/usr/share/pyshared/guidata/hdf5io.py is in python-guidata 1.4.1-1.
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 | # -*- coding: utf-8 -*-
#
# Copyright © 2009-2010 CEA
# Pierre Raybaut
# Licensed under the terms of the CECILL License
# (see guidata/__init__.py for details)
"""
Reader and Writer for the serialization of DataSets into HDF5 files
"""
import sys
import h5py
import numpy as np
from guidata.utils import utf8_to_unicode
class TypeConverter(object):
def __init__(self, to_type, from_type=None):
self._to_type = to_type
if from_type:
self._from_type = from_type
else:
self._from_type = to_type
def to_hdf(self, value):
try:
return self._to_type(value)
except:
print >>sys.stderr, "ERR", repr(value)
raise
def from_hdf(self, value):
return self._from_type(value)
unicode_hdf = TypeConverter(lambda x: x.encode("utf-8"), utf8_to_unicode)
int_hdf = TypeConverter(int)
class Attr(object):
"""Helper class representing class attribute that
should be saved/restored to/from a corresponding HDF5 attribute
hdf_name : name of the attribute in the HDF5 file
struct_name : name of the attribute in the object (default to hdf_name)
type : attribute type (guess it if None)
optional : indicates whether we should fail if the attribute is not present
"""
def __init__(self, hdf_name, struct_name=None, type=None, optional=False):
self.hdf_name = hdf_name
if struct_name is None:
struct_name = hdf_name
self.struct_name = struct_name
self.type = type
self.optional = optional
def get_value(self, struct):
if self.optional:
return getattr(struct, self.struct_name, None)
else:
return getattr(struct, self.struct_name)
def set_value(self, struct, value):
setattr(struct, self.struct_name, value)
def save(self, group, struct):
value = self.get_value(struct)
if self.optional and value is None:
#print ".-", self.hdf_name, value
if self.hdf_name in group.attrs:
del group.attrs[self.hdf_name]
return
if self.type is not None:
value = self.type.to_hdf(value)
#print ".", self.hdf_name, value, self.optional
try:
group.attrs[self.hdf_name] = value
except:
print >>sys.stderr, "ERROR saving:", repr(value), "into", self.hdf_name
raise
def load(self, group, struct):
#print "LoadAttr:", group, self.hdf_name
if self.optional:
if self.hdf_name not in group.attrs:
self.set_value(struct, None)
return
try:
value = group.attrs[self.hdf_name]
except KeyError:
raise KeyError, 'Unable to locate attribute %s' % self.hdf_name
if self.type is not None:
value = self.type.from_hdf(value)
self.set_value(struct, value)
def createdset(group,name,value):
group.create_dataset(name,
compression=None,
#compression_opts=3,
data=value)
class Dset(Attr):
"""
Generic load/save for an hdf5 dataset:
scalar=float -> used to convert the value when it is scalar
"""
def __init__(self, hdf_name, struct_name=None, type=None, scalar=None,
optional=False):
Attr.__init__(self, hdf_name, struct_name, type, optional)
self.scalar = scalar
def save(self, group, struct):
value = self.get_value(struct)
if isinstance(value, float):
value = np.float64(value)
elif isinstance(value, int):
value = np.int32(value)
if value is None or value.size==0:
value = np.array([0.0])
if value.shape == ():
value = value.reshape( (1,) )
group.require_dataset(self.hdf_name, shape=value.shape,
dtype=value.dtype, data=value,
compression="gzip", compression_opts=1)
def load(self, group, struct):
if self.optional:
if self.hdf_name not in group:
self.set_value(struct, None)
return
try:
value = group[self.hdf_name][...]
except KeyError:
raise KeyError, 'Unable to locate dataset %s' % self.hdf_name
if self.scalar is not None:
value = self.scalar(value)
self.set_value(struct, value)
class Dlist(Dset):
def get_value(self, struct):
return np.array( getattr(struct, self.struct_name) )
def set_value(self, struct, value):
setattr(struct, self.struct_name, list(value))
class H5Store(object):
def __init__(self, filename):
self.filename = filename
self.h5 = None
def open(self, mode="a"):
"""Open an hdf5 file"""
if self.h5:
return self.h5
try:
self.h5 = h5py.File(self.filename,mode=mode)
except Exception:
print >>sys.stderr, "Error trying to load:", self.filename, "in mode:", mode
raise
return self.h5
def close(self):
if self.h5:
self.h5.close()
self.h5 = None
def generic_save(self, parent, source, structure):
"""save the data from source into the file using 'structure'
as a descriptor.
structure is a list of Attribute Descriptor (Attr, Dset, Dlist or anything
with a save interface) that describe the conversion of data and the name
of the attribute in the source and in the file
"""
for instr in structure:
instr.save(parent, source)
def generic_load(self, parent, dest, structure):
"""load the data from the file into dest using 'structure'
as a descriptor.
structure is the same as in generic_save
"""
for instr in structure:
try:
instr.load(parent, dest)
except Exception:
print >>sys.stderr, "Error loading HDF5 item:", instr.hdf_name
raise
class HDF5Writer(H5Store):
"""Writer for HDF5 files"""
def __init__(self, filename):
super(HDF5Writer, self).__init__(filename)
self.open("w")
self.option = []
def get_parent_group(self):
parent = self.h5
for p in self.option[:-1]:
parent = parent.require_group( p )
return parent
def write_any(self, val):
group = self.get_parent_group()
group.attrs[self.option[-1]] = val
def write_float(self, val):
group = self.get_parent_group()
group.attrs[self.option[-1]] = val
write_int = write_float
def write_bool(self, val):
self.write_int( int(val))
def write_unicode(self, val):
group = self.get_parent_group()
group.attrs[self.option[-1]] = val.encode("utf-8")
def write_array(self, val):
group = self.get_parent_group()
group[self.option[-1]] = val
write_sequence = write_any
def write_none(self):
group = self.get_parent_group()
group.attrs[self.option[-1]] = ""
def begin(self, section):
self.option.append(section)
def end(self, section):
sect = self.option.pop(-1)
assert sect == section
class HDF5Reader(H5Store):
"""Writer for HDF5 files"""
def __init__(self, filename):
super(HDF5Reader, self).__init__(filename)
self.open("r")
self.option = []
def get_parent_group(self):
parent = self.h5
for p in self.option[:-1]:
parent = parent.require_group( p )
return parent
def read_any(self):
group = self.get_parent_group()
print self.option
return group.attrs[self.option[-1]]
def begin(self, section):
self.option.append(section)
def end(self, section):
sect = self.option.pop(-1)
assert sect == section
def read_int(self):
return int(self.read_any())
def read_float(self):
return float(self.read_any())
def read_unicode(self):
return unicode(self.read_any(), "utf-8")
def read_array(self):
group = self.get_parent_group()
return group[self.option[-1]][...]
def read_sequence(self):
group = self.get_parent_group()
return list(group.attrs[self.option[-1]])
read_none = read_any
|