manner commented on code in PR #29341: URL: https://github.com/apache/flink/pull/29341#discussion_r4185531656
########## flink-table/flink-table-runtime/src/main/java/org/apache/flink/table/runtime/functions/XmlToVariantParser.java: ########## @@ -0,0 +1,623 @@ +/* + * 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.flink.table.runtime.functions; + +import org.apache.flink.annotation.Internal; +import org.apache.flink.annotation.VisibleForTesting; +import org.apache.flink.types.variant.BinaryVariantInternalBuilder; +import org.apache.flink.types.variant.BinaryVariantInternalBuilder.FieldEntry; +import org.apache.flink.types.variant.Variant; +import org.apache.flink.types.variant.VariantTypeException; + +import javax.annotation.Nullable; +import javax.xml.XMLConstants; +import javax.xml.stream.XMLInputFactory; +import javax.xml.stream.XMLStreamConstants; +import javax.xml.stream.XMLStreamException; +import javax.xml.stream.XMLStreamReader; + +import java.io.StringReader; +import java.math.BigDecimal; +import java.math.BigInteger; +import java.time.DateTimeException; +import java.time.Instant; +import java.time.LocalDate; +import java.time.LocalDateTime; +import java.time.LocalTime; +import java.time.OffsetDateTime; +import java.time.ZoneOffset; +import java.time.format.DateTimeFormatter; +import java.time.temporal.ChronoField; +import java.time.temporal.TemporalAccessor; +import java.util.ArrayList; +import java.util.Arrays; +import java.util.LinkedHashMap; +import java.util.List; +import java.util.Map; +import java.util.function.Consumer; + +import static org.apache.flink.types.variant.BinaryVariantInternalBuilder.toVariantDecimal; +import static org.apache.flink.types.variant.BinaryVariantUtil.SIZE_LIMIT; +import static org.apache.flink.types.variant.BinaryVariantUtil.microsSinceEpoch; +import static org.apache.flink.types.variant.BinaryVariantUtil.nanosSinceEpoch; + +/** + * Parses XML into a {@link Variant} for {@code PARSE_XML} and {@code TRY_PARSE_XML}. The mapping is + * described in the documentation of the {@code VARIANT} data type. + * + * <p>Parsing has two steps. First, {@link #read} reads the document into a tree of {@link + * XmlElement}s, which group the children of each element by name and apply {@code xsi:nil} and + * {@code xsi:type}. Then, {@link #encode} writes the tree into a variant. The variant can't be + * written while reading, because the occurrences of a repeated child become one array, and they + * don't have to be next to each other in the document. + * + * <p>Instances are not thread-safe. + */ +@Internal +public final class XmlToVariantParser { + + private static final String ATTRIBUTE_PREFIX = "@"; + private static final String TEXT_KEY = "$"; + private static final String ORDER_KEY = "#"; + + private static final String XSI_NIL = "xsi:nil"; + private static final String XSI_TYPE = "xsi:type"; + private static final String XSI_NAMESPACE_DECLARATION = "xmlns:xsi"; + + private final XMLInputFactory inputFactory = createInputFactory(); + + @VisibleForTesting + Variant parse(String xml, boolean forceArray) { + return encode(read(xml), forceArray); + } + + /** + * Reads the document into the tree of its root element. Calls on the same document can share + * the tree, since {@link #encode} doesn't change it. + */ + XmlElement read(String xml) { + try { + return readDocument(xml); + } catch (XMLStreamException e) { + // Only the message is kept, since the location of the exception is not serializable. + throw new IllegalArgumentException(e.getMessage()); + } + } + + static Variant encode(XmlElement root, boolean forceArray) { + return new VariantEncoder(forceArray).encodeDocument(root); + } + + private static XMLInputFactory createInputFactory() { + // The JDK implementation, regardless of other StAX implementations on the classpath. + final XMLInputFactory factory = XMLInputFactory.newDefaultFactory(); + factory.setProperty(XMLInputFactory.IS_NAMESPACE_AWARE, false); + factory.setProperty(XMLInputFactory.IS_COALESCING, true); + // Internal entities need DTD support. External entities are enabled only so that they + // reach the resolver and fail, instead of being dropped silently. The JDK refuses all + // protocols in case a load bypasses the resolver. + factory.setProperty(XMLInputFactory.SUPPORT_DTD, true); + factory.setProperty(XMLInputFactory.IS_SUPPORTING_EXTERNAL_ENTITIES, true); + factory.setProperty(XMLConstants.ACCESS_EXTERNAL_DTD, ""); + factory.setXMLResolver( + (publicId, systemId, baseUri, namespace) -> { + throw new XMLStreamException( + String.format( + "External entities and DTDs are not supported: %s", systemId)); + }); + // Set here, so that they don't depend on the JDK version or JVM-wide settings. Apart from + // the depth and the entity sizes, these are the defaults of JDK 11 to 21. + // Each element can add an object and an array to the variant, so 500 levels give at most + // the 1000 levels that PARSE_JSON allows. + setLimit(factory, "maxElementDepth", 500); + setLimit(factory, "entityExpansionLimit", 64_000); + setLimit(factory, "totalEntitySizeLimit", SIZE_LIMIT); + setLimit(factory, "maxGeneralEntitySizeLimit", SIZE_LIMIT); + setLimit(factory, "maxParameterEntitySizeLimit", 1_000_000); + setLimit(factory, "elementAttributeLimit", 10_000); + setLimit(factory, "maxXMLNameLimit", 1_000); + return factory; + } + + private static void setLimit(XMLInputFactory factory, String limit, int value) { + factory.setProperty("http://www.oracle.com/xml/jaxp/properties/"; + limit, value); + } + + private XmlElement readDocument(String xml) throws XMLStreamException { + final XMLStreamReader reader = inputFactory.createXMLStreamReader(new StringReader(xml)); + try { + final String version = reader.getVersion(); + if (!XmlVersion.isSupported(version)) { + throw new XMLStreamException( + String.format("XML %s documents are not supported.", version)); + } + while (reader.next() != XMLStreamConstants.START_ELEMENT) { + // Skip the prolog. + } + final XmlElement root = readElement(reader); + // Lets the reader reject content after the root element. + while (reader.hasNext()) { + reader.next(); + } + return root; + } finally { + reader.close(); + } + } + + private static XmlElement readElement(XMLStreamReader reader) throws XMLStreamException { + final XmlElement element = + new XmlElement(qualifiedName(reader.getPrefix(), reader.getLocalName())); + readAttributes(reader, element); + final StringBuilder textRun = new StringBuilder(); + for (int event = reader.next(); + event != XMLStreamConstants.END_ELEMENT; + event = reader.next()) { + switch (event) { + case XMLStreamConstants.START_ELEMENT: + // A child element ends the current text run. + element.addTextRun(textRun); + element.addChild(readElement(reader)); + break; + case XMLStreamConstants.CHARACTERS: + case XMLStreamConstants.CDATA: + case XMLStreamConstants.SPACE: + textRun.append(reader.getText()); + break; + default: + // Comments and processing instructions don't end the text run. + break; + } + } + element.addTextRun(textRun); + element.end(); + return element; + } + + private static void readAttributes(XMLStreamReader reader, XmlElement element) { + for (int i = 0; i < reader.getAttributeCount(); i++) { + element.addAttribute( + qualifiedName(reader.getAttributePrefix(i), reader.getAttributeLocalName(i)), + reader.getAttributeValue(i)); + } + } + + // The reader splits attribute names into prefix and local name, even if not namespace-aware. + private static String qualifiedName(@Nullable String prefix, String localName) { + return prefix == null || prefix.isEmpty() ? localName : prefix + ':' + localName; + } + + /** + * The XML versions that can be parsed. XML 1.1 isn't supported, since it allows characters that + * XML 1.0 can't represent, e.g. most control characters. Accepting only XML 1.0 keeps every + * result representable as XML 1.0. + */ + private enum XmlVersion { + XML_1_0("1.0"); + + private final String version; + + XmlVersion(String version) { + this.version = version; + } + + /** A document without an XML declaration is XML 1.0. */ + static boolean isSupported(@Nullable String version) { + return version == null + || Arrays.stream(values()) + .anyMatch(supported -> supported.version.equals(version)); + } + } + + // -------------------------------------------------------------------------------------------- + + /** Parses the text of an element with {@code xsi:type}. */ + private static final class XmlSchemaTypes { + + // Parsing a BigInteger or BigDecimal is quadratic in the number of digits. + private static final int MAX_NUMBER_LENGTH = 1000; + + private XmlSchemaTypes() {} + + /** + * Returns how to write the text as a value of the type, or null if the type is unknown or + * the text isn't a valid value of it. The value is written later by the {@link + * VariantEncoder}, but whether the type applies has to be known while reading. + */ + static @Nullable Consumer<BinaryVariantInternalBuilder> parse(String text, String xsiType) { + try { + switch (localPart(xsiType.trim())) { + case "string": + return builder -> builder.appendString(text); + case "boolean": + { + final Boolean value = parseBoolean(text); + return value == null ? null : builder -> builder.appendBoolean(value); + } + case "byte": + { + final byte value = Byte.parseByte(text); + return builder -> builder.appendByte(value); + } + case "short": + { + final short value = Short.parseShort(text); + return builder -> builder.appendShort(value); + } + case "int": + { + final int value = Integer.parseInt(text); + return builder -> builder.appendInt(value); + } + case "long": + { + final long value = Long.parseLong(text); + return builder -> builder.appendLong(value); + } + case "integer": + { + // BigInteger rejects a fractional part. The variant has no unbounded + // integer type, so an integer is stored as a decimal. + final BigDecimal value = + toVariantDecimal( + new BigDecimal( + new BigInteger(checkNumberLength(text)))); + return builder -> builder.appendDecimal(value); + } + case "decimal": + { + final BigDecimal value = + toVariantDecimal(new BigDecimal(checkNumberLength(text))); + return builder -> builder.appendDecimal(value); + } + case "float": + { + // Like in PARSE_JSON, NaN and infinity aren't typed. A number out of + // range parses as infinity. + final float value = Float.parseFloat(text); + return Float.isFinite(value) + ? builder -> builder.appendFloat(value) + : null; + } + case "double": + { + final double value = Double.parseDouble(text); + return Double.isFinite(value) + ? builder -> builder.appendDouble(value) + : null; Review Comment: Added a check to make sure the values comply to the XML spec -- 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: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
