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
|
import pickle
import uuid
from collections import MutableMapping
from io import RawIOBase, BufferedIOBase, IOBase
from pathlib import Path
from typing import Union, Optional, Dict, Any, List
import h5py
import numpy as np
from public import public
from copy import deepcopy
from .base import TraceSet
from .. import Trace
@public
class HDF5Meta(MutableMapping):
_dataset: h5py.AttributeManager
def __init__(self, attrs: h5py.AttributeManager):
self._attrs = attrs
super().__init__()
def __getitem__(self, item):
if item not in self._attrs:
raise KeyError
return pickle.loads(self._attrs[item])
def __setitem__(self, key, value):
self._attrs[key] = np.void(pickle.dumps(value))
def __delitem__(self, key):
del self._attrs[key]
def __copy__(self):
return deepcopy(self)
def __deepcopy__(self, memodict={}):
return dict(self)
def __iter__(self):
yield from self._attrs
def __len__(self):
return len(self._attrs)
@public
class HDF5TraceSet(TraceSet):
_file: Optional[h5py.File]
_ordering: Optional[List[str]]
#_meta: Optional[HDF5Meta]
def __init__(self, *traces: Trace, _file: Optional[h5py.File] = None,
_ordering: Optional[List[str]] = None, **kwargs):
#self._meta = HDF5Meta(_file.attrs) if _file is not None else None
self._file = _file
if _ordering is None:
_ordering = [str(uuid.uuid4()) for _ in traces]
super().__init__(*traces, **kwargs, _ordering=_ordering)
@classmethod
def read(cls, input: Union[str, Path, bytes, RawIOBase, BufferedIOBase]) -> "HDF5TraceSet":
if isinstance(input, (str, Path)):
hdf5 = h5py.File(str(input), mode="r")
elif isinstance(input, IOBase):
hdf5 = h5py.File(input, mode="r")
else:
raise ValueError
kwargs = dict(hdf5.attrs)
kwargs["_ordering"] = list(kwargs["_ordering"]) if "_ordering" in kwargs else list(hdf5.keys())
traces = []
for k in kwargs["_ordering"]:
meta = dict(HDF5Meta(hdf5[k].attrs))
samples = hdf5[k]
traces.append(Trace(np.array(samples, dtype=samples.dtype), meta))
hdf5.close()
return HDF5TraceSet(*traces, **kwargs)
@classmethod
def inplace(cls, input: Union[str, Path, bytes, RawIOBase, BufferedIOBase]) -> "HDF5TraceSet":
if isinstance(input, (str, Path)):
hdf5 = h5py.File(str(input), mode="a")
elif isinstance(input, IOBase):
hdf5 = h5py.File(input, mode="a")
else:
raise ValueError
kwargs = dict(hdf5.attrs)
kwargs["_ordering"] = list(kwargs["_ordering"]) if "_ordering" in kwargs else list(hdf5.keys())
traces = []
for k in kwargs["_ordering"]:
meta = HDF5Meta(hdf5[k].attrs)
samples = hdf5[k]
traces.append(Trace(samples, meta))
return HDF5TraceSet(*traces, **kwargs, _file=hdf5)
def insert(self, index: int, value: Trace) -> Trace:
key = str(uuid.uuid4())
self._ordering.insert(index, key)
if self._file is not None:
new_samples = self._file.create_dataset(key, data=value.samples)
new_meta = HDF5Meta(new_samples.attrs)
if value.meta:
for k, v in value.meta.items():
new_meta[k] = v
value = Trace(new_samples, new_meta)
self._file.attrs["_ordering"] = self._ordering
self._traces.insert(index, value)
return value
def get(self, index: int) -> Trace:
return self[index]
def append(self, value: Trace) -> Trace:
return self.insert(len(self), value)
def remove(self, value: Trace):
if value in self._traces:
index = self._traces.index(value)
key = self._ordering[index]
self._ordering.remove(key)
if self._file:
self._file.pop(key)
self._file.attrs["_ordering"] = self._ordering
else:
raise KeyError
def save(self):
if self._file is not None:
self._file.flush()
def close(self):
if self._file is not None:
self._file.close()
# def __getattribute__(self, item):
# if super().__getattribute__("_meta") and item in super().__getattribute__("_meta"):
# return super().__getattribute__("_meta")[item]
# return super().__getattribute__(item)
#
# def __setattr__(self, key, value):
# if key in self._keys and self._meta is not None:
# self._meta[key] = value
# else:
# super().__setattr__(key, value)
def write(self, output: Union[str, Path, RawIOBase, BufferedIOBase]):
if isinstance(output, (str, Path)):
hdf5 = h5py.File(str(output), "w")
elif isinstance(output, IOBase):
hdf5 = h5py.File(output, "w")
else:
raise ValueError
for k in self._keys:
hdf5.attrs[k] = getattr(self, k)
hdf5.attrs["_ordering"] = self._ordering
for i, k in enumerate(self._ordering):
trace = self[i]
dset = hdf5.create_dataset(k, trace.samples)
if trace.meta:
meta = HDF5Meta(dset.attrs)
for k, v in trace.meta.items():
meta[k] = v
hdf5.close()
|
