ClickHouse/src/Interpreters/TableJoin.cpp

242 lines
7.3 KiB
C++
Raw Normal View History

#include <Interpreters/TableJoin.h>
2018-11-02 18:53:23 +00:00
#include <Parsers/ASTExpressionList.h>
2019-09-09 19:43:37 +00:00
#include <Core/Settings.h>
#include <Core/Block.h>
2018-11-02 18:53:23 +00:00
#include <DataTypes/DataTypeNullable.h>
2018-11-02 18:53:23 +00:00
namespace DB
{
2019-09-09 19:43:37 +00:00
namespace ErrorCodes
{
}
TableJoin::TableJoin(const Settings & settings, VolumePtr tmp_volume_)
2019-09-30 14:36:45 +00:00
: size_limits(SizeLimits{settings.max_rows_in_join, settings.max_bytes_in_join, settings.join_overflow_mode})
, default_max_bytes(settings.default_max_bytes_in_join)
2019-09-09 19:43:37 +00:00
, join_use_nulls(settings.join_use_nulls)
2020-01-17 11:45:43 +00:00
, max_joined_block_rows(settings.max_joined_block_size_rows)
, join_algorithm(settings.join_algorithm)
, partial_merge_join_optimizations(settings.partial_merge_join_optimizations)
2019-09-23 19:36:47 +00:00
, partial_merge_join_rows_in_right_blocks(settings.partial_merge_join_rows_in_right_blocks)
, tmp_volume(tmp_volume_)
{
if (settings.partial_merge_join)
join_algorithm = JoinAlgorithm::PREFER_PARTIAL_MERGE;
}
2019-09-09 19:43:37 +00:00
void TableJoin::addUsingKey(const ASTPtr & ast)
{
key_names_left.push_back(ast->getColumnName());
key_names_right.push_back(ast->getAliasOrColumnName());
key_asts_left.push_back(ast);
key_asts_right.push_back(ast);
2019-07-30 18:39:37 +00:00
auto & right_key = key_names_right.back();
if (renames.count(right_key))
right_key = renames[right_key];
}
void TableJoin::addOnKeys(ASTPtr & left_table_ast, ASTPtr & right_table_ast)
{
key_names_left.push_back(left_table_ast->getColumnName());
key_names_right.push_back(right_table_ast->getAliasOrColumnName());
key_asts_left.push_back(left_table_ast);
key_asts_right.push_back(right_table_ast);
}
/// @return how many times right key appears in ON section.
size_t TableJoin::rightKeyInclusion(const String & name) const
{
if (hasUsing())
return 0;
size_t count = 0;
for (const auto & key_name : key_names_right)
if (name == key_name)
++count;
return count;
}
void TableJoin::deduplicateAndQualifyColumnNames(const NameSet & left_table_columns, const String & right_table_prefix)
2019-07-30 18:39:37 +00:00
{
NameSet joined_columns;
NamesAndTypesList dedup_columns;
for (auto & column : columns_from_joined_table)
{
if (joined_columns.count(column.name))
continue;
joined_columns.insert(column.name);
dedup_columns.push_back(column);
auto & inserted = dedup_columns.back();
if (left_table_columns.count(column.name))
inserted.name = right_table_prefix + column.name;
original_names[inserted.name] = column.name;
if (inserted.name != column.name)
renames[column.name] = inserted.name;
}
columns_from_joined_table.swap(dedup_columns);
}
NameSet TableJoin::getQualifiedColumnsSet() const
{
2019-07-30 18:39:37 +00:00
NameSet out;
for (const auto & names : original_names)
out.insert(names.first);
return out;
2018-11-02 18:53:23 +00:00
}
NamesWithAliases TableJoin::getNamesWithAliases(const NameSet & required_columns) const
2019-07-30 18:39:37 +00:00
{
NamesWithAliases out;
2019-07-30 18:39:37 +00:00
for (const auto & column : required_columns)
2019-02-13 19:00:52 +00:00
{
2019-07-30 18:39:37 +00:00
auto it = original_names.find(column);
if (it != original_names.end())
out.emplace_back(it->second, it->first); /// {original_name, name}
2019-02-13 19:00:52 +00:00
}
2019-07-30 18:39:37 +00:00
return out;
2018-11-02 18:53:23 +00:00
}
ASTPtr TableJoin::leftKeysList() const
{
ASTPtr keys_list = std::make_shared<ASTExpressionList>();
keys_list->children = key_asts_left;
return keys_list;
}
ASTPtr TableJoin::rightKeysList() const
{
ASTPtr keys_list = std::make_shared<ASTExpressionList>();
if (hasOn())
keys_list->children = key_asts_right;
return keys_list;
}
Names TableJoin::requiredJoinedNames() const
{
NameSet required_columns_set(key_names_right.begin(), key_names_right.end());
for (const auto & joined_column : columns_added_by_join)
required_columns_set.insert(joined_column.name);
return Names(required_columns_set.begin(), required_columns_set.end());
}
NameSet TableJoin::requiredRightKeys() const
2019-09-10 18:39:10 +00:00
{
2019-09-11 15:57:09 +00:00
NameSet required;
2019-09-10 18:39:10 +00:00
for (const auto & name : key_names_right)
2019-09-11 15:57:09 +00:00
for (const auto & column : columns_added_by_join)
if (name == column.name)
required.insert(name);
2019-09-10 18:39:10 +00:00
return required;
}
NamesWithAliases TableJoin::getRequiredColumns(const Block & sample, const Names & action_required_columns) const
{
NameSet required_columns(action_required_columns.begin(), action_required_columns.end());
for (auto & column : requiredJoinedNames())
if (!sample.has(column))
required_columns.insert(column);
return getNamesWithAliases(required_columns);
}
void TableJoin::addJoinedColumn(const NameAndTypePair & joined_column)
{
if (join_use_nulls && isLeftOrFull(table_join.kind))
{
auto type = joined_column.type->canBeInsideNullable() ? makeNullable(joined_column.type) : joined_column.type;
columns_added_by_join.emplace_back(NameAndTypePair(joined_column.name, std::move(type)));
}
else
columns_added_by_join.push_back(joined_column);
}
void TableJoin::addJoinedColumnsAndCorrectNullability(Block & sample_block) const
{
bool right_or_full_join = isRightOrFull(table_join.kind);
bool left_or_full_join = isLeftOrFull(table_join.kind);
for (auto & col : sample_block)
{
/// Materialize column.
/// Column is not empty if it is constant, but after Join all constants will be materialized.
/// So, we need remove constants from header.
if (col.column)
col.column = nullptr;
bool make_nullable = join_use_nulls && right_or_full_join;
if (make_nullable && col.type->canBeInsideNullable())
col.type = makeNullable(col.type);
}
for (const auto & col : columns_added_by_join)
{
auto res_type = col.type;
bool make_nullable = join_use_nulls && left_or_full_join;
if (make_nullable && res_type->canBeInsideNullable())
res_type = makeNullable(res_type);
sample_block.insert(ColumnWithTypeAndName(nullptr, res_type, col.name));
}
}
bool TableJoin::sameJoin(const TableJoin * x, const TableJoin * y)
{
if (!x && !y)
return true;
if (!x || !y)
return false;
return x->table_join.kind == y->table_join.kind
&& x->table_join.strictness == y->table_join.strictness
&& x->key_names_left == y->key_names_left
&& x->key_names_right == y->key_names_right
&& x->columns_added_by_join == y->columns_added_by_join;
}
bool TableJoin::sameStrictnessAndKind(ASTTableJoin::Strictness strictness_, ASTTableJoin::Kind kind_) const
2020-02-02 16:12:48 +00:00
{
if (strictness_ == strictness() && kind_ == kind())
return true;
/// Compatibility: old ANY INNER == new SEMI LEFT
if (strictness_ == ASTTableJoin::Strictness::Semi && isLeft(kind_) &&
strictness() == ASTTableJoin::Strictness::RightAny && isInner(kind()))
return true;
if (strictness() == ASTTableJoin::Strictness::Semi && isLeft(kind()) &&
strictness_ == ASTTableJoin::Strictness::RightAny && isInner(kind_))
return true;
return false;
}
bool TableJoin::allowMergeJoin() const
{
2020-02-11 18:27:52 +00:00
bool is_any = (strictness() == ASTTableJoin::Strictness::Any);
bool is_all = (strictness() == ASTTableJoin::Strictness::All);
bool is_semi = (strictness() == ASTTableJoin::Strictness::Semi);
2020-02-11 18:27:52 +00:00
bool allow_merge_join = (isLeft(kind()) && (is_any || is_all || is_semi)) || (isInner(kind()) && is_all);
return allow_merge_join;
}
2018-11-02 18:53:23 +00:00
}