Skip to content

Convert mismatched types of arg/sequence when applying optimization of ShuffleByKeys/PartitionsByKeys to empty sequence #1568

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
Merged
Show file tree
Hide file tree
Changes from 2 commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
72 changes: 72 additions & 0 deletions ydb/core/kqp/ut/federated_query/s3/kqp_federated_query_ut.cpp
Original file line number Diff line number Diff line change
Expand Up @@ -1591,6 +1591,78 @@ Y_UNIT_TEST_SUITE(KqpFederatedQuery) {
UNIT_ASSERT_STRING_CONTAINS(queryExecutionOperation.GetIssues().ToString(), "\"/Root/external_table\" is expected to be external data source");
}
}

Y_UNIT_TEST(QueryWithNoDataInS3) {
const TString externalDataSourceName = "tpc_h_s3_storage_connection";
const TString bucket = "test_bucket_no_data";

Aws::S3::S3Client s3Client = MakeS3Client();
CreateBucket(bucket, s3Client);
// Uncomment if you want to compare with query with data
//UploadObject(bucket, "l/l", R"json({"l_extendedprice": 0.0, "l_discount": 1.0, "l_partkey": 1})json", s3Client);
//UploadObject(bucket, "p/p", R"json({"p_partkey": 1, "p_type": "t"})json", s3Client);

auto kikimr = MakeKikimrRunner(NYql::IHTTPGateway::Make());
auto client = kikimr->GetQueryClient();

{
const TString query = fmt::format(R"sql(
CREATE EXTERNAL DATA SOURCE `{external_source}` WITH (
SOURCE_TYPE="ObjectStorage",
LOCATION="{location}",
AUTH_METHOD="NONE"
);
)sql",
"external_source"_a = externalDataSourceName,
"location"_a = GetBucketLocation(bucket)
);
auto result = client.ExecuteQuery(query, TTxControl::NoTx()).GetValueSync();
UNIT_ASSERT_C(result.GetStatus() == NYdb::EStatus::SUCCESS, result.GetIssues().ToString());
}

{
// YQ-2750
const TString query = fmt::format(R"sql(
$border = Date("1994-08-01");
select
100.00 * sum(case
when StartsWith(p.p_type, 'PROMO')
then l.l_extendedprice * (1 - l.l_discount)
else 0
end) / sum(l.l_extendedprice * (1 - l.l_discount)) as promo_revenue
from
{external_source}.`l/` with ( schema (
l_extendedprice double,
l_discount double,
l_partkey int64,
l_shipdate date
),
format = "json_each_row"
) as l
join
{external_source}.`p/` with ( schema (
p_partkey int64,
p_type string
),
format = "json_each_row"
) as p
on
l.l_partkey = p.p_partkey
where
cast(l.l_shipdate as timestamp) >= $border
and cast(l.l_shipdate as timestamp) < ($border + Interval("P31D"));
)sql",
"external_source"_a = externalDataSourceName
);
auto result = client.ExecuteQuery(query, TTxControl::BeginTx().CommitTx()).GetValueSync();
UNIT_ASSERT_C(result.GetStatus() == NYdb::EStatus::SUCCESS, result.GetIssues().ToString());
auto rs = result.GetResultSetParser(0);
UNIT_ASSERT_VALUES_EQUAL(rs.RowsCount(), 1);
rs.TryNextRow();
TMaybe<double> sum = rs.ColumnParser(0).GetOptionalDouble();
UNIT_ASSERT(!sum);
}
}
}

} // namespace NKikimr::NKqp
47 changes: 44 additions & 3 deletions ydb/library/yql/core/common_opt/yql_co_simple1.cpp
Original file line number Diff line number Diff line change
Expand Up @@ -75,7 +75,7 @@ TExprNode::TPtr OptimizePgCastOverPgConst(const TExprNode::TPtr& input, TExprCon
return input;
}
auto val = input->Child(0);
if (!val->IsCallable("PgConst")) {
if (!val->IsCallable("PgConst")) {
return input;
}

Expand All @@ -85,7 +85,7 @@ TExprNode::TPtr OptimizePgCastOverPgConst(const TExprNode::TPtr& input, TExprCon
YQL_CLOG(DEBUG, Core) << "Remove PgCast unknown->text over PgConst";
return ctx.ChangeChild(*val, 1, castToType);
}

return input;
}

Expand Down Expand Up @@ -6285,7 +6285,48 @@ void RegisterCoSimpleCallables1(TCallableOptimizerMap& map) {
map["ShuffleByKeys"] = map["PartitionsByKeys"] = [](const TExprNode::TPtr& node, TExprContext& ctx, TOptimizeContext& optCtx) {
if (IsEmpty(node->Head(), *optCtx.Types)) {
YQL_CLOG(DEBUG, Core) << node->Content() << " over empty input.";
auto lambdaResult = ctx.Builder(node->Pos()).Apply(node->Tail()).With(0, KeepConstraints(node->HeadPtr(), node->Tail().Head().Head(), ctx)).Seal().Build();

// lambda argument type:
auto& lambdaArg = node->Tail().Head().Head();
auto* lambdaArgType = lambdaArg.GetTypeAnn();
YQL_ENSURE(lambdaArgType, "No argument type for lambda in " << node->Content());

// sequence type:
auto* sequenceType = node->Head().GetTypeAnn();
YQL_ENSURE(sequenceType, "No argument type for sequence in " << node->Content());

TStringBuf typeConversionFunc;
if (sequenceType->GetKind() != lambdaArgType->GetKind()) {
switch (lambdaArgType->GetKind()) {
case ETypeAnnotationKind::List:
if (sequenceType->GetKind() == ETypeAnnotationKind::Optional) {
typeConversionFunc = "ToList";
} else if (IsIn({ ETypeAnnotationKind::Stream, ETypeAnnotationKind::Flow }, sequenceType->GetKind())) {
typeConversionFunc = "ForwardList";
}
break;
case ETypeAnnotationKind::Flow:
typeConversionFunc = "ToFlow";
break;
case ETypeAnnotationKind::Stream:
typeConversionFunc = "ToStream";
break;
case ETypeAnnotationKind::Optional:
if (sequenceType->GetKind() == ETypeAnnotationKind::List) {
typeConversionFunc = "ToOptional";
}
// TODO: convert from Stream/Flow to Optional
break;
default:
break;
}
}

TExprNode::TPtr sequence = KeepConstraints(node->HeadPtr(), lambdaArg, ctx);
if (typeConversionFunc) {
sequence = KeepConstraints(ctx.NewCallable(sequence->Pos(), typeConversionFunc, { sequence }), lambdaArg, ctx);
}
auto lambdaResult = ctx.Builder(node->Pos()).Apply(node->Tail()).With(0, sequence).Seal().Build();
if (node->IsCallable("ShuffleByKeys")) {
auto lambdaType = node->Tail().GetTypeAnn();
if (lambdaType->GetKind() == ETypeAnnotationKind::Optional) {
Expand Down