cloud-fan commented on code in PR #52334: URL: https://github.com/apache/spark/pull/52334#discussion_r2375185512
########## sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/util/LiteralToSqlConverter.scala: ########## @@ -0,0 +1,213 @@ +/* + * 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.spark.sql.catalyst.util + +import org.apache.spark.SparkException +import org.apache.spark.sql.catalyst.analysis.UnresolvedFunction +import org.apache.spark.sql.catalyst.expressions._ +import org.apache.spark.sql.errors.QueryCompilationErrors +import org.apache.spark.sql.types._ + +/** + * Utility for converting Catalyst literal expressions to their SQL string representation. + * + * This object provides a specialized implementation for converting Spark SQL literal + * expressions to their equivalent SQL text representation. It is used by the parameter + * substitution system for EXECUTE IMMEDIATE and other parameterized queries. + * + * Key features: + * - Handles all Spark SQL data types for literal values + * - Supports both Scala collections and Spark internal data structures + * - Proper SQL escaping and formatting + * - Optimized for literal expressions only + * + * Supported data types: + * - Primitives: String, Integer, Long, Float, Double, Boolean, Decimal + * - Temporal: Date, Timestamp, TimestampNTZ, Interval + * - Collections: Array, Map (including nested structures) + * - Special: Null values, Binary data + * - Complex: Nested arrays, maps of arrays, arrays of maps + * + * @example Basic usage: + * {{{ + * val result1 = LiteralToSqlConverter.convert(Literal(42)) + * // result1: "42" + * + * val result2 = LiteralToSqlConverter.convert(Literal("hello")) + * // result2: "'hello'" + * + * val arrayLit = Literal.create(Array(1, 2, 3), ArrayType(IntegerType)) + * val result3 = LiteralToSqlConverter.convert(arrayLit) + * // result3: "ARRAY(1, 2, 3)" + * }}} + * + * @example Complex types: + * {{{ + * val mapLit = Literal.create(Map("key" -> "value"), MapType(StringType, StringType)) + * val result = LiteralToSqlConverter.convert(mapLit) + * // result: "MAP('key', 'value')" + * }}} + * + * @note This utility is thread-safe and can be used concurrently. + * @note Only supports Literal expressions - all parameter values must be pre-evaluated. + * @see [[ParameterHandler]] for the main parameter handling entry point + * @since 4.0.0 + */ +object LiteralToSqlConverter { + + /** + * Convert an expression to its SQL string representation. + * + * This method handles both simple literals and complex expressions that result from + * parameter evaluation. For complex types like arrays and maps, the expressions are + * evaluated to internal data structures that need to be converted back to SQL constructors. + * + * @param expr The expression to convert (typically a Literal, but may be other expressions + * for complex types) + * @return SQL string representation of the expression value + */ + def convert(expr: Expression): String = expr match { + case lit: Literal => convertLiteral(lit) + + // Special handling for UnresolvedFunction expressions that don't naturally evaluate + // Only handle functions that are whitelisted in legacy mode but don't eval() naturally + case UnresolvedFunction(name, children, _, _, _, _, _) => + val functionName = name.mkString(".") + functionName.toLowerCase(java.util.Locale.ROOT) match { + case "array" | "map" | "struct" | "map_from_arrays" | "map_from_entries" => + // Convert whitelisted functions to SQL function call syntax + // Note: array, map, struct are naturally foldable but reach here as UnresolvedFunction + // map_from_arrays and map_from_entries are not naturally foldable + val childrenSql = children.map(convert).mkString(", ") + s"${functionName.toUpperCase(java.util.Locale.ROOT)}($childrenSql)" + case _ => + // Non-whitelisted function - not supported in parameter substitution + // Modern parameter substitution supports: legacy whitelist + foldable expressions + // Functions like str_to_map are intentionally not supported + throw QueryCompilationErrors.unsupportedParameterExpression(expr) + } + + case _ => + // For non-literal expressions, they should be resolved before reaching this converter + // If we get an unresolved expression, it indicates a problem in the calling code + if (!expr.resolved) { + throw SparkException.internalError( + s"LiteralToSqlConverter received unresolved expression: " + + s"${expr.getClass.getSimpleName}. All expressions should be resolved before " + + s"parameter conversion.") + } + if (expr.foldable) { + val value = expr.eval() + val dataType = expr.dataType + convertLiteral(Literal.create(value, dataType)) + } else { + throw SparkException.internalError( + s"LiteralToSqlConverter cannot convert non-foldable expression: " + + s"${expr.getClass.getSimpleName}. All parameter values should be evaluable to " + + s"literals before conversion.") + } + } + + private def convertLiteral(lit: Literal): String = { + // For simple cases, delegate to the existing Literal.sql method + // which already has the correct logic for most data types + try { + lit.sql + } catch { + case _: Exception => + // Fallback to manual conversion for cases where Literal.sql doesn't work + // (e.g., complex types that need constructor syntax) + lit match { + case Literal(null, _) => "NULL" + case Literal(value, dataType) => dataType match { + case ArrayType(elementType, _) => convertArrayLiteral(value, elementType) + case MapType(keyType, valueType, _) => convertMapLiteral(value, keyType, valueType) + case _: StructType => + // Struct literals (row values) - convert to ROW constructor + value match { + case row: org.apache.spark.sql.catalyst.InternalRow => + val structType = dataType.asInstanceOf[StructType] + val fieldValues = (0 until row.numFields).map { i => + if (row.isNullAt(i)) { + "NULL" + } else { + val fieldType = structType.fields(i).dataType + val fieldValue = row.get(i, fieldType) + val fieldLiteral = Literal.create(fieldValue, fieldType) + convert(fieldLiteral) + } + } + s"ROW(${fieldValues.mkString(", ")})" + case _ => s"ROW(${value.toString})" + } + case _ => + // For any other unsupported type, fall back to string representation + s"'${value.toString.replace("'", "''")}'" + } + } + } + } + + private def convertArrayLiteral(value: Any, elementType: DataType): String = { Review Comment: how is this different from `Literal#sql`? -- 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. To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org --------------------------------------------------------------------- To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org For additional commands, e-mail: reviews-h...@spark.apache.org
