This file is indexed.

/usr/bin/nib-ls is in python-nibabel 2.1.0-1.

This file is owned by root:root, with mode 0o755.

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
#! /usr/bin/python
# emacs: -*- mode: python-mode; py-indent-offset: 4; indent-tabs-mode: nil -*-
# vi: set ft=python sts=4 ts=4 sw=4 et:
### ### ### ### ### ### ### ### ### ### ### ### ### ### ### ### ### ### ### ##
#
#   See COPYING file distributed along with the NiBabel package for the
#   copyright and license terms.
#
### ### ### ### ### ### ### ### ### ### ### ### ### ### ### ### ### ### ### ##
"""
Output a summary table for neuroimaging files (resolution, dimensionality, etc.)
"""
from __future__ import division, print_function, absolute_import

import re
import sys

import numpy as np
import nibabel as nib

from math import ceil
from collections import defaultdict
from optparse import OptionParser, Option
from io import StringIO
from nibabel.py3k import asunicode
from nibabel.externals.six.moves import xrange

__author__ = 'Yaroslav Halchenko'
__copyright__ = 'Copyright (c) 2011-2016 Yaroslav Halchenko ' \
                'and NiBabel contributors'
__license__ = 'MIT'


# global verbosity switch
verbose_level = 0
MAX_UNIQUE = 1000  # maximal number of unique values to report for --counts

def _err(msg=None):
    """To return a string to signal "error" in output table"""
    if msg is None:
        msg = 'error'
    return '!' + msg

def verbose(l, msg):
    """Print `s` if `l` is less than the `verbose_level`
    """
    # TODO: consider using nibabel's logger
    if l <= int(verbose_level):
        print("%s%s" % (' ' * l, msg))


def error(msg, exit_code):
    print >> sys.stderr, msg
    sys.exit(exit_code)


def table2string(table, out=None):
    """Given list of lists figure out their common widths and print to out

    Parameters
    ----------
    table : list of lists of strings
      What is aimed to be printed
    out : None or stream
      Where to print. If None -- will print and return string

    Returns
    -------
    string if out was None
    """

    print2string = out is None
    if print2string:
        out = StringIO()

    # equalize number of elements in each row
    nelements_max = \
        len(table) and \
        max(len(x) for x in table)

    for i, table_ in enumerate(table):
        table[i] += [''] * (nelements_max - len(table_))

    # figure out lengths within each column
    atable = np.asarray(table)
    # eat whole entry while computing width for @w (for wide)
    markup_strip = re.compile('^@([lrc]|w.*)')
    col_width = [max([len(markup_strip.sub('', x))
                      for x in column]) for column in atable.T]
    string = ""
    for i, table_ in enumerate(table):
        string_ = ""
        for j, item in enumerate(table_):
            item = str(item)
            if item.startswith('@'):
                align = item[1]
                item = item[2:]
                if align not in ['l', 'r', 'c', 'w']:
                    raise ValueError('Unknown alignment %s. Known are l,r,c' %
                                     align)
            else:
                align = 'c'

            nspacesl = max(ceil((col_width[j] - len(item)) / 2.0), 0)
            nspacesr = max(col_width[j] - nspacesl - len(item), 0)

            if align in ['w', 'c']:
                pass
            elif align == 'l':
                nspacesl, nspacesr = 0, nspacesl + nspacesr
            elif align == 'r':
                nspacesl, nspacesr = nspacesl + nspacesr, 0
            else:
                raise RuntimeError('Should not get here with align=%s' % align)

            string_ += "%%%ds%%s%%%ds " \
                       % (nspacesl, nspacesr) % ('', item, '')
        string += string_.rstrip() + '\n'
    out.write(asunicode(string))

    if print2string:
        value = out.getvalue()
        out.close()
        return value


def ap(l, format_, sep=', '):
    """Little helper to enforce consistency"""
    if l == '-':
        return l
    ls = [format_ % x for x in l]
    return sep.join(ls)


def safe_get(obj, name):
    """A getattr which would return '-' if getattr fails
    """
    try:
        f = getattr(obj, 'get_' + name)
        return f()
    except Exception as e:
        verbose(2, "get_%s() failed -- %s" % (name, e))
        return '-'


def get_opt_parser():
    # use module docstring for help output
    p = OptionParser(
        usage="%s [OPTIONS] [FILE ...]\n\n" % sys.argv[0] + __doc__,
        version="%prog " + nib.__version__)

    p.add_options([
        Option("-v", "--verbose", action="count",
               dest="verbose", default=0,
               help="Make more noise.  Could be specified multiple times"),

        Option("-H", "--header-fields",
               dest="header_fields", default='',
               help="Header fields (comma separated) to be printed as well (if present)"),

        Option("-s", "--stats",
               action="store_true", dest='stats', default=False,
               help="Output basic data statistics"),

        Option("-c", "--counts",
               action="store_true", dest='counts', default=False,
               help="Output counts - number of entries for each numeric value "
                    "(useful for int ROI maps)"),

        Option("--all-counts",
               action="store_true", dest='all_counts', default=False,
               help="Output all counts, even if number of unique values > %d" % MAX_UNIQUE),

        Option("-z", "--zeros",
               action="store_true", dest='stats_zeros', default=False,
               help="Include zeros into output basic data statistics (--stats, --counts)"),
    ])

    return p


def proc_file(f, opts):
    verbose(1, "Loading %s" % f)

    row = ["@l%s" % f]
    try:
        vol = nib.load(f)
        h = vol.header
    except Exception as e:
        row += ['failed']
        verbose(2, "Failed to gather information -- %s" % str(e))
        return row

    row += [str(safe_get(h, 'data_dtype')),
            '@l[%s]' % ap(safe_get(h, 'data_shape'), '%3g'),
            '@l%s' % ap(safe_get(h, 'zooms'), '%.2f', 'x')]
    # Slope
    if hasattr(h, 'has_data_slope') and \
            (h.has_data_slope or h.has_data_intercept) and \
            not h.get_slope_inter() in [(1.0, 0.0), (None, None)]:
        row += ['@l*%.3g+%.3g' % h.get_slope_inter()]
    else:
        row += ['']

    if hasattr(h, 'extensions') and len(h.extensions):
        row += ['@l#exts: %d' % len(h.extensions)]
    else:
        row += ['']

    if opts.header_fields:
        # signals "all fields"
        if opts.header_fields == 'all':
            # TODO: might vary across file types, thus prior sensing
            # would be needed
            header_fields = h.keys()
        else:
            header_fields = opts.header_fields.split(',')

        for f in header_fields:
            if not f:  # skip empty
                continue
            try:
                row += [str(h[f])]
            except (KeyError, ValueError):
                row += [_err()]

    try:
        if (hasattr(h, 'get_qform') and hasattr(h, 'get_sform') and
                (h.get_qform() != h.get_sform()).any()):
            row += ['sform']
        else:
            row += ['']
    except Exception as e:
        verbose(2, "Failed to obtain qform or sform -- %s" % str(e))
        if isinstance(h, nib.AnalyzeHeader):
            row += ['']
        else:
            row += [_err()]

    if opts.stats or opts.counts:
        # We are doomed to load data
        try:
            d = vol.get_data()
            if not opts.stats_zeros:
                d = d[np.nonzero(d)]
            else:
                # at least flatten it -- functionality below doesn't
                # depend on the original shape, so let's use a flat view
                d = d.reshape(-1)
            if opts.stats:
                # just # of elements
                row += ["@l[%d]" % np.prod(d.shape)]
                # stats
                row += [len(d) and '@l[%.2g, %.2g]' % (np.min(d), np.max(d)) or '-']
            if opts.counts:
                items, inv = np.unique(d, return_inverse=True)
                if len(items) > 1000 and not opts.all_counts:
                    counts = _err("%d uniques. Use --all-counts" % len(items))
                else:
                    freq = np.bincount(inv)
                    counts = " ".join("%g:%d" % (i, f) for i, f in zip(items, freq))
                row += ["@l" + counts]
        except IOError as e:
            verbose(2, "Failed to obtain stats/counts -- %s" % str(e))
            row += [_err()]
    return row


def main():
    """Show must go on"""

    parser = get_opt_parser()
    (opts, files) = parser.parse_args()

    global verbose_level
    verbose_level = opts.verbose

    if verbose_level < 3:
        # suppress nibabel format-compliance warnings
        nib.imageglobals.logger.level = 50

    rows = [proc_file(f, opts) for f in files]

    print(table2string(rows))


if __name__ == '__main__':
    main()