ClickHouse/dbms/Storages/StorageLog.h

122 lines
4.0 KiB
C++
Raw Normal View History

2011-10-31 17:55:06 +00:00
#pragma once
2010-03-18 19:32:14 +00:00
2015-01-25 05:07:51 +00:00
#include <map>
#include <shared_mutex>
2017-06-06 17:18:32 +00:00
#include <ext/shared_ptr_helper.h>
#include <Disks/IDisk.h>
#include <Storages/IStorage.h>
#include <Common/FileChecker.h>
#include <Common/escapeForFileName.h>
2010-03-18 19:32:14 +00:00
namespace DB
{
/** Implements simple table engine without support of indices.
2017-04-16 15:00:33 +00:00
* The data is stored in a compressed form.
2010-03-18 19:32:14 +00:00
*/
class StorageLog final : public ext::shared_ptr_helper<StorageLog>, public IStorage
2010-03-18 19:32:14 +00:00
{
friend class LogSource;
friend class LogBlockOutputStream;
friend struct ext::shared_ptr_helper<StorageLog>;
2010-03-18 19:32:14 +00:00
public:
String getName() const override { return "Log"; }
Pipes read(
const Names & column_names,
const SelectQueryInfo & query_info,
const Context & context,
QueryProcessingStage::Enum processed_stage,
size_t max_block_size,
2017-06-02 15:54:39 +00:00
unsigned num_streams) override;
BlockOutputStreamPtr write(const ASTPtr & query, const Context & context) override;
void rename(
const String & new_path_to_table_data,
const String & new_database_name,
const String & new_table_name,
TableStructureWriteLockHolder &) override;
2019-07-03 13:17:19 +00:00
CheckResults checkData(const ASTPtr & /* query */, const Context & /* context */) override;
2019-08-27 20:43:08 +00:00
void truncate(const ASTPtr &, const Context &, TableStructureWriteLockHolder &) override;
2018-04-21 00:35:20 +00:00
Strings getDataPaths() const override { return {DB::fullPath(disk, table_path)}; }
protected:
/** Attach the table with the appropriate name, along the appropriate path (with / at the end),
* (the correctness of names and paths is not verified)
* consisting of the specified columns; Create files if they do not exist.
*/
StorageLog(
DiskPtr disk_,
2019-10-25 19:07:47 +00:00
const std::string & relative_path_,
2019-12-04 16:06:55 +00:00
const StorageID & table_id_,
const ColumnsDescription & columns_,
2019-08-24 21:20:20 +00:00
const ConstraintsDescription & constraints_,
size_t max_compress_block_size_);
private:
2017-11-26 19:22:33 +00:00
/** Offsets to some row number in a file for column in table.
* They are needed so that you can read the data in several threads.
*/
struct Mark
{
size_t rows; /// How many rows are before this offset including the block at this offset.
size_t offset; /// The offset in compressed file.
2017-11-26 19:22:33 +00:00
};
using Marks = std::vector<Mark>;
/// Column data
struct ColumnData
{
/// Specifies the column number in the marks file.
/// Does not necessarily match the column number among the columns of the table: columns with lengths of arrays are also numbered here.
size_t column_index;
String data_file_path;
2017-11-26 19:22:33 +00:00
Marks marks;
};
using Files = std::map<String, ColumnData>; /// file name -> column data
DiskPtr disk;
String table_path;
2017-11-26 19:22:33 +00:00
mutable std::shared_mutex rwlock;
2017-11-26 19:22:33 +00:00
Files files;
Names column_names_by_idx; /// column_index -> name
String marks_file_path;
2017-04-16 15:00:33 +00:00
/// The order of adding files should not change: it corresponds to the order of the columns in the marks file.
void addFiles(const String & column_name, const IDataType & type);
2012-08-29 20:07:24 +00:00
bool loaded_marks = false;
size_t max_compress_block_size;
size_t file_count = 0;
2014-03-28 14:36:24 +00:00
FileChecker file_checker;
/// Read marks files if they are not already read.
/// It is done lazily, so that with a large number of tables, the server starts quickly.
/// You can not call with a write locked `rwlock`.
void loadMarks();
2017-04-16 15:00:33 +00:00
/** For normal columns, the number of rows in the block is specified in the marks.
* For array columns and nested structures, there are more than one group of marks that correspond to different files
2017-11-26 19:22:33 +00:00
* - for elements (file name.bin) - the total number of array elements in the block is specified,
2017-04-16 15:00:33 +00:00
* - for array sizes (file name.size0.bin) - the number of rows (the whole arrays themselves) in the block is specified.
*
2017-04-16 15:00:33 +00:00
* Return the first group of marks that contain the number of rows, but not the internals of the arrays.
*/
const Marks & getMarksWithRealRowCount() const;
2010-03-18 19:32:14 +00:00
};
}