This file is indexed.

/usr/lib/python2.7/dist-packages/pyfits/diff.py is in python-pyfits 1:3.4-4.

This file is owned by root:root, with mode 0o644.

The actual contents of the file can be viewed below.

   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
 978
 979
 980
 981
 982
 983
 984
 985
 986
 987
 988
 989
 990
 991
 992
 993
 994
 995
 996
 997
 998
 999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
1260
1261
1262
1263
1264
1265
1266
1267
1268
1269
1270
1271
1272
1273
1274
1275
1276
1277
1278
1279
1280
1281
1282
1283
1284
1285
1286
1287
1288
1289
1290
1291
1292
1293
1294
1295
1296
1297
1298
1299
1300
"""
Facilities for diffing two FITS files.  Includes objects for diffing entire
FITS files, individual HDUs, FITS headers, or just FITS data.

Used to implement the fitsdiff program.
"""


import difflib
import fnmatch
import functools
import glob
import textwrap
import os.path

from collections import defaultdict
from itertools import islice

import numpy as np

from .extern import six
from .extern.six import string_types, StringIO
from .extern.six.moves import zip, range, xrange, reduce

import pyfits
from .card import Card, BLANK_CARD
from .header import Header
# HDUList is used in one of the doctests
from .hdu.hdulist import fitsopen  # pylint: disable=W0611
from .hdu.table import _TableLikeHDU
from .py3compat import getargspec
from .util import indent


__all__ = ['FITSDiff', 'HDUDiff', 'HeaderDiff', 'ImageDataDiff', 'RawDataDiff',
           'TableDataDiff']

# Column attributes of interest for comparison
_COL_ATTRS = [('unit', 'units'), ('null', 'null values'), ('bscale', 'bscales'),
              ('bzero', 'bzeros'), ('disp', 'display formats'),
              ('dim', 'dimensions')]


# Smaller default shift-width for indent:
indent = functools.partial(indent, width=2)


class _BaseDiff(object):
    """
    Base class for all FITS diff objects.

    When instantiating a FITS diff object, the first two arguments are always
    the two objects to diff (two FITS files, two FITS headers, etc.).
    Instantiating a ``_BaseDiff`` also causes the diff itself to be executed.
    The returned ``_BaseDiff`` instance has a number of attribute that describe
    the results of the diff operation.

    The most basic attribute, present on all ``_BaseDiff`` instances, is
    ``.identical`` which is `True` if the two objects being compared are
    identical according to the diff method for objects of that type.
    """

    def __init__(self, a, b):
        """
        The ``_BaseDiff`` class does not implement a ``_diff`` method and
        should not be instantiated directly. Instead instantiate the
        appropriate subclass of ``_BaseDiff`` for the objects being compared
        (for example, use `HeaderDiff` to compare two `Header` objects.
        """

        self.a = a
        self.b = b

        # For internal use in report output
        self._fileobj = None
        self._indent = 0

        self._diff()

    def __nonzero__(self):
        """
        A ``_BaseDiff`` object acts as `True` in a boolean context if the two
        objects compared are identical.  Otherwise it acts as `False`.
        """

        return not self.identical

    if six.PY3:
        __bool__ = __nonzero__
        del __nonzero__

    @classmethod
    def fromdiff(cls, other, a, b):
        """
        Returns a new Diff object of a specific subclass from an existing diff
        object, passing on the values for any arguments they share in common
        (such as ignore_keywords).

        For example::

            >>> from pyfits import HDUList, Header, FITSDiff
            >>> hdul1, hdul2 = HDUList(), HDUList()
            >>> headera, headerb = Header(), Header()
            >>> fd = FITSDiff(hdul1, hdul2, ignore_keywords=['*'])
            >>> hd = HeaderDiff.fromdiff(fd, headera, headerb)
            >>> list(hd.ignore_keywords)
            ['*']
        """

        args, _, _, _ = getargspec(cls.__init__)
        # The first 3 arguments of any Diff initializer are self, a, and b.
        kwargs = {}
        for arg in args[3:]:
            if hasattr(other, arg):
                kwargs[arg] = getattr(other, arg)

        return cls(a, b, **kwargs)

    @property
    def identical(self):
        """
        `True` if all the ``.diff_*`` attributes on this diff instance are
        empty, implying that no differences were found.

        Any subclass of ``_BaseDiff`` must have at least one ``.diff_*``
        attribute, which contains a non-empty value if and only if some
        difference was found between the two objects being compared.
        """

        return not any(getattr(self, attr) for attr in self.__dict__
                       if attr.startswith('diff_'))

    def report(self, fileobj=None, indent=0, clobber=False):
        """
        Generates a text report on the differences (if any) between two
        objects, and either returns it as a string or writes it to a file-like
        object.

        Parameters
        ----------
        fileobj : file-like object, string, or None (optional)
            If `None`, this method returns the report as a string. Otherwise it
            returns `None` and writes the report to the given file-like object
            (which must have a ``.write()`` method at a minimum), or to a new
            file at the path specified.

        indent : int
            The number of 4 space tabs to indent the report.

        clobber : bool
            Whether the report output should overwrite an existing file, when
            fileobj is specified as a path.

        Returns
        -------
        report : str or None
        """

        return_string = False
        filepath = None

        if isinstance(fileobj, string_types):
            if os.path.exists(fileobj) and not clobber:
                raise IOError("File {0} exists, aborting (pass in "
                              "clobber=True to overwrite)".format(fileobj))
            else:
                filepath = fileobj
                fileobj = open(filepath, 'w')
        elif fileobj is None:
            fileobj = StringIO()
            return_string = True

        self._fileobj = fileobj
        self._indent = indent  # This is used internally by _writeln

        try:
            self._report()
        finally:
            if filepath:
                fileobj.close()

        if return_string:
            return fileobj.getvalue()

    def _writeln(self, text):
        self._fileobj.write(indent(text, self._indent) + '\n')

    def _diff(self):
        raise NotImplementedError

    def _report(self):
        raise NotImplementedError


class FITSDiff(_BaseDiff):
    """Diff two FITS files by filename, or two `HDUList` objects.

    `FITSDiff` objects have the following diff attributes:

    - ``diff_hdu_count``: If the FITS files being compared have different
      numbers of HDUs, this contains a 2-tuple of the number of HDUs in each
      file.

    - ``diff_hdus``: If any HDUs with the same index are different, this
      contains a list of 2-tuples of the HDU index and the `HDUDiff` object
      representing the differences between the two HDUs.
    """

    def __init__(self, a, b, ignore_keywords=[], ignore_comments=[],
                 ignore_fields=[], numdiffs=10, tolerance=0.0,
                 ignore_blanks=True, ignore_blank_cards=True):
        """
        Parameters
        ----------
        a : str or `HDUList`
            The filename of a FITS file on disk, or an `HDUList` object.

        b : str or `HDUList`
            The filename of a FITS file on disk, or an `HDUList` object to
            compare to the first file.

        ignore_keywords : sequence, optional
            Header keywords to ignore when comparing two headers; the presence
            of these keywords and their values are ignored.  Wildcard strings
            may also be included in the list.

        ignore_comments : sequence, optional
            A list of header keywords whose comments should be ignored in the
            comparison.  May contain wildcard strings as with ignore_keywords.

        ignore_fields : sequence, optional
            The (case-insensitive) names of any table columns to ignore if any
            table data is to be compared.

        numdiffs : int, optional
            The number of pixel/table values to output when reporting HDU data
            differences.  Though the count of differences is the same either
            way, this allows controlling the number of different values that
            are kept in memory or output.  If a negative value is given, then
            numdiffs is treated as unlimited (default: 10).

        tolerance : float, optional
            The relative difference to allow when comparing two float values
            either in header values, image arrays, or table columns
            (default: 0.0).

        ignore_blanks : bool, optional
            Ignore extra whitespace at the end of string values either in
            headers or data. Extra leading whitespace is not ignored
            (default: True).

        ignore_blank_cards : bool, optional
            Ignore all cards that are blank, i.e. they only contain
            whitespace (default: True).
        """

        if isinstance(a, string_types):
            try:
                a = fitsopen(a)
            except Exception:
                excls, exc = sys.exc_info()[:2]
                raise IOError("error opening file a (%s): %s: %s" %
                              (a, exc.__class__.__name__, exc.args[0]))
            close_a = True
        else:
            close_a = False

        if isinstance(b, string_types):
            try:
                b = fitsopen(b)
            except Exception:
                excls, exc = sys.exc_info()[:2]
                raise IOError("error opening file b (%s): %s: %s" %
                              (b, exc.__class__.__name__, exc.args[0]))
            close_b = True
        else:
            close_b = False

        # Normalize keywords/fields to ignore to upper case
        self.ignore_keywords = set(k.upper() for k in ignore_keywords)
        self.ignore_comments = set(k.upper() for k in ignore_comments)
        self.ignore_fields = set(k.upper() for k in ignore_fields)

        self.numdiffs = numdiffs
        self.tolerance = tolerance
        self.ignore_blanks = ignore_blanks
        self.ignore_blank_cards = ignore_blank_cards

        self.diff_hdu_count = ()
        self.diff_hdus = []

        try:
            super(FITSDiff, self).__init__(a, b)
        finally:
            if close_a:
                a.close()
            if close_b:
                b.close()

    def _diff(self):
        if len(self.a) != len(self.b):
            self.diff_hdu_count = (len(self.a), len(self.b))

        # For now, just compare the extensions one by one in order...might
        # allow some more sophisticated types of diffing later...
        # TODO: Somehow or another simplify the passing around of diff
        # options--this will become important as the number of options grows
        for idx in range(min(len(self.a), len(self.b))):
            hdu_diff = HDUDiff.fromdiff(self, self.a[idx], self.b[idx])

            if not hdu_diff.identical:
                self.diff_hdus.append((idx, hdu_diff))

    def _report(self):
        wrapper = textwrap.TextWrapper(initial_indent='  ',
                                       subsequent_indent='  ')

        # print out heading and parameter values
        filenamea = self.a.filename()
        if not filenamea:
            filenamea = '<%s object at 0x%x>' % (self.a.__class__.__name__,
                                                 id(self.a))

        filenameb = self.b.filename()
        if not filenameb:
            filenameb = '<%s object at 0x%x>' % (self.b.__class__.__name__,
                                                 id(self.b))

        self._fileobj.write('\n')
        self._writeln(' fitsdiff: %s' % pyfits.__version__)
        self._writeln(' a: %s\n b: %s' % (filenamea, filenameb))
        if self.ignore_keywords:
            ignore_keywords = ' '.join(sorted(self.ignore_keywords))
            self._writeln(' Keyword(s) not to be compared:\n%s' %
                          wrapper.fill(ignore_keywords))

        if self.ignore_comments:
            ignore_comments = ' '.join(sorted(self.ignore_comments))
            self._writeln(' Keyword(s) whose comments are not to be compared:'
                          '\n%s' % wrapper.fill(ignore_comments))
        if self.ignore_fields:
            ignore_fields = ' '.join(sorted(self.ignore_fields))
            self._writeln(' Table column(s) not to be compared:\n%s' %
                          wrapper.fill(ignore_fields))
        self._writeln(' Maximum number of different data values to be '
                      'reported: %s' % self.numdiffs)
        self._writeln(' Data comparison level: %s' % self.tolerance)

        if self.diff_hdu_count:
            self._fileobj.write('\n')
            self._writeln('Files contain different numbers of HDUs:')
            self._writeln(' a: %d' % self.diff_hdu_count[0])
            self._writeln(' b: %d' % self.diff_hdu_count[1])

            if not self.diff_hdus:
                self._writeln('No differences found between common HDUs.')
                return
        elif not self.diff_hdus:
            self._fileobj.write('\n')
            self._writeln('No differences found.')
            return

        for idx, hdu_diff in self.diff_hdus:
            # print out the extension heading
            if idx == 0:
                self._fileobj.write('\n')
                self._writeln('Primary HDU:')
            else:
                self._fileobj.write('\n')
                self._writeln('Extension HDU %d:' % idx)
            hdu_diff.report(self._fileobj, indent=self._indent + 1)


class HDUDiff(_BaseDiff):
    """
    Diff two HDU objects, including their headers and their data (but only if
    both HDUs contain the same type of data (image, table, or unknown).

    `HDUDiff` objects have the following diff attributes:

    - ``diff_extnames``: If the two HDUs have different EXTNAME values, this
      contains a 2-tuple of the different extension names.

    - ``diff_extvers``: If the two HDUS have different EXTVER values, this
      contains a 2-tuple of the different extension versions.

    - ``diff_extlevels``: If the two HDUs have different EXTLEVEL values, this
      contains a 2-tuple of the different extension levels.

    - ``diff_extension_types``: If the two HDUs have different XTENSION values,
      this contains a 2-tuple of the different extension types.

    - ``diff_headers``: Contains a `HeaderDiff` object for the headers of the
      two HDUs. This will always contain an object--it may be determined
      whether the headers are different through ``diff_headers.identical``.

    - ``diff_data``: Contains either a `ImageDataDiff`, `TableDataDiff`, or
      `RawDataDiff` as appropriate for the data in the HDUs, and only if the
      two HDUs have non-empty data of the same type (`RawDataDiff` is used for
      HDUs containing non-empty data of an indeterminate type).
    """

    def __init__(self, a, b, ignore_keywords=[], ignore_comments=[],
                 ignore_fields=[], numdiffs=10, tolerance=0.0,
                 ignore_blanks=True, ignore_blank_cards=True):
        """
        See `FITSDiff` for explanations of the initialization parameters.
        """

        self.ignore_keywords = set(k.upper() for k in ignore_keywords)
        self.ignore_comments = set(k.upper() for k in ignore_comments)
        self.ignore_fields = set(k.upper() for k in ignore_fields)

        self.tolerance = tolerance
        self.numdiffs = numdiffs
        self.ignore_blanks = ignore_blanks

        self.diff_extnames = ()
        self.diff_extvers = ()
        self.diff_extlevels = ()
        self.diff_extension_types = ()
        self.diff_headers = None
        self.diff_data = None

        super(HDUDiff, self).__init__(a, b)

    def _diff(self):
        if self.a.name != self.b.name:
            self.diff_extnames = (self.a.name, self.b.name)

        if self.a.ver != self.b.ver:
            self.diff_extvers = (self.a.ver, self.b.ver)

        if self.a.level != self.b.level:
            self.diff_extlevels = (self.a.level, self.b.level)

        if self.a.header.get('XTENSION') != self.b.header.get('XTENSION'):
            self.diff_extension_types = (self.a.header.get('XTENSION'),
                                         self.b.header.get('XTENSION'))

        self.diff_headers = HeaderDiff.fromdiff(self, self.a.header.copy(),
                                                self.b.header.copy())

        if self.a.data is None or self.b.data is None:
            # TODO: Perhaps have some means of marking this case
            pass
        elif self.a.is_image and self.b.is_image:
            self.diff_data = ImageDataDiff.fromdiff(self, self.a.data,
                                                    self.b.data)
        elif (isinstance(self.a, _TableLikeHDU) and
              isinstance(self.b, _TableLikeHDU)):
            # TODO: Replace this if/when _BaseHDU grows a .is_table property
            self.diff_data = TableDataDiff.fromdiff(self, self.a.data,
                                                    self.b.data)
        elif not self.diff_extension_types:
            # Don't diff the data for unequal extension types that are not
            # recognized image or table types
            self.diff_data = RawDataDiff.fromdiff(self, self.a.data,
                                                  self.b.data)

    def _report(self):
        if self.identical:
            self._writeln(" No differences found.")
        if self.diff_extension_types:
            self._writeln(" Extension types differ:\n  a: %s\n  b: %s" %
                          self.diff_extension_types)
        if self.diff_extnames:
            self._writeln(" Extension names differ:\n  a: %s\n  b: %s" %
                          self.diff_extnames)
        if self.diff_extvers:
            self._writeln(" Extension versions differ:\n  a: %s\n  b: %s" %
                          self.diff_extvers)

        if self.diff_extlevels:
            self._writeln(u(" Extension levels differ:\n  a: %s\n  b: %s") %
                          self.diff_extlevels)

        if not self.diff_headers.identical:
            self._fileobj.write('\n')
            self._writeln(" Headers contain differences:")
            self.diff_headers.report(self._fileobj, indent=self._indent + 1)

        if self.diff_data is not None and not self.diff_data.identical:
            self._fileobj.write('\n')
            self._writeln(" Data contains differences:")
            self.diff_data.report(self._fileobj, indent=self._indent + 1)


class HeaderDiff(_BaseDiff):
    """
    Diff two `Header` objects.

    `HeaderDiff` objects have the following diff attributes:

    - ``diff_keyword_count``: If the two headers contain a different number of
      keywords, this contains a 2-tuple of the keyword count for each header.

    - ``diff_keywords``: If either header contains one or more keywords that
      don't appear at all in the other header, this contains a 2-tuple
      consisting of a list of the keywords only appearing in header a, and a
      list of the keywords only appearing in header b.

    - ``diff_duplicate_keywords``: If a keyword appears in both headers at
      least once, but contains a different number of duplicates (for example, a
      different number of HISTORY cards in each header), an item is added to
      this dict with the keyword as the key, and a 2-tuple of the different
      counts of that keyword as the value.  For example::

          {'HISTORY': (20, 19)}

      means that header a contains 20 HISTORY cards, while header b contains
      only 19 HISTORY cards.

    - ``diff_keyword_values``: If any of the common keyword between the two
      headers have different values, they appear in this dict.  It has a
      structure similar to ``diff_duplicate_keywords``, with the keyword as the
      key, and a 2-tuple of the different values as the value.  For example::

          {'NAXIS': (2, 3)}

      means that the NAXIS keyword has a value of 2 in header a, and a value of
      3 in header b.  This excludes any keywords matched by the
      ``ignore_keywords`` list.

    - ``diff_keyword_comments``: Like ``diff_keyword_values``, but contains
      differences between keyword comments.

    `HeaderDiff` objects also have a ``common_keywords`` attribute that lists
    all keywords that appear in both headers.
    """

    def __init__(self, a, b, ignore_keywords=[], ignore_comments=[],
                 tolerance=0.0, ignore_blanks=True, ignore_blank_cards=True):
        """
        See `FITSDiff` for explanations of the initialization parameters.
        """

        self.ignore_keywords = set(k.upper() for k in ignore_keywords)
        self.ignore_comments = set(k.upper() for k in ignore_comments)

        self.tolerance = tolerance
        self.ignore_blanks = ignore_blanks
        self.ignore_blank_cards = ignore_blank_cards

        self.ignore_keyword_patterns = set()
        self.ignore_comment_patterns = set()
        for keyword in list(self.ignore_keywords):
            keyword = keyword.upper()
            if keyword != '*' and glob.has_magic(keyword):
                self.ignore_keywords.remove(keyword)
                self.ignore_keyword_patterns.add(keyword)
        for keyword in list(self.ignore_comments):
            keyword = keyword.upper()
            if keyword != '*' and glob.has_magic(keyword):
                self.ignore_comments.remove(keyword)
                self.ignore_comment_patterns.add(keyword)

        # Keywords appearing in each header
        self.common_keywords = []

        # Set to the number of keywords in each header if the counts differ
        self.diff_keyword_count = ()

        # Set if the keywords common to each header (excluding ignore_keywords)
        # appear in different positions within the header
        # TODO: Implement this
        self.diff_keyword_positions = ()

        # Keywords unique to each header (excluding keywords in
        # ignore_keywords)
        self.diff_keywords = ()

        # Keywords that have different numbers of duplicates in each header
        # (excluding keywords in ignore_keywords)
        self.diff_duplicate_keywords = {}

        # Keywords common to each header but having different values (excluding
        # keywords in ignore_keywords)
        self.diff_keyword_values = defaultdict(lambda: [])

        # Keywords common to each header but having different comments
        # (excluding keywords in ignore_keywords or in ignore_comments)
        self.diff_keyword_comments = defaultdict(lambda: [])

        if isinstance(a, string_types):
            a = Header.fromstring(a)
        if isinstance(b, string_types):
            b = Header.fromstring(b)

        if not (isinstance(a, Header) and isinstance(b, Header)):
            raise TypeError('HeaderDiff can only diff pyfits.Header objects '
                            'or strings containing FITS headers.')

        super(HeaderDiff, self).__init__(a, b)

    # TODO: This doesn't pay much attention to the *order* of the keywords,
    # except in the case of duplicate keywords.  The order should be checked
    # too, or at least it should be an option.
    def _diff(self):
        if self.ignore_blank_cards:
            cardsa = [c for c in self.a.cards if str(c) != BLANK_CARD]
            cardsb = [c for c in self.b.cards if str(c) != BLANK_CARD]
        else:
            cardsa = list(self.a.cards)
            cardsb = list(self.b.cards)

        # build dictionaries of keyword values and comments
        def get_header_values_comments(cards):
            values = {}
            comments = {}
            for card in cards:
                value = card.value
                if self.ignore_blanks and isinstance(value, string_types):
                    value = value.rstrip()
                values.setdefault(card.keyword, []).append(value)
                comments.setdefault(card.keyword, []).append(card.comment)
            return values, comments

        valuesa, commentsa = get_header_values_comments(cardsa)
        valuesb, commentsb = get_header_values_comments(cardsb)

        # Normalize all keyword to upper-case for comparison's sake;
        # TODO: HIERARCH keywords should be handled case-sensitively I think
        keywordsa = set(k.upper() for k in valuesa)
        keywordsb = set(k.upper() for k in valuesb)

        self.common_keywords = sorted(keywordsa.intersection(keywordsb))
        if len(cardsa) != len(cardsb):
            self.diff_keyword_count = (len(cardsa), len(cardsb))

        # Any other diff attributes should exclude ignored keywords
        keywordsa = keywordsa.difference(self.ignore_keywords)
        keywordsb = keywordsb.difference(self.ignore_keywords)
        if self.ignore_keyword_patterns:
            for pattern in self.ignore_keyword_patterns:
                keywordsa = keywordsa.difference(fnmatch.filter(keywordsa,
                                                                pattern))
                keywordsb = keywordsb.difference(fnmatch.filter(keywordsb,
                                                                pattern))

        if '*' in self.ignore_keywords:
            # Any other differences between keywords are to be ignored
            return

        left_only_keywords = sorted(keywordsa.difference(keywordsb))
        right_only_keywords = sorted(keywordsb.difference(keywordsa))

        if left_only_keywords or right_only_keywords:
            self.diff_keywords = (left_only_keywords, right_only_keywords)

        # Compare count of each common keyword
        for keyword in self.common_keywords:
            if keyword in self.ignore_keywords:
                continue
            if self.ignore_keyword_patterns:
                skip = False
                for pattern in self.ignore_keyword_patterns:
                    if fnmatch.fnmatch(keyword, pattern):
                        skip = True
                        break
                if skip:
                    continue

            counta = len(valuesa[keyword])
            countb = len(valuesb[keyword])
            if counta != countb:
                self.diff_duplicate_keywords[keyword] = (counta, countb)

            # Compare keywords' values and comments
            for a, b in zip(valuesa[keyword], valuesb[keyword]):
                if diff_values(a, b, tolerance=self.tolerance):
                    self.diff_keyword_values[keyword].append((a, b))
                else:
                    # If there are duplicate keywords we need to be able to
                    # index each duplicate; if the values of a duplicate
                    # are identical use None here
                    self.diff_keyword_values[keyword].append(None)

            if not any(self.diff_keyword_values[keyword]):
                # No differences found; delete the array of Nones
                del self.diff_keyword_values[keyword]

            if '*' in self.ignore_comments or keyword in self.ignore_comments:
                continue
            if self.ignore_comment_patterns:
                skip = False
                for pattern in self.ignore_comment_patterns:
                    if fnmatch.fnmatch(keyword, pattern):
                        skip = True
                        break
                if skip:
                    continue

            for a, b in zip(commentsa[keyword], commentsb[keyword]):
                if diff_values(a, b):
                    self.diff_keyword_comments[keyword].append((a, b))
                else:
                    self.diff_keyword_comments[keyword].append(None)

            if not any(self.diff_keyword_comments[keyword]):
                del self.diff_keyword_comments[keyword]

    def _report(self):
        if self.diff_keyword_count:
            self._writeln(' Headers have different number of cards:')
            self._writeln('  a: %d' % self.diff_keyword_count[0])
            self._writeln('  b: %d' % self.diff_keyword_count[1])
        if self.diff_keywords:
            for keyword in self.diff_keywords[0]:
                if keyword in Card._commentary_keywords:
                    val = self.a[keyword][0]
                else:
                    val = self.a[keyword]
                self._writeln(' Extra keyword %-8r in a: %r' % (keyword, val))
            for keyword in self.diff_keywords[1]:
                if keyword in Card._commentary_keywords:
                    val = self.b[keyword][0]
                else:
                    val = self.b[keyword]
                self._writeln(' Extra keyword %-8r in b: %r' % (keyword, val))

        if self.diff_duplicate_keywords:
            for keyword, count in sorted(self.diff_duplicate_keywords.items()):
                self._writeln(' Inconsistent duplicates of keyword %-8r:' %
                              keyword)
                self._writeln('  Occurs %d time(s) in a, %d times in (b)' %
                              count)

        if self.diff_keyword_values or self.diff_keyword_comments:
            for keyword in self.common_keywords:
                report_diff_keyword_attr(self._fileobj, 'values',
                                         self.diff_keyword_values, keyword,
                                         ind=self._indent)
                report_diff_keyword_attr(self._fileobj, 'comments',
                                         self.diff_keyword_comments, keyword,
                                         ind=self._indent)

# TODO: It might be good if there was also a threshold option for percentage of
# different pixels: For example ignore if only 1% of the pixels are different
# within some threshold.  There are lots of possibilities here, but hold off
# for now until specific cases come up.


class ImageDataDiff(_BaseDiff):
    """
    Diff two image data arrays (really any array from a PRIMARY HDU or an IMAGE
    extension HDU, though the data unit is assumed to be "pixels").

    `ImageDataDiff` objects have the following diff attributes:

    - ``diff_dimensions``: If the two arrays contain either a different number
      of dimensions or different sizes in any dimension, this contains a
      2-tuple of the shapes of each array.  Currently no further comparison is
      performed on images that don't have the exact same dimensions.

    - ``diff_pixels``: If the two images contain any different pixels, this
      contains a list of 2-tuples of the array index where the difference was
      found, and another 2-tuple containing the different values.  For example,
      if the pixel at (0, 0) contains different values this would look like::

          [(0, 0), (1.1, 2.2)]

      where 1.1 and 2.2 are the values of that pixel in each array.  This
      array only contains up to ``self.numdiffs`` differences, for storage
      efficiency.

    - ``diff_total``: The total number of different pixels found between the
      arrays.  Although ``diff_pixels`` does not necessarily contain all the
      different pixel values, this can be used to get a count of the total
      number of differences found.

    - ``diff_ratio``: Contains the ratio of ``diff_total`` to the total number
      of pixels in the arrays.
    """

    def __init__(self, a, b, numdiffs=10, tolerance=0.0):
        """
        See `FITSDiff` for explanations of the initialization parameters.
        """

        self.numdiffs = numdiffs
        self.tolerance = tolerance

        self.diff_dimensions = ()
        self.diff_pixels = []
        self.diff_ratio = 0

        # self.diff_pixels only holds up to numdiffs differing pixels, but this
        # self.diff_total stores the total count of differences between
        # the images, but not the different values
        self.diff_total = 0

        super(ImageDataDiff, self).__init__(a, b)

    def _diff(self):
        if self.a.shape != self.b.shape:
            self.diff_dimensions = (self.a.shape, self.b.shape)
            # Don't do any further comparison if the dimensions differ
            # TODO: Perhaps we could, however, diff just the intersection
            # between the two images
            return

        # Find the indices where the values are not equal
        # If neither a nor b are floating point, ignore self.tolerance
        if not ((np.issubdtype(self.a.dtype, float) or
                 np.issubdtype(self.a.dtype, complex)) or
                (np.issubdtype(self.b.dtype, float) or
                 np.issubdtype(self.b.dtype, complex))):
            tolerance = 0
        else:
            tolerance = self.tolerance

        diffs = where_not_allclose(self.a, self.b, atol=0.0, rtol=tolerance)

        self.diff_total = len(diffs[0])

        if self.diff_total == 0:
            # Then we're done
            return

        if self.numdiffs < 0:
            numdiffs = self.diff_total
        else:
            numdiffs = self.numdiffs

        self.diff_pixels = [(idx, (self.a[idx], self.b[idx]))
                            for idx in islice(zip(*diffs), 0, numdiffs)]
        self.diff_ratio = float(self.diff_total) / float(len(self.a.flat))

    def _report(self):
        if self.diff_dimensions:
            dimsa = ' x '.join(str(d) for d in
                               reversed(self.diff_dimensions[0]))
            dimsb = ' x '.join(str(d) for d in
                               reversed(self.diff_dimensions[1]))
            self._writeln(' Data dimensions differ:')
            self._writeln('  a: %s' % dimsa)
            self._writeln('  b: %s' % dimsb)
            # For now we don't do any further comparison if the dimensions
            # differ; though in the future it might be nice to be able to
            # compare at least where the images intersect
            self._writeln(' No further data comparison performed.')
            return

        if not self.diff_pixels:
            return

        for index, values in self.diff_pixels:
            index = [x + 1 for x in reversed(index)]
            self._writeln(' Data differs at %s:' % index)
            report_diff_values(self._fileobj, values[0], values[1],
                               ind=self._indent + 1)

        if self.diff_total > self.numdiffs:
            self._writeln(' ...')
        self._writeln(' %d different pixels found (%.2f%% different).' %
                      (self.diff_total, self.diff_ratio * 100))


class RawDataDiff(ImageDataDiff):
    """
    `RawDataDiff` is just a special case of `ImageDataDiff` where the images
    are one-dimensional, and the data is treated as a 1-dimensional array of
    bytes instead of pixel values.  This is used to compare the data of two
    non-standard extension HDUs that were not recognized as containing image or
    table data.

    `ImageDataDiff` objects have the following diff attributes:

    - ``diff_dimensions``: Same as the ``diff_dimensions`` attribute of
      `ImageDataDiff` objects. Though the "dimension" of each array is just an
      integer representing the number of bytes in the data.

    - ``diff_bytes``: Like the ``diff_pixels`` attribute of `ImageDataDiff`
      objects, but renamed to reflect the minor semantic difference that these
      are raw bytes and not pixel values.  Also the indices are integers
      instead of tuples.

    - ``diff_total`` and ``diff_ratio``: Same as `ImageDataDiff`.
    """

    def __init__(self, a, b, numdiffs=10):
        """
        See `FITSDiff` for explanations of the initialization parameters.
        """

        self.diff_dimensions = ()
        self.diff_bytes = []

        super(RawDataDiff, self).__init__(a, b, numdiffs=numdiffs)

    def _diff(self):
        super(RawDataDiff, self)._diff()
        if self.diff_dimensions:
            self.diff_dimensions = (self.diff_dimensions[0][0],
                                    self.diff_dimensions[1][0])

        self.diff_bytes = [(x[0], y) for x, y in self.diff_pixels]
        del self.diff_pixels

    def _report(self):
        if self.diff_dimensions:
            self._writeln(' Data sizes differ:')
            self._writeln('  a: %d bytes' % self.diff_dimensions[0])
            self._writeln('  b: %d bytes' % self.diff_dimensions[1])
            # For now we don't do any further comparison if the dimensions
            # differ; though in the future it might be nice to be able to
            # compare at least where the images intersect
            self._writeln(' No further data comparison performed.')
            return

        if not self.diff_bytes:
            return

        for index, values in self.diff_bytes:
            self._writeln(' Data differs at byte %d:' % index)
            report_diff_values(self._fileobj, values[0], values[1],
                               ind=self._indent + 1)

        self._writeln(' ...')
        self._writeln(' %d different bytes found (%.2f%% different).' %
                      (self.diff_total, self.diff_ratio * 100))


class TableDataDiff(_BaseDiff):
    """
    Diff two table data arrays. It doesn't matter whether the data originally
    came from a binary or ASCII table--the data should be passed in as a
    recarray.

    `TableDataDiff` objects have the following diff attributes:

    - ``diff_column_count``: If the tables being compared have different
      numbers of columns, this contains a 2-tuple of the column count in each
      table.  Even if the tables have different column counts, an attempt is
      still made to compare any columns they have in common.

    - ``diff_columns``: If either table contains columns unique to that table,
      either in name or format, this contains a 2-tuple of lists. The first
      element is a list of columns (these are full `Column` objects) that
      appear only in table a.  The second element is a list of tables that
      appear only in table b.  This only lists columns with different column
      definitions, and has nothing to do with the data in those columns.

    - ``diff_column_names``: This is like ``diff_columns``, but lists only the
      names of columns unique to either table, rather than the full `Column`
      objects.

    - ``diff_column_attributes``: Lists columns that are in both tables but
      have different secondary attributes, such as TUNIT or TDISP.  The format
      is a list of 2-tuples: The first a tuple of the column name and the
      attribute, the second a tuple of the different values.

    - ``diff_values``: `TableDataDiff` compares the data in each table on a
      column-by-column basis.  If any different data is found, it is added to
      this list.  The format of this list is similar to the ``diff_pixels``
      attribute on `ImageDataDiff` objects, though the "index" consists of a
      (column_name, row) tuple.  For example::

          [('TARGET', 0), ('NGC1001', 'NGC1002')]

      shows that the tables contain different values in the 0-th row of the
      'TARGET' column.

    - ``diff_total`` and ``diff_ratio``: Same as `ImageDataDiff`.

    `TableDataDiff` objects also have a ``common_columns`` attribute that lists
    the `Column` objects for columns that are identical in both tables, and a
    ``common_column_names`` attribute which contains a set of the names of
    those columns.
    """

    def __init__(self, a, b, ignore_fields=[], numdiffs=10, tolerance=0.0):
        """
        See `FITSDiff` for explanations of the initialization parameters.
        """

        self.ignore_fields = set(ignore_fields)
        self.numdiffs = numdiffs
        self.tolerance = tolerance

        self.common_columns = []
        self.common_column_names = set()

        # self.diff_columns contains columns with different column definitions,
        # but not different column data. Column data is only compared in
        # columns that have the same definitions
        self.diff_rows = ()
        self.diff_column_count = ()
        self.diff_columns = ()

        # If two columns have the same name+format, but other attributes are
        # different (such as TUNIT or such) they are listed here
        self.diff_column_attributes = []

        # Like self.diff_columns, but just contains a list of the column names
        # unique to each table, and in the order they appear in the tables
        self.diff_column_names = ()
        self.diff_values = []

        self.diff_ratio = 0
        self.diff_total = 0

        super(TableDataDiff, self).__init__(a, b)

    def _diff(self):
        # Much of the code for comparing columns is similar to the code for
        # comparing headers--consider refactoring
        colsa = self.a.columns
        colsb = self.b.columns

        if len(colsa) != len(colsb):
            self.diff_column_count = (len(colsa), len(colsb))

        # Even if the number of columns are unequal, we still do comparison of
        # any common columns
        colsa = dict((c.name.lower(), c) for c in colsa)
        colsb = dict((c.name.lower(), c) for c in colsb)

        if '*' in self.ignore_fields:
            # If all columns are to be ignored, ignore any further differences
            # between the columns
            return

        # Keep the user's original ignore_fields list for reporting purposes,
        # but internally use a case-insensitive version
        ignore_fields = set([f.lower() for f in self.ignore_fields])

        # It might be nice if there were a cleaner way to do this, but for now
        # it'll do
        for fieldname in ignore_fields:
            fieldname = fieldname.lower()
            if fieldname in colsa:
                del colsa[fieldname]
            if fieldname in colsb:
                del colsb[fieldname]

        colsa_set = set(colsa.values())
        colsb_set = set(colsb.values())
        self.common_columns = sorted(colsa_set.intersection(colsb_set),
                                     key=lambda c: c.name)

        self.common_column_names = set([col.name.lower()
                                        for col in self.common_columns])

        left_only_columns = dict((col.name.lower(), col)
                                 for col in colsa_set.difference(colsb_set))
        right_only_columns = dict((col.name.lower(), col)
                                  for col in colsb_set.difference(colsa_set))

        if left_only_columns or right_only_columns:
            self.diff_columns = (left_only_columns, right_only_columns)
            self.diff_column_names = ([], [])

        if left_only_columns:
            for col in self.a.columns:
                if col.name.lower() in left_only_columns:
                    self.diff_column_names[0].append(col.name)

        if right_only_columns:
            for col in self.b.columns:
                if col.name.lower() in right_only_columns:
                    self.diff_column_names[1].append(col.name)

        # If the tables have a different number of rows, we don't compare the
        # columns right now.
        # TODO: It might be nice to optionally compare the first n rows where n
        # is the minimum of the row counts between the two tables.
        if len(self.a) != len(self.b):
            self.diff_rows = (len(self.a), len(self.b))
            return

        # If the tables contain no rows there's no data to compare, so we're
        # done at this point. (See ticket #178)
        if len(self.a) == len(self.b) == 0:
            return

        # Like in the old fitsdiff, compare tables on a column by column basis
        # The difficulty here is that, while FITS column names are meant to be
        # case-insensitive, PyFITS still allows, for the sake of flexibility,
        # two columns with the same name but different case.  When columns are
        # accessed in FITS tables, a case-sensitive is tried first, and failing
        # that a case-insensitive match is made.
        # It's conceivable that the same column could appear in both tables
        # being compared, but with different case.
        # Though it *may* lead to inconsistencies in these rare cases, this
        # just assumes that there are no duplicated column names in either
        # table, and that the column names can be treated case-insensitively.
        for col in self.common_columns:
            name_lower = col.name.lower()
            if name_lower in ignore_fields:
                continue

            cola = colsa[name_lower]
            colb = colsb[name_lower]

            for attr, _ in _COL_ATTRS:
                vala = getattr(cola, attr, None)
                valb = getattr(colb, attr, None)
                if diff_values(vala, valb):
                    self.diff_column_attributes.append(
                        ((col.name.upper(), attr), (vala, valb)))

            arra = self.a[col.name]
            arrb = self.b[col.name]

            if (np.issubdtype(arra.dtype, float) and
                    np.issubdtype(arrb.dtype, float)):
                diffs = where_not_allclose(arra, arrb, atol=0.0,
                                           rtol=self.tolerance)
            elif 'P' in col.format:
                diffs = ([idx for idx in xrange(len(arra))
                          if not np.allclose(arra[idx], arrb[idx], atol=0.0,
                                             rtol=self.tolerance)],)
            else:
                diffs = np.where(arra != arrb)

            self.diff_total += len(set(diffs[0]))

            if self.numdiffs >= 0:
                if len(self.diff_values) >= self.numdiffs:
                    # Don't save any more diff values
                    continue

                # Add no more diff'd values than this
                max_diffs = self.numdiffs - len(self.diff_values)
            else:
                max_diffs = len(diffs[0])

            last_seen_idx = None
            for idx in islice(diffs[0], 0, max_diffs):
                if idx == last_seen_idx:
                    # Skip duplicate indices, which my occur when the column
                    # data contains multi-dimensional values; we're only
                    # interested in storing row-by-row differences
                    continue
                last_seen_idx = idx
                self.diff_values.append(((col.name, idx),
                                         (arra[idx], arrb[idx])))

        total_values = len(self.a) * len(self.a.dtype.fields)
        self.diff_ratio = float(self.diff_total) / float(total_values)

    def _report(self):
        if self.diff_column_count:
            self._writeln(' Tables have different number of columns:')
            self._writeln('  a: %d' % self.diff_column_count[0])
            self._writeln('  b: %d' % self.diff_column_count[1])

        if self.diff_column_names:
            # Show columns with names unique to either table
            for name in self.diff_column_names[0]:
                format = self.diff_columns[0][name.lower()].format
                self._writeln(' Extra column %s of format %s in a' %
                              (name, format))
            for name in self.diff_column_names[1]:
                format = self.diff_columns[1][name.lower()].format
                self._writeln(' Extra column %s of format %s in b' %
                              (name, format))

        col_attrs = dict(_COL_ATTRS)
        # Now go through each table again and show columns with common
        # names but other property differences...
        for col_attr, vals in self.diff_column_attributes:
            name, attr = col_attr
            self._writeln(' Column %s has different %s:' %
                          (name, col_attrs[attr]))
            report_diff_values(self._fileobj, vals[0], vals[1],
                               ind=self._indent + 1)

        if self.diff_rows:
            self._writeln(' Table rows differ:')
            self._writeln('  a: %s' % self.diff_rows[0])
            self._writeln('  b: %s' % self.diff_rows[1])
            self._writeln(' No further data comparison performed.')
            return

        if not self.diff_values:
            return

        # Finally, let's go through and report column data differences:
        for indx, values in self.diff_values:
            self._writeln(' Column %s data differs in row %d:' % indx)
            report_diff_values(self._fileobj, values[0], values[1],
                               ind=self._indent + 1)

        if self.diff_values and self.numdiffs < self.diff_total:
            self._writeln(' ...%d additional difference(s) found.' %
                          (self.diff_total - self.numdiffs))

        if self.diff_total > self.numdiffs:
            self._writeln(' ...')

        self._writeln(' %d different table data element(s) found '
                      '(%.2f%% different).' %
                      (self.diff_total, self.diff_ratio * 100))


def diff_values(a, b, tolerance=0.0):
    """
    Diff two scalar values.  If both values are floats they are compared to
    within the given relative tolerance.
    """

    if isinstance(a, float) and isinstance(b, float):
        if np.isnan(a) and np.isnan(b):
            return False
        return not np.allclose(a, b, tolerance, 0.0)
    else:
        return a != b


def report_diff_values(fileobj, a, b, ind=0):
    """Write a diff between two values to the specified file-like object."""

    typea = type(a)
    typeb = type(b)

    if (isinstance(a, string_types) and not isinstance(b, string_types)):
        a = repr(a).lstrip('u')
    elif (isinstance(b, string_types) and not isinstance(a, string_types)):
        b = repr(b).lstrip('u')

    if isinstance(a, (int, float, complex, np.number)):
        a = repr(a)

    if isinstance(b, (int, float, complex, np.number)):
        b = repr(b)

    if isinstance(a, np.ndarray) and isinstance(b, np.ndarray):
        diff_indices = np.where(a != b)
        num_diffs = reduce(lambda x, y: x * y,
                           (len(d) for d in diff_indices), 1)
        for idx in islice(zip(*diff_indices), 3):
            fileobj.write(indent(u('  at %r:\n') % list(idx), ind))
            report_diff_values(fileobj, a[idx], b[idx], ind=ind + 1)

        if num_diffs:
            fileobj.write(indent('  ...and at %d more indices.\n' %
                          (num_diffs - 3), ind))
        return

    padding = max(len(typea.__name__), len(typeb.__name__)) + 3

    for line in difflib.ndiff(str(a).splitlines(), str(b).splitlines()):
        if line[0] == '-':
            line = 'a>' + line[1:]
            if typea != typeb:
                typename = '(' + typea.__name__ + ') '
                line = typename.rjust(padding) + line

        elif line[0] == '+':
            line = 'b>' + line[1:]
            if typea != typeb:
                typename = '(' + typeb.__name__ + ') '
                line = typename.rjust(padding) + line
        else:
            line = ' ' + line
            if typea != typeb:
                line = ' ' * padding + line
        fileobj.write(indent('  %s\n' % line.rstrip('\n'), ind))


def report_diff_keyword_attr(fileobj, attr, diffs, keyword, ind=0):
    """
    Write a diff between two header keyword values or comments to the specified
    file-like object.
    """

    if keyword in diffs:
        vals = diffs[keyword]
        for idx, val in enumerate(vals):
            if val is None:
                continue
            if idx == 0:
                dup = ''
            else:
                dup = '[%d]' % (idx + 1)
            fileobj.write(indent(' Keyword %-8s%s has different %s:\n' %
                          (keyword, dup, attr), ind))
            report_diff_values(fileobj, val[0], val[1], ind=ind + 1)


def where_not_allclose(a, b, rtol=1e-5, atol=1e-8):
    """
    A version of numpy.allclose that returns the indices where the two arrays
    differ, instead of just a boolean value.
    """

    # Create fixed mask arrays to handle INF and NaN; currently INF and NaN
    # are handled as equivalent
    if not np.all(np.isfinite(a)):
        a = np.ma.fix_invalid(a).data
    if not np.all(np.isfinite(b)):
        b = np.ma.fix_invalid(b).data

    if atol == 0.0 and rtol == 0.0:
        # Use a faster comparison for the most simple (and common) case
        return np.where(a != b)
    return np.where(np.abs(a - b) > (atol + rtol * np.abs(b)))