Fix variable types collected and graph crash (#4724)

* Fix variable types collected and graph crash

add test cases

small fix

* unskip some test cases related to multiple query parts

* small delete

* fmt

* Fix ldbc BI R5 implementation

Co-authored-by: Harris.Chu <1726587+HarrisChu@users.noreply.github.com>
Co-authored-by: Sophie <84560950+Sophie-Xie@users.noreply.github.com>

src/graph/planner/match/MatchPlanner.cpp

@@ -78,17 +78,19 @@ Status MatchPlanner::connectMatchPlan(SubPlan& queryPlan, MatchClauseContext* ma
  if (matchCtx->isOptional) {
  // connect LeftJoin match filter
  auto& whereCtx = matchCtx->where;
- if (whereCtx.get() != nullptr) {
- auto exprs =
- ExpressionUtils::collectAll(whereCtx->filter, {Expression::Kind::kLabelTagProperty});
+ if (whereCtx.get() != nullptr && whereCtx->filter != nullptr) {
+ auto exprs = ExpressionUtils::collectAll(
+ whereCtx->filter, {Expression::Kind::kVarProperty, Expression::Kind::kLabel});
 
  // Check if all aliases in where clause are generated by the current match statement pattern
  std::vector<std::string> aliases;
  for (const auto* expr : exprs) {
- DCHECK_EQ(expr->kind(), Expression::Kind::kLabelTagProperty);
- auto* labelExpr = static_cast<const LabelTagPropertyExpression*>(expr)->label();
- DCHECK_EQ(labelExpr->kind(), Expression::Kind::kVarProperty);
- aliases.emplace_back(static_cast<const PropertyExpression*>(labelExpr)->prop());
+ if (expr->kind() == Expression::Kind::kVarProperty) {
+ aliases.emplace_back(static_cast<const PropertyExpression*>(expr)->prop());
+ } else {
+ DCHECK_EQ(expr->kind(), Expression::Kind::kLabel);
+ aliases.emplace_back(static_cast<const LabelExpression*>(expr)->name());
+ }
  }
  auto aliasesGenerated = matchCtx->aliasesGenerated;
  if (!std::all_of(aliases.begin(), aliases.end(), [&aliasesGenerated](std::string& alias) {

tests/tck/features/match/MultiLineMultiQueryParts.feature

@@ -172,6 +172,37 @@ Feature: Multi Line Multi Query Parts
  | "Tim Duncan" | "Manu Ginobili" | NULL |
  | "Tim Duncan" | "Manu Ginobili" | NULL |
  | "Tim Duncan" | "Manu Ginobili" | NULL |
+ When executing query:
+ """
+ MATCH (m)-[]->(n) WHERE id(m)=="Tim Duncan"
+ OPTIONAL MATCH (n)-[]->(l) WHERE id(n)=="Tony Parker"
+ RETURN id(m) AS m, id(n) AS n, id(l) AS l;
+ """
+ Then the result should be, in any order:
+ | m | n | l |
+ | "Tim Duncan" | "Spurs" | NULL |
+ | "Tim Duncan" | "LaMarcus Aldridge" | NULL |
+ | "Tim Duncan" | "Tony Parker" | "Spurs" |
+ | "Tim Duncan" | "Tony Parker" | "LaMarcus Aldridge" |
+ | "Tim Duncan" | "Tony Parker" | "LaMarcus Aldridge" |
+ | "Tim Duncan" | "Tony Parker" | "Kyle Anderson" |
+ | "Tim Duncan" | "Tony Parker" | "Tim Duncan" |
+ | "Tim Duncan" | "Tony Parker" | "Tim Duncan" |
+ | "Tim Duncan" | "Tony Parker" | "Manu Ginobili" |
+ | "Tim Duncan" | "Tony Parker" | "Manu Ginobili" |
+ | "Tim Duncan" | "Tony Parker" | "Hornets" |
+ | "Tim Duncan" | "Tony Parker" | "Spurs" |
+ | "Tim Duncan" | "Tony Parker" | "LaMarcus Aldridge" |
+ | "Tim Duncan" | "Tony Parker" | "LaMarcus Aldridge" |
+ | "Tim Duncan" | "Tony Parker" | "Kyle Anderson" |
+ | "Tim Duncan" | "Tony Parker" | "Tim Duncan" |
+ | "Tim Duncan" | "Tony Parker" | "Tim Duncan" |
+ | "Tim Duncan" | "Tony Parker" | "Manu Ginobili" |
+ | "Tim Duncan" | "Tony Parker" | "Manu Ginobili" |
+ | "Tim Duncan" | "Tony Parker" | "Hornets" |
+ | "Tim Duncan" | "Danny Green" | NULL |
+ | "Tim Duncan" | "Manu Ginobili" | NULL |
+ | "Tim Duncan" | "Manu Ginobili" | NULL |
  When executing query:
  """
  OPTIONAL match (v:player) WHERE v.player.age > 41
@@ -245,7 +276,6 @@ Feature: Multi Line Multi Query Parts
  | count |
  | 4 |
 
- @skip
  Scenario: Multi Line Multi Query Parts
  When executing query:
  """
@@ -288,15 +318,6 @@ Feature: Multi Line Multi Query Parts
  Then the result should be, in order:
  | scount |
  | 270 |
- When executing query:
- """
- MATCH (m)-[]-(n) WHERE id(m)=="Tim Duncan"
- OPTIONAL MATCH (n)-->(v) WHERE v.player.age < m.player.age
- RETURN count(*) AS count
- """
- Then the result should be, in order:
- | count |
- | 45 |
  When executing query:
  """
  MATCH (a:player{age:42}) WITH a
@@ -361,3 +382,9 @@ Feature: Multi Line Multi Query Parts
  RETURN n,v
  """
  Then a SemanticError should be raised at runtime: The where clause of optional match statement that reference variables defined by other statements is not supported yet.
+ When executing query:
+ """
+ MATCH (m)-[]-(n) WHERE id(m)=="Tim Duncan"
+ OPTIONAL MATCH (n)-->(v) WHERE id(v) < id(m) RETURN count(*) AS count
+ """
+ Then a SemanticError should be raised at runtime: The where clause of optional match statement that reference variables defined by other statements is not supported yet.

tests/tck/ldbc/business_intelligence_workload/Read.feature

@@ -156,6 +156,7 @@ Feature: LDBC Business Intelligence Workload - Read
  (forum)-[:HAS_MEMBER]->(person:Person)
  OPTIONAL MATCH
  (person)<-[:HAS_CREATOR]-(post:Post)<-[:CONTAINER_OF]-(popularForum:Forum)
+ WITH popularForums, popularForum, person, post
  WHERE popularForum IN popularForums
  RETURN
  person.Person.id AS personId,