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
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
|
#include <Core/Defines.h>
#include <Core/ProtocolDefines.h>
#include <IO/ReadHelpers.h>
#include <IO/VarInt.h>
#include <Compression/CompressedReadBufferFromFile.h>
#include <DataTypes/DataTypeFactory.h>
#include <Common/typeid_cast.h>
#include <base/range.h>
#include <Formats/NativeReader.h>
#include <Formats/insertNullAsDefaultIfNeeded.h>
#include <DataTypes/DataTypeLowCardinality.h>
#include <DataTypes/Serializations/SerializationInfo.h>
#include <DataTypes/DataTypeAggregateFunction.h>
#include <Interpreters/castColumn.h>
namespace DB
{
namespace ErrorCodes
{
extern const int INCORRECT_INDEX;
extern const int LOGICAL_ERROR;
extern const int CANNOT_READ_ALL_DATA;
extern const int INCORRECT_DATA;
extern const int TOO_LARGE_ARRAY_SIZE;
}
NativeReader::NativeReader(ReadBuffer & istr_, UInt64 server_revision_)
: istr(istr_), server_revision(server_revision_)
{
}
NativeReader::NativeReader(
ReadBuffer & istr_,
const Block & header_,
UInt64 server_revision_,
bool skip_unknown_columns_,
bool null_as_default_,
bool allow_types_conversion_,
BlockMissingValues * block_missing_values_)
: istr(istr_)
, header(header_)
, server_revision(server_revision_)
, skip_unknown_columns(skip_unknown_columns_)
, null_as_default(null_as_default_)
, allow_types_conversion(allow_types_conversion_)
, block_missing_values(block_missing_values_)
{
}
NativeReader::NativeReader(ReadBuffer & istr_, UInt64 server_revision_,
IndexForNativeFormat::Blocks::const_iterator index_block_it_,
IndexForNativeFormat::Blocks::const_iterator index_block_end_)
: istr(istr_), server_revision(server_revision_),
use_index(true), index_block_it(index_block_it_), index_block_end(index_block_end_)
{
istr_concrete = typeid_cast<CompressedReadBufferFromFile *>(&istr);
if (!istr_concrete)
throw Exception(ErrorCodes::LOGICAL_ERROR, "When need to use index for NativeReader, istr must be CompressedReadBufferFromFile.");
if (index_block_it == index_block_end)
return;
index_column_it = index_block_it->columns.begin();
/// Initialize header from the index.
for (const auto & column : index_block_it->columns)
{
auto type = DataTypeFactory::instance().get(column.type);
header.insert(ColumnWithTypeAndName{ type, column.name });
}
}
void NativeReader::resetParser()
{
istr_concrete = nullptr;
use_index = false;
}
void NativeReader::readData(const ISerialization & serialization, ColumnPtr & column, ReadBuffer & istr, size_t rows, double avg_value_size_hint)
{
ISerialization::DeserializeBinaryBulkSettings settings;
settings.getter = [&](ISerialization::SubstreamPath) -> ReadBuffer * { return &istr; };
settings.avg_value_size_hint = avg_value_size_hint;
settings.position_independent_encoding = false;
settings.native_format = true;
ISerialization::DeserializeBinaryBulkStatePtr state;
serialization.deserializeBinaryBulkStatePrefix(settings, state);
serialization.deserializeBinaryBulkWithMultipleStreams(column, rows, settings, state, nullptr);
if (column->size() != rows)
throw Exception(ErrorCodes::CANNOT_READ_ALL_DATA,
"Cannot read all data in NativeReader. Rows read: {}. Rows expected: {}", column->size(), rows);
}
Block NativeReader::getHeader() const
{
return header;
}
Block NativeReader::read()
{
Block res;
const DataTypeFactory & data_type_factory = DataTypeFactory::instance();
if (use_index && index_block_it == index_block_end)
return res;
if (istr.eof())
{
if (use_index)
throw ParsingException(ErrorCodes::CANNOT_READ_ALL_DATA, "Input doesn't contain all data for index.");
return res;
}
/// Additional information about the block.
if (server_revision > 0)
res.info.read(istr);
/// Dimensions
size_t columns = 0;
size_t rows = 0;
if (!use_index)
{
readVarUInt(columns, istr);
readVarUInt(rows, istr);
if (columns > 1'000'000uz)
throw Exception(ErrorCodes::TOO_LARGE_ARRAY_SIZE, "Suspiciously many columns in Native format: {}", columns);
if (rows > 1'000'000'000'000uz)
throw Exception(ErrorCodes::TOO_LARGE_ARRAY_SIZE, "Suspiciously many rows in Native format: {}", rows);
}
else
{
columns = index_block_it->num_columns;
rows = index_block_it->num_rows;
}
if (columns == 0 && !header && rows != 0)
throw Exception(ErrorCodes::INCORRECT_DATA, "Zero columns but {} rows in Native format.", rows);
for (size_t i = 0; i < columns; ++i)
{
if (use_index)
{
/// If the current position is what is required, the real seek does not occur.
istr_concrete->seek(index_column_it->location.offset_in_compressed_file, index_column_it->location.offset_in_decompressed_block);
}
ColumnWithTypeAndName column;
/// Name
readBinary(column.name, istr);
/// Type
String type_name;
readBinary(type_name, istr);
column.type = data_type_factory.get(type_name);
setVersionToAggregateFunctions(column.type, true, server_revision);
SerializationPtr serialization;
if (server_revision >= DBMS_MIN_REVISION_WITH_CUSTOM_SERIALIZATION)
{
auto info = column.type->createSerializationInfo({});
UInt8 has_custom;
readBinary(has_custom, istr);
if (has_custom)
info->deserializeFromKindsBinary(istr);
serialization = column.type->getSerialization(*info);
}
else
{
serialization = column.type->getDefaultSerialization();
}
if (use_index)
{
/// Index allows to do more checks.
if (index_column_it->name != column.name)
throw Exception(ErrorCodes::INCORRECT_INDEX, "Index points to column with wrong name: corrupted index or data");
if (index_column_it->type != type_name)
throw Exception(ErrorCodes::INCORRECT_INDEX, "Index points to column with wrong type: corrupted index or data");
}
/// Data
ColumnPtr read_column = column.type->createColumn(*serialization);
double avg_value_size_hint = avg_value_size_hints.empty() ? 0 : avg_value_size_hints[i];
if (rows) /// If no rows, nothing to read.
readData(*serialization, read_column, istr, rows, avg_value_size_hint);
column.column = std::move(read_column);
bool use_in_result = true;
if (header)
{
if (header.has(column.name))
{
auto & header_column = header.getByName(column.name);
if (null_as_default)
insertNullAsDefaultIfNeeded(column, header_column, header.getPositionByName(column.name), block_missing_values);
if (!header_column.type->equals(*column.type))
{
if (allow_types_conversion)
{
try
{
column.column = castColumn(column, header_column.type);
}
catch (Exception & e)
{
e.addMessage(fmt::format(
"while converting column \"{}\" from type {} to type {}",
column.name,
column.type->getName(),
header_column.type->getName()));
throw;
}
}
else
{
/// Support insert from old clients without low cardinality type.
column.column = recursiveLowCardinalityTypeConversion(column.column, column.type, header_column.type);
}
column.type = header_column.type;
}
}
else
{
if (!skip_unknown_columns)
throw Exception(ErrorCodes::INCORRECT_DATA, "Unknown column with name {} found while reading data in Native format", column.name);
use_in_result = false;
}
}
if (use_in_result)
res.insert(std::move(column));
if (use_index)
++index_column_it;
}
if (use_index)
{
if (index_column_it != index_block_it->columns.end())
throw Exception(ErrorCodes::INCORRECT_INDEX, "Inconsistent index: not all columns were read");
++index_block_it;
if (index_block_it != index_block_end)
index_column_it = index_block_it->columns.begin();
}
if (rows && header)
{
/// Allow to skip columns. Fill them with default values.
Block tmp_res;
for (size_t column_i = 0; column_i != header.columns(); ++column_i)
{
auto & col = header.getByPosition(column_i);
if (res.has(col.name))
{
tmp_res.insert(res.getByName(col.name));
}
else
{
tmp_res.insert({col.type->createColumn()->cloneResized(rows), col.type, col.name});
if (block_missing_values)
block_missing_values->setBits(column_i, rows);
}
}
tmp_res.info = res.info;
res.swap(tmp_res);
}
if (res.rows() != rows)
throw Exception(ErrorCodes::LOGICAL_ERROR, "Row count mismatch after desirialization, got: {}, expected: {}", res.rows(), rows);
return res;
}
void NativeReader::updateAvgValueSizeHints(const Block & block)
{
auto rows = block.rows();
if (rows < 10)
return;
avg_value_size_hints.resize_fill(block.columns(), 0);
for (auto idx : collections::range(0, block.columns()))
{
auto & avg_value_size_hint = avg_value_size_hints[idx];
IDataType::updateAvgValueSizeHint(*block.getByPosition(idx).column, avg_value_size_hint);
}
}
}
|