thevar1able/ClickHouse
1
0
Fork 0
mirror of https://github.com/ClickHouse/ClickHouse.git synced 2024-11-27 10:02:01 +00:00
Code Issues Packages Projects Releases Wiki Activity

Properly resolve array join columns.

Browse Source
This commit is contained in:
Nikolai Kochetov 2024-05-23 16:34:11 +00:00
parent 7f450cfbdd
commit b4581286f7
5 changed files with 75 additions and 20 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

56
src/Analyzer/Passes/QueryAnalysisPass.cpp
View File

@ -607,6 +607,8 @@ struct ScopeAliases
std::unordered_set<QueryTreeNodePtr> nodes_with_duplicated_aliases;
std::vector<QueryTreeNodePtr> cloned_nodes_with_duplicated_aliases;
std::unordered_set<std::string> array_join_aliases;
std::unordered_map<std::string, QueryTreeNodePtr> & getAliasMap(IdentifierLookupContext lookup_context)
{
switch (lookup_context)
@ -2875,7 +2877,7 @@ QueryTreeNodePtr QueryAnalyzer::tryResolveIdentifierFromExpressionArguments(cons
bool QueryAnalyzer::tryBindIdentifierToAliases(const IdentifierLookup & identifier_lookup, const IdentifierResolveScope & scope)
{
return scope.aliases.find(identifier_lookup, ScopeAliases::FindOption::FIRST_NAME) != nullptr;
return scope.aliases.find(identifier_lookup, ScopeAliases::FindOption::FIRST_NAME) != nullptr || scope.aliases.array_join_aliases.contains(identifier_lookup.identifier.front());
}
/** Resolve identifier from scope aliases.
@ -2924,6 +2926,7 @@ QueryTreeNodePtr QueryAnalyzer::tryResolveIdentifierFromAliases(const Identifier
IdentifierResolveSettings identifier_resolve_settings)
{
const auto & identifier_bind_part = identifier_lookup.identifier.front();
// std::cerr << "tryResolveIdentifierFromAliases " << identifier_lookup.dump() << std::endl;
auto * it = scope.aliases.find(identifier_lookup, ScopeAliases::FindOption::FIRST_NAME);
if (it == nullptr)
@ -2952,6 +2955,7 @@ QueryTreeNodePtr QueryAnalyzer::tryResolveIdentifierFromAliases(const Identifier
}
auto node_type = alias_node->getNodeType();
// std::cerr << "tryResolveIdentifierFromAliases 1.5 \n" << alias_node->dumpTree() << std::endl;
/// Resolve expression if necessary
if (node_type == QueryTreeNodeType::IDENTIFIER)
@ -2960,6 +2964,7 @@ QueryTreeNodePtr QueryAnalyzer::tryResolveIdentifierFromAliases(const Identifier
auto & alias_identifier_node = alias_node->as<IdentifierNode &>();
auto identifier = alias_identifier_node.getIdentifier();
// std::cerr << "tryResolveIdentifierFromAliases 2 " << identifier.getFullName() << std::endl;
auto lookup_result = tryResolveIdentifier(IdentifierLookup{identifier, identifier_lookup.lookup_context}, scope, identifier_resolve_settings);
if (!lookup_result.resolved_identifier)
{
@ -3136,6 +3141,7 @@ QueryTreeNodePtr QueryAnalyzer::tryResolveIdentifierFromStorage(
size_t identifier_column_qualifier_parts,
bool can_be_not_found)
{
// std::cerr << "tryResolveIdentifierFromStorage " << identifier.getFullName() << std::endl;
auto identifier_without_column_qualifier = identifier;
identifier_without_column_qualifier.popFirst(identifier_column_qualifier_parts);
@ -3278,6 +3284,7 @@ QueryTreeNodePtr QueryAnalyzer::tryResolveIdentifierFromStorage(
{
auto qualified_identifier_with_removed_part = qualified_identifier;
qualified_identifier_with_removed_part.popFirst();
// std::cerr << "tryResolveIdentifierFromStorage qualified_identifier_with_removed_part" << qualified_identifier_with_removed_part.getFullName() << std::endl;
if (qualified_identifier_with_removed_part.empty())
break;
@ -3896,7 +3903,7 @@ QueryTreeNodePtr QueryAnalyzer::tryResolveIdentifierFromArrayJoin(const Identifi
auto resolved_identifier = tryResolveIdentifierFromJoinTreeNode(identifier_lookup, from_array_join_node.getTableExpression(), scope);
// std::cerr << "tryResolveIdentifierFromArrayJoin 2 " << scope.table_expressions_in_resolve_process.contains(table_expression_node.get())
// << ' ' << identifier_lookup.dump() << '\n' << table_expression_node->dumpTree() << std::endl;
// << ' ' << identifier_lookup.dump() << ' ' << (resolved_identifier ? resolved_identifier->dumpTree() : "not resolved ") << std::endl;
if (scope.table_expressions_in_resolve_process.contains(table_expression_node.get()) || !identifier_lookup.isExpressionLookup())
return resolved_identifier;
@ -3914,14 +3921,48 @@ QueryTreeNodePtr QueryAnalyzer::tryResolveIdentifierFromArrayJoin(const Identifi
auto & array_join_column_expression_typed = array_join_column_expression->as<ColumnNode &>();
// std::cerr << "========== " << identifier_lookup.identifier.getFullName() << ' ' << from_array_join_node.getAlias() << ' ' << array_join_column_expression_typed.getAlias() << std::endl;
const auto & parts = identifier_lookup.identifier.getParts();
if (array_join_column_expression_typed.getAlias() == identifier_lookup.identifier.getFullName() ||
(parts.size() == 2 && parts.front() == from_array_join_node.getAlias() && parts.back() == array_join_column_expression_typed.getAlias()))
IdentifierView identifier_view(identifier_lookup.identifier);
if (identifier_view.isCompound() && from_array_join_node.hasAlias() && identifier_view.front() == from_array_join_node.getAlias())
identifier_view.popFirst();
const auto & alias_or_name = array_join_column_expression_typed.hasAlias()
? array_join_column_expression_typed.getAlias()
: array_join_column_expression_typed.getColumnName();
if (identifier_view.front() == alias_or_name)
identifier_view.popFirst();
else if (identifier_view.getFullName() == alias_or_name)
identifier_view.popFirst(identifier_view.getPartsSize()); /// Clear
else
continue;
if (identifier_view.empty())
{
auto array_join_column = std::make_shared<ColumnNode>(array_join_column_expression_typed.getColumn(),
array_join_column_expression_typed.getColumnSource());
return array_join_column;
}
auto compound_expr = tryResolveIdentifierFromCompoundExpression(
identifier_lookup.identifier,
identifier_lookup.identifier.getPartsSize() - identifier_view.getPartsSize() /*identifier_bind_size*/,
array_join_column_expression,
{} /* compound_expression_source */,
scope,
true /* can_be_not_found */);
if (compound_expr)
return compound_expr;
// const auto & parts = identifier_lookup.identifier.getParts();
// if (array_join_column_expression_typed.getAlias() == identifier_lookup.identifier.getFullName() ||
// (parts.size() == 2 && parts.front() == from_array_join_node.getAlias() && parts.back() == array_join_column_expression_typed.getAlias()))
// {
// auto array_join_column = std::make_shared<ColumnNode>(array_join_column_expression_typed.getColumn(),
// array_join_column_expression_typed.getColumnSource());
// return array_join_column;
// }
}
if (!resolved_identifier)
@ -3993,7 +4034,7 @@ QueryTreeNodePtr QueryAnalyzer::tryResolveIdentifierFromJoinTree(const Identifie
if (identifier_lookup.isFunctionLookup())
return {};
// std::cerr << "tryResolveIdentifier " << identifier_lookup.identifier.getFullName() << std::endl;
// std::cerr << "tryResolveIdentifierFromJoinTree " << identifier_lookup.identifier.getFullName() << std::endl;
/// Try to resolve identifier from table columns
if (auto resolved_identifier = tryResolveIdentifierFromTableColumns(identifier_lookup, scope))
@ -7613,9 +7654,12 @@ void QueryAnalyzer::resolveArrayJoin(QueryTreeNodePtr & array_join_node, Identif
for (const auto & elem : array_join_nodes)
{
if (elem->hasAlias())
scope.aliases.array_join_aliases.insert(elem->getAlias());
for (auto & child : elem->getChildren())
{
//std::cerr << "<<<<<<<<<< " << child->dumpTree() << std::endl;
if (child)
expressions_visitor.visit(child);
//visit(child);
}

16
tests/queries/0_stateless/02374_analyzer_array_join.reference
View File

@ -45,7 +45,13 @@ SELECT id, value, value_1, value_2 FROM test_table ARRAY JOIN [[1, 2, 3]] AS val
0 Value [1,2,3] 1
0 Value [1,2,3] 2
0 Value [1,2,3] 3
SELECT 1 AS value FROM test_table ARRAY JOIN [1,2,3] AS value; -- { serverError 179 }
SELECT 1 AS value FROM test_table ARRAY JOIN [1,2,3] AS value;
1
2
3
1
2
3
SELECT 'ARRAY JOIN with column';
ARRAY JOIN with column
SELECT id, value, test_table.value_array FROM test_table ARRAY JOIN value_array;
@ -84,7 +90,13 @@ SELECT id, value, value_array AS value_array_array_alias FROM test_table ARRAY J
0 Value [4,5,6]
SELECT '--';
--
SELECT id AS value FROM test_table ARRAY JOIN value_array AS value; -- { serverError 179 }
SELECT id AS value FROM test_table ARRAY JOIN value_array AS value;
1
2
3
4
5
6
SELECT '--';
--
SELECT id, value, value_array AS value_array_array_alias, value_array_array_alias_element FROM test_table ARRAY JOIN value_array_array_alias AS value_array_array_alias_element;

4
tests/queries/0_stateless/02374_analyzer_array_join.sql
View File

@ -33,7 +33,7 @@ SELECT '--';
SELECT id, value, value_1, value_2 FROM test_table ARRAY JOIN [[1, 2, 3]] AS value_1 ARRAY JOIN value_1 AS value_2;
SELECT 1 AS value FROM test_table ARRAY JOIN [1,2,3] AS value; -- { serverError 179 }
SELECT 1 AS value FROM test_table ARRAY JOIN [1,2,3] AS value;
SELECT 'ARRAY JOIN with column';
@ -53,7 +53,7 @@ SELECT id, value, value_array AS value_array_array_alias FROM test_table ARRAY J
SELECT '--';
SELECT id AS value FROM test_table ARRAY JOIN value_array AS value; -- { serverError 179 }
SELECT id AS value FROM test_table ARRAY JOIN value_array AS value;
SELECT '--';

9
tests/queries/0_stateless/02521_analyzer_array_join_crash.reference
View File

@ -1,11 +1,10 @@
-- { echoOn }
SELECT id, value_element, value FROM test_table ARRAY JOIN [[1,2,3]] AS value_element, value_element AS value;
0 [1,2,3] [1,2,3]
SELECT id, value_element, value FROM test_table ARRAY JOIN [[1,2,3]] AS value_element, value_element AS value; -- { serverError UNKNOWN_IDENTIFIER }
SELECT id, value_element, value FROM test_table ARRAY JOIN [[1,2,3]] AS value_element ARRAY JOIN value_element AS value;
0 [1,2,3] 1
0 [1,2,3] 2
0 [1,2,3] 3
SELECT value_element, value FROM test_table ARRAY JOIN [1048577] AS value_element, arrayMap(x -> value_element, ['']) AS value;
1048577 [1048577]
SELECT arrayFilter(x -> notEmpty(concat(x)), [NULL, NULL]) FROM system.one ARRAY JOIN [1048577] AS elem, arrayMap(x -> splitByChar(x, elem), ['']) AS unused; -- { serverError 44 }
SELECT value_element, value FROM test_table ARRAY JOIN [1048577] AS value_element ARRAY JOIN arrayMap(x -> value_element, ['']) AS value;
1048577 1048577
SELECT arrayFilter(x -> notEmpty(concat(x)), [NULL, NULL]) FROM system.one ARRAY JOIN [1048577] AS elem ARRAY JOIN arrayMap(x -> splitByChar(x, elem), ['']) AS unused; -- { serverError ILLEGAL_COLUMN }

6
tests/queries/0_stateless/02521_analyzer_array_join_crash.sql
View File

@ -11,13 +11,13 @@ INSERT INTO test_table VALUES (0, 'Value');
-- { echoOn }
SELECT id, value_element, value FROM test_table ARRAY JOIN [[1,2,3]] AS value_element, value_element AS value;
SELECT id, value_element, value FROM test_table ARRAY JOIN [[1,2,3]] AS value_element, value_element AS value; -- { serverError UNKNOWN_IDENTIFIER }
SELECT id, value_element, value FROM test_table ARRAY JOIN [[1,2,3]] AS value_element ARRAY JOIN value_element AS value;
SELECT value_element, value FROM test_table ARRAY JOIN [1048577] AS value_element, arrayMap(x -> value_element, ['']) AS value;
SELECT value_element, value FROM test_table ARRAY JOIN [1048577] AS value_element ARRAY JOIN arrayMap(x -> value_element, ['']) AS value;
SELECT arrayFilter(x -> notEmpty(concat(x)), [NULL, NULL]) FROM system.one ARRAY JOIN [1048577] AS elem, arrayMap(x -> splitByChar(x, elem), ['']) AS unused; -- { serverError 44 }
SELECT arrayFilter(x -> notEmpty(concat(x)), [NULL, NULL]) FROM system.one ARRAY JOIN [1048577] AS elem ARRAY JOIN arrayMap(x -> splitByChar(x, elem), ['']) AS unused; -- { serverError ILLEGAL_COLUMN }
-- { echoOff }