstance(pathOrFile, (bytes, unicode)):
        pathOrFile = open(pathOrFile, 'rb')
        didOpen = True
    try:
        reader = PlistReader(pathOrFile)
        result = reader.parse()
    except NotBinaryPlistException as e:
        try:
            pathOrFile.seek(0)
            result = None
            if hasattr(plistlib, 'loads'):
                contents = None
                if isinstance(pathOrFile, (bytes, unicode)):
                    with open(pathOrFile, 'rb') as f:
                        contents = f.read()
                else:
                    contents = pathOrFile.read()
                result = plistlib.loads(contents)
            else:
                result = plistlib.readPlist(pathOrFile)
            result = wrapDataObject(result, for_binary=True)
        except Exception as e:
            raise InvalidPlistException(e)
    finally:
        if didOpen:
            pathOrFile.close()
    return result

def wrapDataObject(o, for_binary=False):
    if isinstance(o, Data) and not for_binary:
        v = sys.version_info
        if not (v[0] >= 3 and v[1] >= 4):
            o = plistlib.Data(o)
    elif isinstance(o, (bytes, plistlib.Data)) and for_binary:
        if hasattr(o, 'data'):
            o = Data(o.data)
    elif isinstance(o, tuple):
        o = wrapDataObject(list(o), for_binary)
        o = tuple(o)
    elif isinstance(o, list):
        for i in range(len(o)):
            o[i] = wrapDataObject(o[i], for_binary)
    elif isinstance(o, dict):
        for k in o:
            o[k] = wrapDataObject(o[k], for_binary)
    return o

def writePlist(rootObject, pathOrFile, binary=True):
    if not binary:
        rootObject = wrapDataObject(rootObject, binary)
        if hasattr(plistlib, "dump"):
            if isinstance(pathOrFile, (bytes, unicode)):
                with open(pathOrFile, 'wb') as f:
                    return plistlib.dump(rootObject, f)
            else:
                return plistlib.dump(rootObject, pathOrFile)
        else:
            return plistlib.writePlist(rootObject, pathOrFile)
    else:
        didOpen = False
        if isinstance(pathOrFile, (bytes, unicode)):
            pathOrFile = open(pathOrFile, 'wb')
            didOpen = True
        writer = PlistWriter(pathOrFile)
        result = writer.writeRoot(rootObject)
        if didOpen:
            pathOrFile.close()
        return result

def readPlistFromString(data):
    return readPlist(io.BytesIO(data))

def writePlistToString(rootObject, binary=True):
    if not binary:
        rootObject = wrapDataObject(rootObject, binary)
        if hasattr(plistlib, "dumps"):
            return plistlib.dumps(rootObject)
        elif hasattr(plistlib, "writePlistToBytes"):
            return plistlib.writePlistToBytes(rootObject)
        else:
            return plistlib.writePlistToString(rootObject)
    else:
        ioObject = io.BytesIO()
        writer = PlistWriter(ioObject)
        writer.writeRoot(rootObject)
        return ioObject.getvalue()

def is_stream_binary_plist(stream):
    stream.seek(0)
    header = stream.read(7)
    if header == b'bplist0':
        return True
    else:
        return False

PlistTrailer = namedtuple('PlistTrailer', 'offsetSize, objectRefSize, offsetCount, topLevelObjectNumber, offsetTableOffset')
PlistByteCounts = namedtuple('PlistByteCounts', 'nullBytes, boolBytes, intBytes, realBytes, dateBytes, dataBytes, stringBytes, uidBytes, arrayBytes, setBytes, dictBytes')

class PlistReader(object):
    file = None
    contents = ''
    offsets = None
    trailer = None
    currentOffset = 0
    # Used to detect recursive object references.
    offsetsStack = []
    
    def __init__(self, fileOrStream):
        """Raises NotBinaryPlistException."""
        self.reset()
        self.file = fileOrStream
    
    def parse(self):
        return self.readRoot()
    
    def reset(self):
        self.trailer = None
        self.contents = ''
        self.offsets = []
        self.currentOffset = 0
        self.offsetsStack = []
    
    def readRoot(self):
        result = None
        self.reset()
        # Get the header, make sure it's a valid file.
        if not is_stream_binary_plist(self.file):
            raise NotBinaryPlistException()
        self.file.seek(0)
        self.contents = self.file.read()
        if len(self.contents) < 32:
            raise InvalidPlistException("File is too short.")
        trailerContents = self.contents[-32:]
        try:
            self.trailer = PlistTrailer._make(unpack("!xxxxxxBBQQQ", trailerContents))
            
            if pow(2, self.trailer.offsetSize*8) < self.trailer.offsetTableOffset:
                raise InvalidPlistException("Offset size insufficient to reference all objects.")
            
            if pow(2, self.trailer.objectRefSize*8) < self.trailer.offsetCount:
                raise InvalidPlistException("Too many offsets to represent in size of object reference representation.")
            
            offset_size = self.trailer.offsetSize * self.trailer.offsetCount
            offset = self.trailer.offsetTableOffset
            
            if offset + offset_size > pow(2, 64):
                raise InvalidPlistException("Offset table is excessively long.")
            
            if self.trailer.offsetSize > 16:
                raise InvalidPlistException("Offset size is greater than maximum integer size.")
            
            if self.trailer.objectRefSize == 0:
                raise InvalidPlistException("Object reference size is zero.")
            
            if offset >= len(self.contents) - 32:
                raise InvalidPlistException("Offset table offset is too large.")
            
            if offset < len("bplist00x"):
                raise InvalidPlistException("Offset table offset is too small.")
            
            if self.trailer.topLevelObjectNumber >= self.trailer.offsetCount:
                raise InvalidPlistException("Top level object number is larger than the number of objects.")
            
            offset_contents = self.contents[offset:offset+offset_size]
            offset_i = 0
            offset_table_length = len(offset_contents)
            
            while offset_i < self.trailer.offsetCount:
                begin = self.trailer.offsetSize*offset_i
                end = begin+self.trailer.offsetSize
                if end > offset_table_length:
                    raise InvalidPlistException("End of object is at invalid offset %d in offset table of length %d" % (end, offset_table_length))
                tmp_contents = offset_contents[begin:end]
                tmp_sized = self.getSizedInteger(tmp_contents, self.trailer.offsetSize)
                self.offsets.append(tmp_sized)
                offset_i += 1
            self.setCurrentOffsetToObjectNumber(self.trailer.topLevelObjectNumber)
            result = self.readObject()
        except TypeError as e:
            raise InvalidPlistException(e)
        return result
    
    def setCurrentOffsetToObjectNumber(self, objectNumber):
        if objectNumber > len(self.offsets) - 1:
            raise InvalidPlistException("Invalid offset number: %d" % objectNumber)
        self.currentOffset = self.offsets[objectNumber]
        if self.currentOffset in self.offsetsStack:
            raise InvalidPlistException("Recursive data structure detected in object: %d" % objectNumber)
    
    def beginOffsetProtection(self):
        self.offsetsStack.append(self.currentOffset)
        return self.currentOffset
    
    def endOffsetProtection(self, offset):
        try:
            index = self.offsetsStack.index(offset)
            self.offsetsStack = self.offsetsStack[:index]
        except ValueError as e:
            pass
    
    def readObject(self):
        protection = self.beginOffsetProtection()
        result = None
        tmp_byte = self.contents[self.currentOffset:self.currentOffset+1]
        if len(tmp_byte) != 1:
            raise InvalidPlistException("No object found at offset: %d" % self.currentOffset)
        marker_byte = unpack("!B", tmp_byte)[0]
        format = (marker_byte >> 4) & 0x0f
        extra = marker_byte & 0x0f
        self.currentOffset += 1
        
        def proc_extra(extra):
            if extra == 0b1111:
                extra = self.readObject()
            return extra
        
        # bool, null, or fill byte
        if format == 0b0000:
            if extra == 0b0000:
                result = None
            elif extra == 0b1000:
                result = False
            elif extra == 0b1001:
                result = True
            elif extra == 0b1111:
                pass # fill byte
            else:
                raise InvalidPlistException("Invalid object found at offset: %d" % (self.currentOffset - 1))
        # int
        elif format == 0b0001:
            result = self.readInteger(pow(2, extra))
        # real
        elif format == 0b0010:
            result = self.readReal(extra)
        # date
        elif format == 0b0011 and extra == 0b0011:
            result = self.readDate()
        # data
        elif format == 0b0100:
            extra = proc_extra(extra)
            result = self.readData(extra)
        # ascii string
        elif format == 0b0101:
            extra = proc_extra(extra)
            result = self.readAsciiString(extra)
        # Unicode string
        elif format == 0b0110:
            extra = proc_extra(extra)
            result = self.readUnicode(extra)
        # uid
        elif format == 0b1000:
            result = self.readUid(extra)
        # array
        elif format == 0b1010:
            extra = proc_extra(extra)
            result = self.readArray(extra)
        # set
        elif format == 0b1100:
            extra = proc_extra(extra)
            result = set(self.readArray(extra))
        # dict
        elif format == 0b1101:
            extra = proc_extra(extra)
            result = self.readDict(extra)
        else:    
            raise InvalidPlistException("Invalid object found: {format: %s, extra: %s}" % (bin(format), bin(extra)))
        self.endOffsetProtection(protection)
        return result
    
    def readContents(self, length, description="Object contents"):
        end = self.currentOffset + length
        if end >= len(self.contents) - 32:
            raise InvalidPlistException("%s extends into trailer" % description)
        elif length < 0:
            raise InvalidPlistException("%s length is less than zero" % length)
        data = self.contents[self.currentOffset:end]
        return data
    
    def readInteger(self, byteSize):
        data = self.readContents(byteSize, "Integer")
        self.currentOffset = self.currentOffset + byteSize
        return self.getSizedInteger(data, byteSize, as_number=True)
    
    def readReal(self, length):
        to_read = pow(2, length)
        data = self.readContents(to_read, "Real")
        if length == 2: # 4 bytes
            result = unpack('>f', data)[0]
        elif length == 3: # 8 bytes
            result = unpack('>d', data)[0]
        else:
            raise InvalidPlistException("Unknown Real of length %d bytes" % to_read)
        return result
    
    def readRefs(self, count):    
        refs = []
        i = 0
        while i < count:
            fragment = self.readContents(self.trailer.objectRefSize, "Object reference")
            ref = self.getSizedInteger(fragment, len(fragment))
            refs.append(ref)
            self.currentOffset += self.trailer.objectRefSize
            i += 1
        return refs
    
    def readArray(self, count):
        if not isinstance(count, (int, long)):
            raise InvalidPlistException("Count of entries in dict isn't of integer type.")
        result = []
        values = self.readRefs(count)
        i = 0
        while i < len(values):
            self.setCurrentOffsetToObjectNumber(values[i])
            value = self.readObject()
            result.append(value)
            i += 1
        return result
    
    def readDict(self, count):
        if not isinstance(count, (int, long)):
            raise InvalidPlistException("Count of keys/values in dict isn't of integer type.")
        result = {}
        keys = self.readRefs(count)
        values = self.readRefs(count)
        i = 0
        while i < len(keys):
            self.setCurrentOffsetToObjectNumber(keys[i])
            key = self.readObject()
            self.setCurrentOffsetToObjectNumber(values[i])
            value = self.readObject()
            result[key] = value
            i += 1
        return result
    
    def readAsciiString(self, length):
        if not isinstance(length, (int, long)):
            raise InvalidPlistException("Length of ASCII string isn't of integer type.")
        data = self.readContents(length, "ASCII string")
        result = unpack("!%ds" % length, data)[0]
        self.currentOffset += length
        return str(result.decode('ascii'))
    
    def readUnicode(self, length):
        if not isinstance(length, (int, long)):
            raise InvalidPlistException("Length of Unicode string isn't of integer type.")
        actual_length = length*2
        data = self.readContents(actual_length, "Unicode string")
        self.currentOffset += actual_length
        return data.decode('utf_16_be')
    
    def readDate(self):
        data = self.readContents(8, "Date")
        x = unpack(">d", data)[0]
        if math.isnan(x):
            raise InvalidPlistException("Date is NaN")
        # Use timedelta to workaround time_t size limitation on 32-bit python.
        try:
            result = datetime.timedelta(seconds=x) + apple_reference_date
        except OverflowError:
            if x > 0:
                result = datetime.datetime.max
            else:
                result = datetime.datetime.min
        self.currentOffset += 8
        return result
    
    def readData(self, length):
        if not isinstance(length, (int, long)):
            raise InvalidPlistException("Length of data isn't of integer type.")
        result = self.readContents(length, "Data")
        self.currentOffset += length
        return Data(result)
    
    def readUid(self, length):
        if not isinstance(length, (int, long)):
            raise InvalidPlistException("Uid length isn't of integer type.")
        return Uid(self.readInteger(length+1))
    
    def getSizedInteger(self, data, byteSize, as_number=False):
        """Numbers of 8 bytes are signed integers when they refer to numbers, but unsigned otherwise."""
        result = 0
        if byteSize == 0:
            raise InvalidPlistException("Encountered integer with byte size of 0.")
        # 1, 2, and 4 byte integers are unsigned
        elif byteSize == 1:
            result = unpack('>B', data)[0]
        elif byteSize == 2:
            result = unpack('>H', data)[0]
        elif byteSize == 4:
            result = unpack('>L', data)[0]
        elif byteSize == 8:
            if as_number:
                result = unpack('>q', data)[0]
            else:
                result = unpack('>Q', data)[0]
        elif byteSize <= 16:
            # Handle odd-sized or integers larger than 8 bytes
            # Don't naively go over 16 bytes, in order to prevent infinite loops.
            result = 0
            if hasattr(int, 'from_bytes'):
                result = int.from_bytes(data, 'big')
            else:
                for byte in data:
                    if not isinstance(byte, int): # Python3.0-3.1.x return ints, 2.x return str
                        byte = unpack_from('>B', byte)[0]
                    result = (result << 8) | byte
        else:
            raise InvalidPlistException("Encountered integer longer than 16 bytes.")
        return result

class HashableWrapper(object):
    def __init__(self, value):
        self.value = value
    def __repr__(self):
        return "<HashableWrapper: %s>" % [self.value]

class BoolWrapper(object):
    def __init__(self, value):
        self.value = value
    def __repr__(self):
        return "<BoolWrapper: %s>" % self.value

class FloatWrapper(object):
    _instances = {}
    def __new__(klass, value):
        # Ensure FloatWrapper(x) for a given float x is always the same object
        wrapper = klass._instances.get(value)
        if wrapper is None:
            wrapper = object.__new__(klass)
            wrapper.value = value
            klass._instances[value] = wrapper
        return wrapper
    def __repr__(self):
        return "<FloatWrapper: %s>" % self.value

class StringWrapper(object):
    __instances = {}
    
    encodedValue = None
    encoding = None
    
    def __new__(cls, value):
        '''Ensure we only have a only one instance for any string,
         and that we encode ascii as 1-byte-per character when possible'''
        
        encodedValue = None
        
        for encoding in ('ascii', 'utf_16_be'):
            try:
               encodedValue = value.encode(encoding)
            except: pass
            if encodedValue is not None:
                if encodedValue not in cls.__instances:
                    cls.__instances[encodedValue] = super(StringWrapper, cls).__new__(cls)
                    cls.__instances[encodedValue].encodedValue = encodedValue
                    cls.__instances[encodedValue].encoding = encoding
                return cls.__instances[encodedValue]
        
        raise ValueError('Unable to get ascii or utf_16_be encoding for %s' % repr(value))
    
    def __len__(self):
        '''Return roughly the number of characters in this string (half the byte length)'''
        if self.encoding == 'ascii':
            return len(self.encodedValue)
        else:
            return len(self.encodedValue)//2
    
    def __lt__(self, other):
        return self.encodedValue < other.encodedValue
    
    @property
    def encodingMarker(self):
        if self.encoding == 'ascii':
            return 0b0101
        else:
            return 0b0110
    
    def __repr__(self):
        return '<StringWrapper (%s): %s>' % (self.encoding, self.encodedValue)

class PlistWriter(object):
    header = b'bplist00bybiplist1.0'
    file = None
    byteCounts = None
    trailer = None
    computedUniques = None
    writtenReferences = None
    referencePositions = None
    wrappedTrue = None
    wrappedFalse = None
    # Used to detect recursive object references.
    objectsStack = []
    
    def __init__(self, file):
        self.reset()
        self.file = file
        self.wrappedTrue = BoolWrapper(True)
        self.wrappedFalse = BoolWrapper(False)

    def reset(self):
        self.byteCounts = PlistByteCounts(0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0)
        self.trailer = PlistTrailer(0, 0, 0, 0, 0)
        
        # A set of all the uniques which have been computed.
        self.computedUniques = set()
        # A list of all the uniques which have been written.
        self.writtenReferences = {}
        # A dict of the positions of the written uniques.
        self.referencePositions = {}
        
        self.objectsStack = []
        
    def positionOfObjectReference(self, obj):
        """If the given object has been written already, return its
           position in the offset table. Otherwise, return None."""
        return self.writtenReferences.get(obj)
        
    def writeRoot(self, root):
        """
        Strategy is:
        - write header
        - wrap root object so everything is hashable
        - compute size of objects which will be written
          - need to do this in order to know how large the object refs
            will be in the list/dict/set reference lists
        - write objects
          - keep objects in writtenReferences
          - keep positions of object references in referencePositions
          - write object references with the length computed previously
        - computer object reference length
        - write object reference positions
        - write trailer
        """
        output = self.header
        wrapped_root = self.wrapRoot(root)
        self.computeOffsets(wrapped_root, asReference=True, isRoot=True)
        self.trailer = self.trailer._replace(**{'objectRefSize':self.intSize(len(self.computedUniques))})
        self.writeObjectReference(wrapped_root, output)
        output = self.writeObject(wrapped_root, output, setReferencePosition=True)
        
        # output size at this point is an upper bound on how big the
        # object reference offsets need to be.
        self.trailer = self.trailer._replace(**{
            'offsetSize':self.intSize(len(output)),
            'offsetCount':len(self.computedUniques),
            'offsetTableOffset':len(output),
            'topLevelObjectNumber':0
            })
        
        output = self.writeOffsetTable(output)
        output += pack('!xxxxxxBBQQQ', *self.trailer)
        self.file.write(output)
    
    def beginRecursionProtection(self, obj):
        if not isinstance(obj, (set, dict, list, tuple)):
            return
        if id(obj) in self.objectsStack:
            raise InvalidPlistException("Recursive containers are not allowed in plists.")
        self.objectsStack.append(id(obj))
    
    def endRecursionProtection(self, obj):
        if not isinstance(obj, (set, dict, list, tuple)):
            return
        try:
            index = self.objectsStack.index(id(obj))
            self.objectsStack = self.objectsStack[:index]
        except ValueError as e:
            pass

    def wrapRoot(self, root):
        result = None
        self.beginRecursionProtection(root)
        
        if isinstance(root, bool):
            if root is True:
                result = self.wrappedTrue
            else:
                result = self.wrappedFalse
        elif isinstance(root, float):
            result = FloatWrapper(root)
        elif isinstance(root, set):
            n = set()
            for value in root:
                n.add(self.wrapRoot(value))
            result = HashableWrapper(n)
        elif isinstance(root, dict):
            n = {}
            for key, value in iteritems(root):
                n[self.wrapRoot(key)] = self.wrapRoot(value)
            result = HashableWrapper(n)
        elif isinstance(root, list):
            n = []
            for value in root:
                n.append(self.wrapRoot(value))
            result = HashableWrapper(n)
        elif isinstance(root, tuple):
            n = tuple([self.wrapRoot(value) for value in root])
            result = HashableWrapper(n)
        elif isinstance(root, (str, unicode)) and not isinstance(root, Data):
            result =  StringWrapper(root)
        elif isinstance(root, bytes):
            result = Data(root)
        else:
            result = root
        
        self.endRecursionProtection(root)
        return result

    def incrementByteCount(self, field, incr=1):
        self.byteCounts = self.byteCounts._replace(**{field:self.byteCounts.__getattribute__(field) + incr})

    def computeOffsets(self, obj, asReference=False, isRoot=False):
        def check_key(key):
            if key is None:
                raise InvalidPlistException('Dictionary keys cannot be null in plists.')
            elif isinstance(key, Data):
                raise InvalidPlistException('Data cannot be dictionary keys in plists.')
            elif not isinstance(key, StringWrapper):
                raise InvalidPlistException('Keys must be strings.')
        
        def proc_size(size):
            if size > 0b1110:
                size += self.intSize(size)
            return size
        # If this should be a reference, then we keep a record of it in the
        # uniques table.
        if asReference:
            if obj in self.computedUniques:
                return
            else:
                self.computedUniques.add(obj)
        
        if obj is None:
            self.incrementByteCount('nullBytes')
        elif isinstance(obj, BoolWrapper):
            self.incrementByteCount('boolBytes')
        elif isinstance(obj, Uid):
            size = self.intSize(obj.integer)
            self.incrementByteCount('uidBytes', incr=1+size)
        elif isinstance(obj, (int, long)):
            size = self.intSize(obj)
            self.incrementByteCount('intBytes', incr=1+size)
        elif isinstance(obj, FloatWrapper):
            size = self.realSize(obj)
            self.incrementByteCount('realBytes', incr=1+size)
        elif isinstance(obj, datetime.datetime):    
            self.incrementByteCount('dateBytes', incr=2)
        elif isinstance(obj, Data):
            size = proc_size(len(obj))
            self.incrementByteCount('dataBytes', incr=1+size)
        elif isinstance(obj, StringWrapper):
            size = proc_size(len(obj))
            self.incrementByteCount('stringBytes', incr=1+size)
        elif isinstance(obj, HashableWrapper):
            obj = obj.value
            if isinstance(obj, set):
                size = proc_size(len(obj))
                self.incrementByteCount('setBytes', incr=1+size)
                for value in obj:
                    self.computeOffsets(value, asReference=True)
            elif isinstance(obj, (list, tuple)):
                size = proc_size(len(obj))
                self.incrementByteCount('arrayBytes', incr=1+size)
                for value in obj:
                    asRef = True
                    self.computeOffsets(value, asReference=True)
            elif isinstance(obj, dict):
                size = proc_size(len(obj))
                self.incrementByteCount('dictBytes', incr=1+size)
                for key, value in iteritems(obj):
                    check_key(key)
                    self.computeOffsets(key, asReference=True)
                    self.computeOffsets(value, asReference=True)
        else:
            raise InvalidPlistException("Unknown object type: %s (%s)" % (type(obj).__name__, repr(obj)))

    def writeObjectReference(self, obj, output):
        """Tries to write an object reference, adding it to the references
           table. Does not write the actual object bytes or set the reference
           position. Returns a tuple of whether the object was a new reference
           (True if it was, False if it already was in the reference table)
           and the new output.
        """
        position = self.positionOfObjectReference(obj)
        if position is None:
            self.writtenReferences[obj] = len(self.writtenReferences)
            output += self.binaryInt(len(self.writtenReferences) - 1, byteSize=self.trailer.objectRefSize)
            return (True, output)
        else:
            output += self.binaryInt(position, byteSize=self.trailer.objectRefSize)
            return (False, output)

    def writeObject(self, obj, output, setReferencePosition=False):
        """Serializes the given object to the output. Returns output.
           If setReferencePosition is True, will set the position the
           object was written.
        """
        def proc_variable_length(format, length):
            result = b''
            if length > 0b1110:
                result += pack('!B', (format << 4) | 0b1111)
                result = self.writeObject(length, result)
            else:
                result += pack('!B', (format << 4) | length)
            return result
        
        def timedelta_total_seconds(td):
            # Shim for Python 2.6 compatibility, which doesn't have total_seconds.
            # Make one argument a float to ensure the right calculation.
            return (td.microseconds + (td.seconds + td.days * 24 * 3600) * 10.0**6) / 10.0**6
       
        if setReferencePosition:
            self.referencePositions[obj] = len(output)
        
        if obj is None:
            output += pack('!B', 0b00000000)
        elif isinstance(obj, BoolWrapper):
            if obj.value is False:
                output += pack('!B', 0b00001000)
            else:
                output += pack('!B', 0b00001001)
        elif isinstance(obj, Uid):
            size = self.intSize(obj.integer)
            output += pack('!B', (0b1000 << 4) | size - 1)
            output += self.binaryInt(obj.integer)
        elif isinstance(obj, (int, long)):
            byteSize = self.intSize(obj)
            root = math.log(byteSize, 2)
            output += pack('!B', (0b0001 << 4) | int(root))
            output += self.binaryInt(obj, as_number=True)
        elif isinstance(obj, FloatWrapper):
            # just use doubles
            output += pack('!B', (0b0010 << 4) | 3)
            output += self.binaryReal(obj)
        elif isinstance(obj, datetime.datetime):
            try:
                timestamp = (obj - apple_reference_date).total_seconds()
            except AttributeError:
                timestamp = timedelta_total_seconds(obj - apple_reference_date)
            output += pack('!B', 0b00110011)
            output += pack('!d', float(timestamp))
        elif isinstance(obj, Data):
            output += proc_variable_length(0b0100, len(obj))
            output += obj
        elif isinstance(obj, StringWrapper):
            output += proc_variable_length(obj.encodingMarker, len(obj))
            output += obj.encodedValue
        elif isinstance(obj, bytes):
            output += proc_variable_length(0b0101, len(obj))
            output += obj
        elif isinstance(obj, HashableWrapper):
            obj = obj.value
            if isinstance(obj, (set, list, tuple)):
                if isinstance(obj, set):
                    output += proc_variable_length(0b1100, len(obj))
                else:
                    output += proc_variable_length(0b1010, len(obj))
            
                objectsToWrite = []
                for objRef in sorted(obj) if isinstance(obj, set) else obj:
                    (isNew, output) = self.writeObjectReference(objRef, output)
                    if isNew:
                        objectsToWrite.append(objRef)
                for objRef in objectsToWrite:
                    output = self.writeObject(objRef, output, setReferencePosition=True)
            elif isinstance(obj, dict):
                output += proc_variable_length(0b1101, len(obj))
                keys = []
                values = []
                objectsToWrite = []
                for key, value in sorted(iteritems(obj)):
                    keys.append(key)
                    values.append(value)
                for key in keys:
                    (isNew, output) = self.writeObjectReference(key, output)
                    if isNew:
                        objectsToWrite.append(key)
                for value in values:
                    (isNew, output) = self.writeObjectReference(value, output)
                    if isNew:
                        objectsToWrite.append(value)
                for objRef in objectsToWrite:
                    output = self.writeObject(objRef, output, setReferencePosition=True)
        return output
    
    def writeOffsetTable(self, output):
        """Writes all of the object reference offsets."""
        all_positions = []
        writtenReferences = list(self.writtenReferences.items())
        writtenReferences.sort(key=lambda x: x[1])
        for obj,order in writtenReferences:
            # Porting note: Elsewhere we deliberately replace empty unicdoe strings
            # with empty binary strings, but the empty unicode string
            # goes into writtenReferences.  This isn't an issue in Py2
            # because u'' and b'' have the same hash; but it is in
            # Py3, where they don't.
            if bytes != str and obj == unicodeEmpty:
                obj = b''
            position = self.referencePositions.get(obj)
            if position is None:
                raise InvalidPlistException("Error while writing offsets table. Object not found. %s" % obj)
            output += self.binaryInt(position, self.trailer.offsetSize)
            all_positions.append(position)
        return output
    
    def binaryReal(self, obj):
        # just use doubles
        result = pack('>d', obj.value)
        return result
    
    def binaryInt(self, obj, byteSize=None, as_number=False):
        result = b''
        if byteSize is None:
            byteSize = self.intSize(obj)
        if byteSize == 1:
            result += pack('>B', obj)
        elif byteSize == 2:
            result += pack('>H', obj)
        elif byteSize == 4:
            result += pack('>L', obj)
        elif byteSize == 8:
            if as_number:
                result += pack('>q', obj)
            else:
                result += pack('>Q', obj)
        elif byteSize <= 16:
            try:
                result = pack('>Q', 0) + pack('>Q', obj)
            except struct_error as e:
                raise InvalidPlistException("Unable to pack integer %d: %s" % (obj, e))
        else:
            raise InvalidPlistException("Core Foundation can't handle integers with size greater than 16 bytes.")
        return result
    
    def intSize(self, obj):
        """Returns the number of bytes necessary to store the given integer."""
        # SIGNED
        if obj < 0: # Signed integer, always 8 bytes
            return 8
        # UNSIGNED
        elif obj <= 0xFF: # 1 byte
            return 1
        elif obj <= 0xFFFF: # 2 bytes
            return 2
        elif obj <= 0xFFFFFFFF: # 4 bytes
            return 4
        # SIGNED
        # 0x7FFFFFFFFFFFFFFF is the max.
        elif obj <= 0x7FFFFFFFFFFFFFFF: # 8 bytes signed
            return 8
        elif obj <= 0xffffffffffffffff: # 8 bytes unsigned
            return 16
        else:
            raise InvalidPlistException("Core Foundation can't handle integers with size greater than 8 bytes.")
    
    def realSize(self, obj):
        return 8