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
|
from typing import List, Dict, Tuple, Union, Optional
from fakeredis import _msgs as msgs
from fakeredis._helpers import Database, SimpleError
class TimeSeries:
def __init__(
self,
name: bytes,
database: Database,
retention: int = 0,
encoding: bytes = b"compressed",
chunk_size: int = 4096,
duplicate_policy: bytes = b"block",
ignore_max_time_diff: int = 0,
ignore_max_val_diff: int = 0,
labels: Dict[str, str] = None,
source_key: Optional[bytes] = None,
):
super().__init__()
self.name = name
self._db = database
self.retention = retention
self.encoding = encoding
self.chunk_size = chunk_size
self.duplicate_policy = duplicate_policy
self.ts_ind_map: Dict[int, int] = dict() # Map from timestamp to index in sorted_list
self.sorted_list: List[Tuple[int, float]] = list()
self.max_timestamp: int = 0
self.labels = labels or {}
self.source_key = source_key
self.ignore_max_time_diff = ignore_max_time_diff
self.ignore_max_val_diff = ignore_max_val_diff
self.rules: List[TimeSeriesRule] = list()
def add(
self, timestamp: int, value: float, duplicate_policy: Optional[bytes] = None
) -> Union[int, None, List[None]]:
if self.retention != 0 and self.max_timestamp - timestamp > self.retention:
raise SimpleError(msgs.TIMESERIES_TIMESTAMP_OLDER_THAN_RETENTION)
if duplicate_policy is None:
duplicate_policy = self.duplicate_policy
if timestamp in self.ts_ind_map: # Duplicate policy
if duplicate_policy == b"block":
raise SimpleError(msgs.TIMESERIES_DUPLICATE_POLICY_BLOCK)
if duplicate_policy == b"first":
return timestamp
ind = self.ts_ind_map[timestamp]
curr_value = self.sorted_list[ind][1]
if duplicate_policy == b"max":
value = max(curr_value, value)
elif duplicate_policy == b"min":
value = min(curr_value, value)
self.sorted_list[ind] = (timestamp, value)
return timestamp
self.sorted_list.append((timestamp, value))
self.ts_ind_map[timestamp] = len(self.sorted_list) - 1
self.rules = [rule for rule in self.rules if rule.dest_key.name in self._db]
for rule in self.rules:
rule.add_record((timestamp, value))
self.max_timestamp = max(self.max_timestamp, timestamp)
return timestamp
def incrby(self, timestamp: int, value: float) -> Union[int, None]:
if len(self.sorted_list) == 0:
return self.add(timestamp, value)
if timestamp == self.max_timestamp:
ind = self.ts_ind_map[timestamp]
self.sorted_list[ind] = (timestamp, self.sorted_list[ind][1] + value)
elif timestamp > self.max_timestamp:
ind = self.ts_ind_map[self.max_timestamp]
self.add(timestamp, self.sorted_list[ind][1] + value)
else: # timestamp < self.sorted_list[ind][0]
raise ValueError()
return timestamp
def get(self) -> Optional[List[Union[int, float]]]:
if len(self.sorted_list) == 0:
return None
ind = self.ts_ind_map[self.max_timestamp]
return [self.sorted_list[ind][0], self.sorted_list[ind][1]]
def delete(self, from_ts: int, to_ts: int) -> int:
prev_size = len(self.sorted_list)
self.sorted_list = [x for x in self.sorted_list if not (from_ts <= x[0] <= to_ts)]
self.ts_ind_map = {k: v for k, v in self.ts_ind_map.items() if not (from_ts <= k <= to_ts)}
return prev_size - len(self.sorted_list)
def get_rule(self, dest_key: bytes) -> Optional["TimeSeriesRule"]:
for rule in self.rules:
if rule.dest_key.name == dest_key:
return rule
return None
def add_rule(self, rule: "TimeSeriesRule") -> None:
self.rules.append(rule)
def delete_rule(self, rule: "TimeSeriesRule") -> None:
self.rules.remove(rule)
rule.dest_key.source_key = None
def range(
self,
from_ts: int,
to_ts: int,
value_min: Optional[float],
value_max: Optional[float],
count: Optional[int],
filter_ts: Optional[List[int]],
reverse: bool,
) -> List[Tuple[int, float]]:
value_min = value_min or float("-inf")
value_max = value_max or float("inf")
res: List[Tuple[int, float]] = [
x
for x in self.sorted_list
if (from_ts <= x[0] <= to_ts)
and value_min <= x[1] <= value_max
and (filter_ts is None or x[0] in filter_ts)
]
if reverse:
res.reverse()
if count is not None:
return res[:count]
return res
def aggregate(
self,
from_ts: int,
to_ts: int,
latest: bool,
value_min: Optional[float],
value_max: Optional[float],
count: Optional[int],
filter_ts: Optional[List[int]],
align: Optional[int],
aggregator: bytes,
bucket_duration: int,
bucket_timestamp: Optional[bytes],
empty: Optional[bool],
reverse: bool,
) -> List[Tuple[int, float]]:
align = align or 0
value_min = value_min or float("-inf")
value_max = value_max or float("inf")
rule = TimeSeriesRule(self, TimeSeries(b"", self._db), aggregator, bucket_duration)
for x in self.sorted_list:
if from_ts <= x[0] <= to_ts and value_min <= x[1] <= value_max and (filter_ts is None or x[0] in filter_ts):
rule.add_record((x[0], x[1]), bucket_timestamp)
if latest and len(rule.current_bucket) > 0:
rule.apply_curr_bucket(bucket_timestamp)
if empty:
min_bucket_ts = rule.dest_key.sorted_list[0][0]
for ts in range(min_bucket_ts, rule.current_bucket_start_ts, bucket_duration):
if ts not in rule.dest_key.ts_ind_map:
rule.dest_key.add(ts, float("nan"))
rule.dest_key.sorted_list = sorted(rule.dest_key.sorted_list)
if reverse:
rule.dest_key.sorted_list.reverse()
if count:
return rule.dest_key.sorted_list[:count]
return rule.dest_key.sorted_list
class Aggregators:
@staticmethod
def var_p(values: List[float]) -> float:
if len(values) == 0:
return 0
avg = sum(values) / len(values)
return sum((x - avg) ** 2 for x in values) / len(values)
@staticmethod
def var_s(values: List[float]) -> float:
if len(values) == 0:
return 0
avg = sum(values) / len(values)
return sum((x - avg) ** 2 for x in values) / (len(values) - 1)
@staticmethod
def std_p(values: List[float]) -> float:
return Aggregators.var_p(values) ** 0.5
@staticmethod
def std_s(values: List[float]) -> float:
return Aggregators.var_s(values) ** 0.5
AGGREGATORS = {
b"avg": lambda x: sum(x) / len(x),
b"sum": sum,
b"min": min,
b"max": max,
b"range": lambda x: max(x) - min(x),
b"count": len,
b"first": lambda x: x[0],
b"last": lambda x: x[-1],
b"std.p": Aggregators.std_p,
b"std.s": Aggregators.std_s,
b"var.p": Aggregators.var_p,
b"var.s": Aggregators.var_s,
b"twa": lambda x: 0,
}
def apply_aggregator(
bucket: List[Tuple[int, float]], bucket_start_ts: int, bucket_duration: int, aggregator: bytes
) -> float:
if len(bucket) == 0:
return 0.0
if aggregator == b"twa":
total = 0.0
curr_ts = bucket_start_ts
for i, (ts, val) in enumerate(bucket):
# next_ts = bucket[i + 1][0] if len(bucket) > i + 1 else bucket_start_ts + bucket_duration
total += (ts - curr_ts) * val
curr_ts = ts
total += val * (bucket_start_ts + bucket_duration - curr_ts)
return total / bucket_duration
relevant_values = [x[1] for x in bucket]
return AGGREGATORS[aggregator](relevant_values)
class TimeSeriesRule:
def __init__(
self,
source_key: TimeSeries,
dest_key: TimeSeries,
aggregator: bytes,
bucket_duration: int,
align_timestamp: int = 0,
):
self.source_key = source_key
self.dest_key = dest_key
self.aggregator = aggregator.lower()
self.bucket_duration = bucket_duration
self.align_timestamp = align_timestamp
self.current_bucket_start_ts: int = 0
self.current_bucket: List[Tuple[int, float]] = list()
self.dest_key.source_key = source_key.name
def add_record(self, record: Tuple[int, float], bucket_timestamp: Optional[bytes] = None) -> bool:
ts, val = record
bucket_start_ts = ts - (ts % self.bucket_duration) + self.align_timestamp
if self.current_bucket_start_ts == bucket_start_ts:
self.current_bucket.append(record)
if (
self.current_bucket_start_ts != bucket_start_ts
or ts == self.current_bucket_start_ts + self.bucket_duration - 1
):
should_add = self.current_bucket_start_ts != bucket_start_ts
self.apply_curr_bucket(bucket_timestamp)
self.current_bucket_start_ts = (
bucket_start_ts
if self.current_bucket_start_ts != bucket_start_ts
else self.current_bucket_start_ts + self.bucket_duration
)
if should_add:
self.current_bucket.append(record)
return True
return False
def apply_curr_bucket(self, bucket_timestamp: Optional[bytes] = None) -> None:
if len(self.current_bucket) == 0:
return
value = apply_aggregator(
self.current_bucket, self.current_bucket_start_ts, self.bucket_duration, self.aggregator
)
self.current_bucket = list()
timestamp = self.current_bucket_start_ts
if bucket_timestamp == b"+":
timestamp = int(self.current_bucket_start_ts + self.bucket_duration)
elif bucket_timestamp == b"~":
timestamp = int(self.current_bucket_start_ts + self.bucket_duration / 2)
self.dest_key.add(timestamp, value)
|
