ClickHouse/dbms/include/DB/Storages/MergeTree/MergeTreeBlockInputStream.h

396 lines
14 KiB
C
Raw Normal View History

2013-04-24 10:31:32 +00:00
#pragma once
#include <DB/DataStreams/IProfilingBlockInputStream.h>
2014-03-09 17:36:01 +00:00
#include <DB/Storages/MergeTree/MergeTreeData.h>
#include <DB/Storages/MergeTree/PKCondition.h>
2013-11-26 11:55:11 +00:00
#include <DB/Storages/MergeTree/MergeTreeReader.h>
2013-04-24 10:31:32 +00:00
namespace DB
{
/// Для чтения из одного куска. Для чтения сразу из многих, Storage использует сразу много таких объектов.
class MergeTreeBlockInputStream : public IProfilingBlockInputStream
{
public:
MergeTreeBlockInputStream(const String & path_, /// Путь к куску
2014-07-17 13:41:47 +00:00
size_t block_size_, Names column_names,
2014-03-09 17:36:01 +00:00
MergeTreeData & storage_, const MergeTreeData::DataPartPtr & owned_data_part_,
const MarkRanges & mark_ranges_, bool use_uncompressed_cache_,
2015-04-12 04:39:20 +00:00
ExpressionActionsPtr prewhere_actions_, String prewhere_column_, bool check_columns,
size_t min_bytes_to_use_direct_io_, size_t max_read_buffer_size_)
:
2014-07-17 13:41:47 +00:00
path(path_), block_size(block_size_),
storage(storage_), owned_data_part(owned_data_part_),
2014-07-14 14:07:47 +00:00
part_columns_lock(new Poco::ScopedReadRWLock(owned_data_part->columns_lock)),
2013-11-26 11:55:11 +00:00
all_mark_ranges(mark_ranges_), remaining_mark_ranges(mark_ranges_),
use_uncompressed_cache(use_uncompressed_cache_),
2014-03-13 12:48:07 +00:00
prewhere_actions(prewhere_actions_), prewhere_column(prewhere_column_),
2015-02-16 17:01:38 +00:00
log(&Logger::get("MergeTreeBlockInputStream")),
2015-04-12 04:39:20 +00:00
ordered_names{column_names},
min_bytes_to_use_direct_io(min_bytes_to_use_direct_io_), max_read_buffer_size(max_read_buffer_size_)
2013-04-24 10:31:32 +00:00
{
2013-11-26 11:55:11 +00:00
std::reverse(remaining_mark_ranges.begin(), remaining_mark_ranges.end());
2015-04-01 11:44:42 +00:00
/// inject columns required for defaults evaluation
const auto injected_columns = injectRequiredColumns(column_names);
/// insert injected columns into ordered columns list to avoid exception about different block structures
ordered_names.insert(std::end(ordered_names), std::begin(injected_columns), std::end(injected_columns));
2014-07-17 13:41:47 +00:00
Names pre_column_names;
2014-04-22 22:52:00 +00:00
if (prewhere_actions)
{
pre_column_names = prewhere_actions->getRequiredColumns();
2015-04-01 11:44:42 +00:00
/// @todo somehow decide which injected columns belong to PREWHERE, optimizing reads
pre_column_names.insert(std::end(pre_column_names),
std::begin(injected_columns), std::end(injected_columns));
if (pre_column_names.empty())
pre_column_names.push_back(column_names[0]);
2015-04-01 11:44:42 +00:00
const NameSet pre_name_set(pre_column_names.begin(), pre_column_names.end());
/// Если выражение в PREWHERE - не столбец таблицы, не нужно отдавать наружу столбец с ним
/// (от storage ожидают получить только столбцы таблицы).
remove_prewhere_column = !pre_name_set.count(prewhere_column);
2015-04-01 11:44:42 +00:00
Names post_column_names;
for (const auto & name : column_names)
if (!pre_name_set.count(name))
post_column_names.push_back(name);
2015-04-01 11:44:42 +00:00
column_names = post_column_names;
}
2015-04-01 11:44:42 +00:00
/// will be used to distinguish between PREWHERE and WHERE columns when applying filter
column_name_set = NameSet{column_names.begin(), column_names.end()};
2014-07-17 13:41:47 +00:00
if (check_columns)
{
/// Под owned_data_part->columns_lock проверим, что все запрошенные столбцы в куске того же типа, что в таблице.
/// Это может быть не так во время ALTER MODIFY.
if (!pre_column_names.empty())
storage.check(owned_data_part->columns, pre_column_names);
if (!column_names.empty())
storage.check(owned_data_part->columns, column_names);
pre_columns = storage.getColumnsList().addTypes(pre_column_names);
columns = storage.getColumnsList().addTypes(column_names);
}
else
{
pre_columns = owned_data_part->columns.addTypes(pre_column_names);
columns = owned_data_part->columns.addTypes(column_names);
}
/// Оценим общее количество строк - для прогресс-бара.
size_t total_rows = 0;
for (const auto & range : all_mark_ranges)
total_rows += range.end - range.begin;
total_rows *= storage.index_granularity;
2014-03-13 12:48:07 +00:00
LOG_TRACE(log, "Reading " << all_mark_ranges.size() << " ranges from part " << owned_data_part->name
<< ", approx. " << total_rows
<< (all_mark_ranges.size() > 1
? ", up to " + toString((all_mark_ranges.back().end - all_mark_ranges.front().begin) * storage.index_granularity)
: "")
2013-11-26 11:55:11 +00:00
<< " rows starting from " << all_mark_ranges.front().begin * storage.index_granularity);
setTotalRowsApprox(total_rows);
2013-04-24 10:31:32 +00:00
}
String getName() const override { return "MergeTreeBlockInputStream"; }
String getID() const override
{
std::stringstream res;
res << "MergeTree(" << path << ", columns";
2014-07-17 13:41:47 +00:00
for (const NameAndTypePair & column : columns)
res << ", " << column.name;
if (prewhere_actions)
res << ", prewhere, " << prewhere_actions->getID();
res << ", marks";
2013-11-26 11:55:11 +00:00
for (size_t i = 0; i < all_mark_ranges.size(); ++i)
res << ", " << all_mark_ranges[i].begin << ", " << all_mark_ranges[i].end;
res << ")";
return res.str();
}
2013-04-24 10:31:32 +00:00
protected:
/// Будем вызывать progressImpl самостоятельно.
void progress(const Progress & value) override {}
2015-04-02 03:08:43 +00:00
/** Если некоторых запрошенных столбцов нет в куске,
* то выясняем, какие столбцы может быть необходимо дополнительно прочитать,
* чтобы можно было вычислить DEFAULT выражение для этих столбцов.
2015-04-02 03:17:10 +00:00
* Добавляет их в columns.
2015-04-02 03:08:43 +00:00
*/
2015-04-02 03:17:10 +00:00
NameSet injectRequiredColumns(Names & columns) const
2015-04-02 03:08:43 +00:00
{
2015-04-01 11:44:42 +00:00
NameSet required_columns{std::begin(columns), std::end(columns)};
NameSet injected_columns;
2014-10-24 10:48:47 +00:00
2015-04-01 11:44:42 +00:00
for (size_t i = 0; i < columns.size(); ++i)
2014-10-24 10:48:47 +00:00
{
2015-04-01 11:44:42 +00:00
const auto & column_name = columns[i];
2014-10-24 10:48:47 +00:00
2015-04-01 11:44:42 +00:00
/// column has files and hence does not require evaluation
if (owned_data_part->hasColumnFiles(column_name))
continue;
2014-10-24 10:48:47 +00:00
2015-04-01 11:44:42 +00:00
const auto default_it = storage.column_defaults.find(column_name);
/// columns has no explicit default expression
if (default_it == std::end(storage.column_defaults))
continue;
2015-03-31 12:57:14 +00:00
2015-04-01 11:44:42 +00:00
/// collect identifiers required for evaluation
IdentifierNameSet identifiers;
default_it->second.expression->collectIdentifierNames(identifiers);
for (const auto & identifier : identifiers)
2014-10-24 10:48:47 +00:00
{
2015-04-01 11:44:42 +00:00
if (storage.hasColumn(identifier))
2014-10-24 10:48:47 +00:00
{
2015-04-01 11:44:42 +00:00
/// ensure each column is added only once
if (required_columns.count(identifier) == 0)
2014-10-24 10:48:47 +00:00
{
2015-04-01 11:44:42 +00:00
columns.emplace_back(identifier);
required_columns.emplace(identifier);
injected_columns.emplace(identifier);
2014-10-24 10:48:47 +00:00
}
}
}
}
2015-04-01 11:44:42 +00:00
return injected_columns;
2014-10-24 10:48:47 +00:00
}
2015-04-02 03:08:43 +00:00
Block readImpl() override
2013-04-24 10:31:32 +00:00
{
Block res;
2013-11-26 11:55:11 +00:00
if (remaining_mark_ranges.empty())
return res;
if (!reader)
2013-04-24 10:31:32 +00:00
{
2015-04-16 06:12:35 +00:00
if (use_uncompressed_cache)
owned_uncompressed_cache = storage.context.getUncompressedCache();
owned_mark_cache = storage.context.getMarkCache();
2015-04-02 03:08:43 +00:00
2015-04-12 04:39:20 +00:00
reader.reset(new MergeTreeReader(
2015-04-16 06:12:35 +00:00
path, owned_data_part, columns, owned_uncompressed_cache.get(), owned_mark_cache.get(), storage,
2015-04-12 04:39:20 +00:00
all_mark_ranges, min_bytes_to_use_direct_io, max_read_buffer_size));
2015-04-02 03:08:43 +00:00
if (prewhere_actions)
2015-04-12 04:39:20 +00:00
pre_reader.reset(new MergeTreeReader(
2015-04-16 06:12:35 +00:00
path, owned_data_part, pre_columns, owned_uncompressed_cache.get(), owned_mark_cache.get(), storage,
2015-04-12 04:39:20 +00:00
all_mark_ranges, min_bytes_to_use_direct_io, max_read_buffer_size));
2013-04-24 10:31:32 +00:00
}
2013-11-26 11:55:11 +00:00
if (prewhere_actions)
2013-04-24 10:31:32 +00:00
{
do
{
/// Прочитаем полный блок столбцов, нужных для вычисления выражения в PREWHERE.
size_t space_left = std::max(1LU, block_size / storage.index_granularity);
MarkRanges ranges_to_read;
while (!remaining_mark_ranges.empty() && space_left)
{
MarkRange & range = remaining_mark_ranges.back();
size_t marks_to_read = std::min(range.end - range.begin, space_left);
pre_reader->readRange(range.begin, range.begin + marks_to_read, res);
ranges_to_read.push_back(MarkRange(range.begin, range.begin + marks_to_read));
space_left -= marks_to_read;
range.begin += marks_to_read;
if (range.begin == range.end)
remaining_mark_ranges.pop_back();
}
2015-04-02 03:08:43 +00:00
progressImpl(Progress(res.rowsInFirstColumn(), res.bytes()));
2015-02-16 17:01:38 +00:00
pre_reader->fillMissingColumns(res, ordered_names);
/// Вычислим выражение в PREWHERE.
prewhere_actions->execute(res);
ColumnPtr column = res.getByName(prewhere_column).column;
if (remove_prewhere_column)
res.erase(prewhere_column);
size_t pre_bytes = res.bytes();
/** Если фильтр - константа (например, написано PREWHERE 1),
2015-04-02 03:08:43 +00:00
* то либо вернём пустой блок, либо вернём блок без изменений.
*/
if (ColumnConstUInt8 * column_const = typeid_cast<ColumnConstUInt8 *>(&*column))
{
if (!column_const->getData())
{
res.clear();
return res;
}
2014-04-22 22:52:00 +00:00
for (size_t i = 0; i < ranges_to_read.size(); ++i)
{
const MarkRange & range = ranges_to_read[i];
reader->readRange(range.begin, range.end, res);
}
progressImpl(Progress(0, res.bytes() - pre_bytes));
}
else if (ColumnUInt8 * column_vec = typeid_cast<ColumnUInt8 *>(&*column))
{
size_t index_granularity = storage.index_granularity;
2013-06-25 12:19:10 +00:00
const IColumn::Filter & pre_filter = column_vec->getData();
IColumn::Filter post_filter(pre_filter.size());
2013-11-11 05:35:58 +00:00
/// Прочитаем в нужных отрезках остальные столбцы и составим для них свой фильтр.
size_t pre_filter_pos = 0;
size_t post_filter_pos = 0;
2014-04-22 22:52:00 +00:00
for (size_t i = 0; i < ranges_to_read.size(); ++i)
{
const MarkRange & range = ranges_to_read[i];
size_t begin = range.begin;
size_t pre_filter_begin_pos = pre_filter_pos;
for (size_t mark = range.begin; mark <= range.end; ++mark)
{
UInt8 nonzero = 0;
if (mark != range.end)
{
size_t limit = std::min(pre_filter.size(), pre_filter_pos + index_granularity);
for (size_t row = pre_filter_pos; row < limit; ++row)
nonzero |= pre_filter[row];
}
if (!nonzero)
{
if (mark > begin)
{
memcpy(
&post_filter[post_filter_pos],
&pre_filter[pre_filter_begin_pos],
pre_filter_pos - pre_filter_begin_pos);
post_filter_pos += pre_filter_pos - pre_filter_begin_pos;
reader->readRange(begin, mark, res);
}
begin = mark + 1;
pre_filter_begin_pos = std::min(pre_filter_pos + index_granularity, pre_filter.size());
}
if (mark < range.end)
pre_filter_pos = std::min(pre_filter_pos + index_granularity, pre_filter.size());
}
}
if (!post_filter_pos)
{
res.clear();
continue;
}
progressImpl(Progress(0, res.bytes() - pre_bytes));
post_filter.resize(post_filter_pos);
/// Отфильтруем столбцы, относящиеся к PREWHERE, используя pre_filter,
/// остальные столбцы - используя post_filter.
size_t rows = 0;
for (size_t i = 0; i < res.columns(); ++i)
{
ColumnWithNameAndType & column = res.getByPosition(i);
if (column.name == prewhere_column && res.columns() > 1)
continue;
column.column = column.column->filter(column_name_set.count(column.name) ? post_filter : pre_filter);
rows = column.column->size();
}
/// Заменим столбец со значением условия из PREWHERE на константу.
if (!remove_prewhere_column)
res.getByName(prewhere_column).column = new ColumnConstUInt8(rows, 1);
}
else
throw Exception("Illegal type " + column->getName() + " of column for filter. Must be ColumnUInt8 or ColumnConstUInt8.", ErrorCodes::ILLEGAL_TYPE_OF_COLUMN_FOR_FILTER);
2015-04-02 03:08:43 +00:00
reader->fillMissingColumnsAndReorder(res, ordered_names);
}
while (!remaining_mark_ranges.empty() && !res && !isCancelled());
2013-04-24 10:31:32 +00:00
}
else
{
size_t space_left = std::max(1LU, block_size / storage.index_granularity);
while (!remaining_mark_ranges.empty() && space_left)
{
MarkRange & range = remaining_mark_ranges.back();
size_t marks_to_read = std::min(range.end - range.begin, space_left);
reader->readRange(range.begin, range.begin + marks_to_read, res);
2013-11-11 05:35:58 +00:00
space_left -= marks_to_read;
range.begin += marks_to_read;
if (range.begin == range.end)
remaining_mark_ranges.pop_back();
}
2015-04-02 03:08:43 +00:00
progressImpl(Progress(res.rowsInFirstColumn(), res.bytes()));
2015-02-16 17:01:38 +00:00
reader->fillMissingColumns(res, ordered_names);
}
2013-11-26 11:55:11 +00:00
if (remaining_mark_ranges.empty())
2013-04-24 10:31:32 +00:00
{
/** Закрываем файлы (ещё до уничтожения объекта).
2014-07-23 09:15:41 +00:00
* Чтобы при создании многих источников, но одновременном чтении только из нескольких,
* буферы не висели в памяти.
*/
2014-04-22 22:58:05 +00:00
reader.reset();
2014-07-14 14:07:47 +00:00
pre_reader.reset();
part_columns_lock.reset();
owned_data_part.reset();
2013-04-24 10:31:32 +00:00
}
2013-11-26 11:55:11 +00:00
2013-04-24 10:31:32 +00:00
return res;
}
2013-04-24 10:31:32 +00:00
private:
const String path;
size_t block_size;
2014-07-17 13:41:47 +00:00
NamesAndTypesList columns;
NameSet column_name_set;
2014-07-17 13:41:47 +00:00
NamesAndTypesList pre_columns;
2014-03-09 17:36:01 +00:00
MergeTreeData & storage;
MergeTreeData::DataPartPtr owned_data_part; /// Кусок не будет удалён, пока им владеет этот объект.
2014-07-14 14:07:47 +00:00
std::unique_ptr<Poco::ScopedReadRWLock> part_columns_lock; /// Не дадим изменить список столбцов куска, пока мы из него читаем.
2013-11-26 11:55:11 +00:00
MarkRanges all_mark_ranges; /// В каких диапазонах засечек читать. В порядке возрастания номеров.
MarkRanges remaining_mark_ranges; /// В каких диапазонах засечек еще не прочли.
/// В порядке убывания номеров, чтобы можно было выбрасывать из конца.
bool use_uncompressed_cache;
2014-04-22 22:58:05 +00:00
std::unique_ptr<MergeTreeReader> reader;
std::unique_ptr<MergeTreeReader> pre_reader;
ExpressionActionsPtr prewhere_actions;
String prewhere_column;
bool remove_prewhere_column;
2014-03-13 12:48:07 +00:00
Logger * log;
2015-02-16 17:01:38 +00:00
2015-04-01 11:44:42 +00:00
/// column names in specific order as expected by other stages
Names ordered_names;
2015-04-12 04:39:20 +00:00
size_t min_bytes_to_use_direct_io;
size_t max_read_buffer_size;
2015-04-16 06:12:35 +00:00
UncompressedCachePtr owned_uncompressed_cache;
MarkCachePtr owned_mark_cache;
2013-04-24 10:31:32 +00:00
};
}