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
|
# (C) Copyright 2021 ECMWF.
#
# This software is licensed under the terms of the Apache Licence Version 2.0
# which can be obtained at http://www.apache.org/licenses/LICENSE-2.0.
# In applying this licence, ECMWF does not waive the privileges and immunities
# granted to it by virtue of its status as an intergovernmental organisation
# nor does it submit to any jurisdiction.
#
import logging
import re
from collections import namedtuple
LOG = logging.getLogger(__name__)
Part = namedtuple("Part", ["offset", "length"])
def round_down(a, b):
return (a // b) * b
def round_up(a, b):
return ((a + b - 1) // b) * b
class HierarchicalClustering:
def __init__(self, min_clusters=5):
self.min_clusters = min_clusters
def __call__(self, parts):
clusters = [p for p in parts]
while len(clusters) > self.min_clusters:
min_dist = min(
clusters[i].offset - clusters[i - 1].offset + clusters[i - 1].length
for i in range(1, len(clusters))
)
i = 1
while i < len(clusters):
d = clusters[i].offset - clusters[i - 1].offset + clusters[i - 1].length
if d <= min_dist:
clusters[i - 1] = Part(
clusters[i - 1].offset,
clusters[i].offset
+ clusters[i].length
- clusters[i - 1].offset,
)
clusters.pop(i)
else:
i += 1
return clusters
def __repr__(self):
return f"cluster({self.min_clusters})"
class BlockGrouping:
def __init__(self, block_size):
self.block_size = block_size
def __call__(self, parts):
blocks = []
last_block_offset = -1
last_offset = 0
for offset, length in parts:
assert offset >= last_offset
block_offset = round_down(offset, self.block_size)
block_length = round_up(offset + length, self.block_size) - block_offset
if block_offset <= last_block_offset:
prev_offset, prev_length = blocks.pop()
end_offset = block_offset + block_length
prev_end_offset = prev_offset + prev_length
block_offset = min(block_offset, prev_offset)
assert block_offset == prev_offset
block_length = max(end_offset, prev_end_offset) - block_offset
blocks.append(Part(block_offset, block_length))
last_block_offset = block_offset + block_length
last_offset = offset + length
return blocks
def __repr__(self):
return f"blocked({self.block_size})"
class Automatic:
def __call__(self, parts):
smallest = min(x.length for x in parts)
range_method = round_up(max(x.length for x in parts), 1024)
while range_method >= smallest:
blocks = BlockGrouping(range_method)(parts)
range_method //= 2
# Max number of parts
return blocks
def __repr__(self):
return "auto"
class Pipe:
def __init__(self, first, second):
self.first = first
self.second = second
def __call__(self, parts):
return self.first(self.second(parts))
def __repr__(self):
return f"{self.second}|{self.first}"
class Debug:
def __call__(self, parts):
print("DEBUG", parts)
return parts
def __repr__(self):
return "debug"
HEURISTICS = {
"auto": Automatic,
"cluster": HierarchicalClustering,
"blocked": BlockGrouping,
"debug": Debug,
}
def parts_heuristics(method, statistics_gatherer):
if isinstance(method, int):
return BlockGrouping(method)
result = None
for name in method.split("|"):
if "(" in name:
m = re.match(r"(.+)\((.+)\)", name)
name = m.group(1)
args = []
for a in m.group(2).split(","):
try:
args.append(int(a))
except ValueError:
args.append(float(a))
else:
args = []
obj = HEURISTICS[name](*args)
if result is None:
result = obj
else:
result = Pipe(obj, result)
statistics_gatherer(
"parts-heuristics",
full_method=str(method),
method_args=args,
)
return result
|
