This file is indexed.

/usr/lib/python2.7/dist-packages/surfer/io.py is in python-surfer 0.7-2.

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
import os
from tempfile import mktemp

from subprocess import Popen, PIPE, check_output
import gzip
import numpy as np
import nibabel as nib
from nibabel.spatialimages import ImageFileError

from .utils import verbose

import logging
logger = logging.getLogger('surfer')


def read_scalar_data(filepath):
    """Load in scalar data from an image.

    Parameters
    ----------
    filepath : str
        path to scalar data file

    Returns
    -------
    scalar_data : numpy array
        flat numpy array of scalar data
    """
    try:
        scalar_data = nib.load(filepath).get_data()
        scalar_data = np.ravel(scalar_data, order="F")
        return scalar_data

    except ImageFileError:
        ext = os.path.splitext(filepath)[1]
        if ext == ".mgz":
            openfile = gzip.open
        elif ext == ".mgh":
            openfile = open
        else:
            raise ValueError("Scalar file format must be readable "
                             "by Nibabel or .mg{hz} format")

    fobj = openfile(filepath, "rb")
    # We have to use np.fromstring here as gzip fileobjects don't work
    # with np.fromfile; same goes for try/finally instead of with statement
    try:
        v = np.fromstring(fobj.read(4), ">i4")[0]
        if v != 1:
            # I don't actually know what versions this code will read, so to be
            # on the safe side, let's only let version 1 in for now.
            # Scalar data might also be in curv format (e.g. lh.thickness)
            # in which case the first item in the file is a magic number.
            raise NotImplementedError("Scalar data file version not supported")
        ndim1 = np.fromstring(fobj.read(4), ">i4")[0]
        ndim2 = np.fromstring(fobj.read(4), ">i4")[0]
        ndim3 = np.fromstring(fobj.read(4), ">i4")[0]
        nframes = np.fromstring(fobj.read(4), ">i4")[0]
        datatype = np.fromstring(fobj.read(4), ">i4")[0]
        # Set the number of bytes per voxel and numpy data type according to
        # FS codes
        databytes, typecode = {0: (1, ">i1"), 1: (4, ">i4"), 3: (4, ">f4"),
                               4: (2, ">h")}[datatype]
        # Ignore the rest of the header here, just seek to the data
        fobj.seek(284)
        nbytes = ndim1 * ndim2 * ndim3 * nframes * databytes
        # Read in all the data, keep it in flat representation
        # (is this ever a problem?)
        scalar_data = np.fromstring(fobj.read(nbytes), typecode)
    finally:
        fobj.close()

    return scalar_data


def read_stc(filepath):
    """Read an STC file from the MNE package

    STC files contain activations or source reconstructions
    obtained from EEG and MEG data.

    Parameters
    ----------
    filepath: string
        Path to STC file

    Returns
    -------
    data: dict
        The STC structure. It has the following keys:
           tmin           The first time point of the data in seconds
           tstep          Time between frames in seconds
           vertices       vertex indices (0 based)
           data           The data matrix (nvert * ntime)
    """
    fid = open(filepath, 'rb')

    stc = dict()

    fid.seek(0, 2)  # go to end of file
    file_length = fid.tell()
    fid.seek(0, 0)  # go to beginning of file

    # read tmin in ms
    stc['tmin'] = float(np.fromfile(fid, dtype=">f4", count=1))
    stc['tmin'] /= 1000.0

    # read sampling rate in ms
    stc['tstep'] = float(np.fromfile(fid, dtype=">f4", count=1))
    stc['tstep'] /= 1000.0

    # read number of vertices/sources
    vertices_n = int(np.fromfile(fid, dtype=">u4", count=1))

    # read the source vector
    stc['vertices'] = np.fromfile(fid, dtype=">u4", count=vertices_n)

    # read the number of timepts
    data_n = int(np.fromfile(fid, dtype=">u4", count=1))

    if ((file_length / 4 - 4 - vertices_n) % (data_n * vertices_n)) != 0:
        raise ValueError('incorrect stc file size')

    # read the data matrix
    stc['data'] = np.fromfile(fid, dtype=">f4", count=vertices_n * data_n)
    stc['data'] = stc['data'].reshape([data_n, vertices_n]).T

    # close the file
    fid.close()
    return stc


@verbose
def project_volume_data(filepath, hemi, reg_file=None, subject_id=None,
                        projmeth="frac", projsum="avg", projarg=[0, 1, .1],
                        surf="white", smooth_fwhm=3, mask_label=None,
                        target_subject=None, verbose=None):
    """Sample MRI volume onto cortical manifold.

    Note: this requires Freesurfer to be installed with correct
    SUBJECTS_DIR definition (it uses mri_vol2surf internally).

    Parameters
    ----------
    filepath : string
        Volume file to resample (equivalent to --mov)
    hemi : [lh, rh]
        Hemisphere target
    reg_file : string
        Path to TKreg style affine matrix file
    subject_id : string
        Use if file is in register with subject's orig.mgz
    projmeth : [frac, dist]
        Projection arg should be understood as fraction of cortical
        thickness or as an absolute distance (in mm)
    projsum : [avg, max, point]
        Average over projection samples, take max, or take point sample
    projarg : single float or sequence of three floats
        Single float for point sample, sequence for avg/max specifying
        start, stop, and step
    surf : string
        Target surface
    smooth_fwhm : float
        FWHM of surface-based smoothing to apply; 0 skips smoothing
    mask_label : string
        Path to label file to constrain projection; otherwise uses cortex
    target_subject : string
        Subject to warp data to in surface space after projection
    verbose : bool, str, int, or None
        If not None, override default verbose level (see surfer.verbose).
    """

    env = os.environ
    if 'FREESURFER_HOME' not in env:
        raise RuntimeError('FreeSurfer environment not defined. Define the '
                           'FREESURFER_HOME environment variable.')
    # Run FreeSurferEnv.sh if not most recent script to set PATH
    if not env['PATH'].startswith(os.path.join(env['FREESURFER_HOME'], 'bin')):
        cmd = ['bash', '-c', 'source {} && env'.format(
               os.path.join(env['FREESURFER_HOME'], 'FreeSurferEnv.sh'))]
        envout = check_output(cmd)
        env = dict(line.split('=', 1)
                   for line in envout.decode('utf-8').split('\n')
                   if '=' in line)

    # Set the basic commands
    cmd_list = ["mri_vol2surf",
                "--mov", filepath,
                "--hemi", hemi,
                "--surf", surf]

    # Specify the affine registration
    if reg_file is not None:
        cmd_list.extend(["--reg", reg_file])
    elif subject_id is not None:
        cmd_list.extend(["--regheader", subject_id])
    else:
        raise ValueError("Must specify reg_file or subject_id")

    # Specify the projection
    proj_flag = "--proj" + projmeth
    if projsum != "point":
        proj_flag += "-"
        proj_flag += projsum
    if hasattr(projarg, "__iter__"):
        proj_arg = list(map(str, projarg))
    else:
        proj_arg = [str(projarg)]
    cmd_list.extend([proj_flag] + proj_arg)

    # Set misc args
    if smooth_fwhm:
        cmd_list.extend(["--surf-fwhm", str(smooth_fwhm)])
    if mask_label is not None:
        cmd_list.extend(["--mask", mask_label])
    if target_subject is not None:
        cmd_list.extend(["--trgsubject", target_subject])

    # Execute the command
    out_file = mktemp(prefix="pysurfer-v2s", suffix='.mgz')
    cmd_list.extend(["--o", out_file])
    logger.info(" ".join(cmd_list))
    p = Popen(cmd_list, stdout=PIPE, stderr=PIPE, env=env)
    stdout, stderr = p.communicate()
    out = p.returncode
    if out:
        raise RuntimeError(("mri_vol2surf command failed "
                            "with command-line: ") + " ".join(cmd_list))

    # Read in the data
    surf_data = read_scalar_data(out_file)
    os.remove(out_file)
    return surf_data