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
|
#include "data_statistics.h"
#include <yt/yt/client/chunk_client/data_statistics.h>
#include <yt/yt/core/ytree/fluent.h>
#include <yt/yt/core/misc/protobuf_helpers.h>
namespace NYT::NChunkClient {
using namespace NCompression;
using namespace NYTree;
using namespace NYson;
using ::ToString;
using ::FromString;
////////////////////////////////////////////////////////////////////////////////
namespace NProto {
////////////////////////////////////////////////////////////////////////////////
bool HasInvalidDataWeight(const TDataStatistics& statistics)
{
return statistics.has_data_weight() && statistics.data_weight() == -1;
}
bool HasInvalidUnmergedRowCount(const TDataStatistics& statistics)
{
return statistics.has_unmerged_row_count() && statistics.unmerged_row_count() == -1;
}
bool HasInvalidUnmergedDataWeight(const TDataStatistics& statistics)
{
return statistics.has_unmerged_data_weight() && statistics.unmerged_data_weight() == -1;
}
TDataStatistics& operator += (TDataStatistics& lhs, const TDataStatistics& rhs)
{
lhs.set_uncompressed_data_size(lhs.uncompressed_data_size() + rhs.uncompressed_data_size());
lhs.set_compressed_data_size(lhs.compressed_data_size() + rhs.compressed_data_size());
lhs.set_chunk_count(lhs.chunk_count() + rhs.chunk_count());
lhs.set_row_count(lhs.row_count() + rhs.row_count());
lhs.set_regular_disk_space(lhs.regular_disk_space() + rhs.regular_disk_space());
lhs.set_erasure_disk_space(lhs.erasure_disk_space() + rhs.erasure_disk_space());
if (HasInvalidDataWeight(lhs) || HasInvalidDataWeight(rhs)) {
lhs.set_data_weight(-1);
} else {
lhs.set_data_weight(lhs.data_weight() + rhs.data_weight());
}
if (HasInvalidUnmergedRowCount(lhs) || HasInvalidUnmergedRowCount(rhs)) {
lhs.set_unmerged_row_count(-1);
} else {
lhs.set_unmerged_row_count(lhs.unmerged_row_count() + rhs.unmerged_row_count());
}
if (HasInvalidUnmergedDataWeight(lhs) || HasInvalidUnmergedDataWeight(rhs)) {
lhs.set_unmerged_data_weight(-1);
} else {
lhs.set_unmerged_data_weight(lhs.unmerged_data_weight() + rhs.unmerged_data_weight());
}
return lhs;
}
TDataStatistics operator + (const TDataStatistics& lhs, const TDataStatistics& rhs)
{
auto result = lhs;
result += rhs;
return result;
}
bool operator == (const TDataStatistics& lhs, const TDataStatistics& rhs)
{
return
lhs.uncompressed_data_size() == rhs.uncompressed_data_size() &&
lhs.compressed_data_size() == rhs.compressed_data_size() &&
lhs.row_count() == rhs.row_count() &&
lhs.chunk_count() == rhs.chunk_count() &&
lhs.regular_disk_space() == rhs.regular_disk_space() &&
lhs.erasure_disk_space() == rhs.erasure_disk_space() &&
(HasInvalidDataWeight(lhs) || HasInvalidDataWeight(rhs) || lhs.data_weight() == rhs.data_weight()) &&
(
HasInvalidUnmergedRowCount(lhs) ||
HasInvalidUnmergedRowCount(rhs) ||
lhs.unmerged_row_count() == rhs.unmerged_row_count()) &&
(
HasInvalidUnmergedDataWeight(lhs) ||
HasInvalidUnmergedDataWeight(rhs) ||
lhs.unmerged_data_weight() == rhs.unmerged_data_weight());
}
bool operator != (const TDataStatistics& lhs, const TDataStatistics& rhs)
{
return !(lhs == rhs);
}
void Serialize(const TDataStatistics& statistics, NYson::IYsonConsumer* consumer)
{
// TODO(max42): replace all Item with OptionalItem in order to expose only meaningful
// fields in each particular context. This would require fixing some tests or using
// manually constructed data statistics with some fields being explicitly set to zero
// as a neutral element.
BuildYsonFluently(consumer).BeginMap()
.Item("chunk_count").Value(statistics.chunk_count())
.Item("row_count").Value(statistics.row_count())
.Item("uncompressed_data_size").Value(statistics.uncompressed_data_size())
.Item("compressed_data_size").Value(statistics.compressed_data_size())
.Item("data_weight").Value(statistics.data_weight())
.Item("regular_disk_space").Value(statistics.regular_disk_space())
.Item("erasure_disk_space").Value(statistics.erasure_disk_space())
.Item("unmerged_row_count").Value(statistics.unmerged_row_count())
.Item("unmerged_data_weight").Value(statistics.unmerged_data_weight())
.EndMap();
}
void SetDataStatisticsField(TDataStatistics& statistics, TStringBuf key, i64 value)
{
if (key == "chunk_count") {
statistics.set_chunk_count(value);
} else if (key == "row_count") {
statistics.set_row_count(value);
} else if (key == "uncompressed_data_size") {
statistics.set_uncompressed_data_size(value);
} else if (key == "compressed_data_size") {
statistics.set_compressed_data_size(value);
} else if (key == "data_weight") {
statistics.set_data_weight(value);
} else if (key == "regular_disk_space") {
statistics.set_regular_disk_space(value);
} else if (key == "erasure_disk_space") {
statistics.set_erasure_disk_space(value);
} else if (key == "unmerged_row_count") {
statistics.set_unmerged_row_count(value);
} else if (key == "unmerged_data_weight") {
statistics.set_unmerged_data_weight(value);
} // Else we have a strange situation on our hands but we intentionally ignore it.
}
void FormatValue(TStringBuilderBase* builder, const TDataStatistics& statistics, TStringBuf /*spec*/)
{
builder->AppendFormat(
"{UncompressedDataSize: %v, CompressedDataSize: %v, DataWeight: %v, RowCount: %v, "
"ChunkCount: %v, RegularDiskSpace: %v, ErasureDiskSpace: %v, "
"UnmergedRowCount: %v, UnmergedDataWeight: %v}",
statistics.uncompressed_data_size(),
statistics.compressed_data_size(),
statistics.data_weight(),
statistics.row_count(),
statistics.chunk_count(),
statistics.regular_disk_space(),
statistics.erasure_disk_space(),
statistics.unmerged_row_count(),
statistics.unmerged_data_weight());
}
void FormatValue(TStringBuilderBase* builder, const TDataStatistics* statistics, TStringBuf spec)
{
if (statistics) {
FormatValue(builder, *statistics, spec);
} else {
FormatValue(builder, std::nullopt, spec);
}
}
TString ToString(const TDataStatistics& statistics)
{
return ToStringViaBuilder(statistics);
}
////////////////////////////////////////////////////////////////////////////////
} // namespace NProto
////////////////////////////////////////////////////////////////////////////////
TCodecStatistics& TCodecStatistics::Append(const TCodecDuration& codecTime)
{
return Append(std::make_pair(codecTime.Codec, codecTime.CpuDuration));
}
TCodecStatistics& TCodecStatistics::Append(const std::pair<ECodec, TDuration>& codecTime)
{
CodecToDuration_[codecTime.first] += codecTime.second;
TotalDuration_ += codecTime.second;
return *this;
}
TCodecStatistics& TCodecStatistics::operator+=(const TCodecStatistics& other)
{
for (const auto& pair : other.CodecToDuration_) {
Append(pair);
}
return *this;
}
TDuration TCodecStatistics::GetTotalDuration() const
{
return TotalDuration_;
}
void FormatValue(TStringBuilderBase* builder, const TCodecStatistics& statistics, TStringBuf /* spec */)
{
FormatKeyValueRange(builder, statistics.CodecToDuration(), TDefaultFormatter());
}
TString ToString(const TCodecStatistics& statistics)
{
return ToStringViaBuilder(statistics);
}
////////////////////////////////////////////////////////////////////////////////
} // namespace NYT::NChunkClient
|
