# -*- coding: UTF-8 -*- # Copyright (c) 2018, Dirk Gütlin & Thomas Hartmann # All rights reserved. # # This file is part of the pymatreader Project, see: # https://gitlab.com/obob/pymatreader # # Redistribution and use in source and binary forms, with or without # modification, are permitted provided that the following conditions are met: # # * Redistributions of source code must retain the above copyright notice, this # list of conditions and the following disclaimer. # # * Redistributions in binary form must reproduce the above copyright notice, # this list of conditions and the following disclaimer in the documentation # and/or other materials provided with the distribution. # # THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" # AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE # IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE # ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE # LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR # CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF # SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS # INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN # CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) # ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE # POSSIBILITY OF SUCH DAMAGE. import types import sys import numpy import scipy.io if sys.version_info <= (2, 7): chr = unichr # noqa This is needed for python 2 and 3 compatibility def _import_h5py(): try: import h5py except Exception as exc: raise ImportError('h5py is required to read MATLAB files >= v7.3 ' '(%s)' % (exc,)) return h5py def _hdf5todict(hdf5_object, variable_names=None, ignore_fields=None): """ Recursively converts a hdf5 object to a python dictionary, converting all types as well. Parameters ---------- hdf5_object: Union[h5py.Group, h5py.Dataset] Object to convert. Can be a h5py File, Group or Dataset variable_names: iterable, optional Tuple or list of variables to include. If set to none, all variable are read. ignore_fields: iterable, optional Tuple or list of fields to ignore. If set to none, all fields will be read. Returns ------- dict Python dictionary """ h5py = _import_h5py() if isinstance(hdf5_object, h5py.Group): return _handle_hdf5_group(hdf5_object, variable_names=variable_names, ignore_fields=ignore_fields) elif isinstance(hdf5_object, h5py.Dataset): return _handle_hdf5_dataset(hdf5_object) elif isinstance(hdf5_object, (list, types.GeneratorType)): return [_hdf5todict(item) for item in hdf5_object] raise TypeError('Unknown type in hdf5 file') def _handle_hdf5_group(hdf5_object, variable_names=None, ignore_fields=None): all_keys = set(hdf5_object.keys()) if ignore_fields: all_keys = all_keys - set(ignore_fields) if variable_names: all_keys = all_keys & set(variable_names) return_dict = dict() for key in all_keys: return_dict[key] = _hdf5todict(hdf5_object[key], variable_names=None, ignore_fields=ignore_fields) return return_dict def _handle_hdf5_dataset(hdf5_object): if 'MATLAB_empty' in hdf5_object.attrs.keys(): data = numpy.empty((0,)) else: data = hdf5_object.value if isinstance(data, numpy.ndarray) and \ data.dtype == numpy.dtype('object'): data = [hdf5_object.file[cur_data] for cur_data in data.flatten()] if len(data) == 1 and hdf5_object.attrs['MATLAB_class'] == b'cell': data = data[0] data = data.value return _assign_types(data) data = _hdf5todict(data) return _assign_types(data) def _convert_string_hdf5(values): if values.size > 1: assigned_values = u''.join(chr(c) for c in values.flatten()) else: assigned_values = chr(values) return assigned_values def _assign_types(values): """private function, which assigns correct types to h5py extracted values from _browse_dataset()""" if type(values) == numpy.ndarray: assigned_values = _handle_ndarray(values) elif type(values) == numpy.float64: assigned_values = float(values) else: assigned_values = values return assigned_values def _handle_ndarray(values): """Handle conversion of ndarrays.""" values = numpy.squeeze(values).T if values.dtype in ("uint8", "uint16", "uint32", "uint64"): values = _handle_hdf5_strings(values) if isinstance(values, numpy.ndarray) and \ values.size == 1: values = values.item() return values def _handle_hdf5_strings(values): if values.ndim in (0, 1): values = _convert_string_hdf5(values) elif values.ndim == 2: values = [_convert_string_hdf5(cur_val) for cur_val in values] else: raise RuntimeError('String arrays with more than 2 dimensions' 'are not supported at the moment.') return values def _check_for_scipy_mat_struct(data): """ Private function to check all entries of data for occurrences of scipy.io.matlab.mio5_params.mat_struct and convert them. Parameters ========== data: any data to be checked Returns ========= object checked and converted data """ if isinstance(data, dict): for key in data: data[key] = _check_for_scipy_mat_struct(data[key]) if isinstance(data, numpy.ndarray): data = _handle_scipy_ndarray(data) return data def _handle_scipy_ndarray(data): if data.dtype == numpy.dtype('object') and not \ isinstance(data, scipy.io.matlab.mio5.MatlabFunction): as_list = [] for element in data: as_list.append(_check_for_scipy_mat_struct(element)) data = as_list elif isinstance(data.dtype.names, tuple): data = _todict_from_np_struct(data) data = _check_for_scipy_mat_struct(data) if isinstance(data, numpy.ndarray): data = numpy.array(data) return data def _todict_from_np_struct(data): data_dict = dict() for cur_field_name in data.dtype.names: try: n_items = len(data[cur_field_name]) cur_list = list() for idx in numpy.arange(n_items): cur_value = data[cur_field_name].item(idx) cur_value = _check_for_scipy_mat_struct(cur_value) cur_list.append(cur_value) data_dict[cur_field_name] = cur_list except TypeError: cur_value = data[cur_field_name].item(0) cur_value = _check_for_scipy_mat_struct(cur_value) data_dict[cur_field_name] = cur_value return data_dict