twalthr commented on a change in pull request #11034: [FLINK-15802][table] Support new type inference for table functions URL: https://github.com/apache/flink/pull/11034#discussion_r378770124
########## File path: flink-table/flink-table-planner-blink/src/main/java/org/apache/calcite/sql/validate/ProcedureNamespace.java ########## @@ -0,0 +1,80 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to you under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ +package org.apache.calcite.sql.validate; + +import org.apache.flink.annotation.Internal; +import org.apache.flink.table.planner.functions.utils.TableSqlFunction; + +import org.apache.calcite.rel.type.RelDataType; +import org.apache.calcite.sql.SqlCall; +import org.apache.calcite.sql.SqlCallBinding; +import org.apache.calcite.sql.SqlKind; +import org.apache.calcite.sql.SqlNode; +import org.apache.calcite.sql.SqlOperator; +import org.apache.calcite.sql.SqlUtil; +import org.apache.calcite.sql.type.SqlTypeName; + +/** + * Namespace whose contents are defined by the result of a call to a user-defined procedure. + * + * <p>Note: Compared to Calcite, this class implements custom logic for dealing with collection tables + * like {@code TABLE(function(...))} procedures. Compared to the SQL standard, Flink's table functions + * can return arbitrary types that are wrapped into a ROW type if necessary. We don't interpret ARRAY + * or MULTISET types as it would be standard. + */ +@Internal +public final class ProcedureNamespace extends AbstractNamespace { + + private final SqlValidatorScope scope; + + private final SqlCall call; + + ProcedureNamespace( + SqlValidatorImpl validator, + SqlValidatorScope scope, + SqlCall call, + SqlNode enclosingNode) { + super(validator, enclosingNode); + this.scope = scope; + this.call = call; + } + + public RelDataType validateImpl(RelDataType targetRowType) { + validator.inferUnknownTypes(validator.unknownType, scope, call); + final RelDataType type = validator.deriveTypeImpl(scope, call); + final SqlOperator operator = call.getOperator(); + final SqlCallBinding callBinding = + new SqlCallBinding(validator, scope, call); + // legacy table functions + if (operator instanceof TableSqlFunction) { Review comment: I will use the previous `instanceof SqlUserDefinedTableFunction` to avoid any classloading issues. I copied the `ProcedureNamespace` in both planners so we should not have problems. ---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
