Optimize IP field parsing

server/src/main/java/org/elasticsearch/index/mapper/ESInetAddressPoint.java

@@ -0,0 +1,91 @@
+/*
+ * Copyright Elasticsearch B.V. and/or licensed to Elasticsearch B.V. under one
+ * or more contributor license agreements. Licensed under the "Elastic License
+ * 2.0", the "GNU Affero General Public License v3.0 only", and the "Server Side
+ * Public License v 1"; you may not use this file except in compliance with, at
+ * your election, the "Elastic License 2.0", the "GNU Affero General Public
+ * License v3.0 only", or the "Server Side Public License, v 1".
+ */
+
+package org.elasticsearch.index.mapper;
+
+import org.apache.lucene.document.Field;
+import org.apache.lucene.document.FieldType;
+import org.apache.lucene.document.InetAddressPoint;
+import org.apache.lucene.util.BytesRef;
+import org.elasticsearch.common.network.CIDRUtils;
+import org.elasticsearch.common.network.InetAddresses;
+import org.elasticsearch.core.SuppressForbidden;
+import org.elasticsearch.xcontent.XContentString;
+
+import java.net.InetAddress;
+
+/**
+ * A Lucene {@link Field} that stores an IP address as a point.
+ * This is similar to {@link InetAddressPoint} but uses a more efficient way to parse IP addresses
+ * that doesn't require the address to be an {@link InetAddress} object.
+ * Otherwise, it behaves just like the {@link InetAddressPoint} field.
+ */
+class ESInetAddressPoint extends Field {
+    private static final FieldType TYPE;
+
+    static {
+        TYPE = new FieldType();
+        TYPE.setDimensions(1, InetAddressPoint.BYTES);
+        TYPE.freeze();
+    }
+
+    private final XContentString ipString;
+    private final InetAddress inetAddress;
+
+    protected ESInetAddressPoint(String name, XContentString ipString) {
+        super(name, TYPE);
+        this.fieldsData = new BytesRef(InetAddresses.encodeAsIpv6(ipString));
+        this.ipString = ipString;
+        this.inetAddress = null;
+    }
+
+    protected ESInetAddressPoint(String name, InetAddress inetAddress) {
+        super(name, TYPE);
+        this.fieldsData = new BytesRef(CIDRUtils.encode(inetAddress.getAddress()));
+        this.inetAddress = inetAddress;
+        this.ipString = null;
+    }
+
+    public InetAddress getInetAddress() {
+        if (ipString != null) {
+            return InetAddresses.forString(ipString.bytes());
+        }
+        if (inetAddress != null) {
+            return inetAddress;
+        }
+        throw new IllegalStateException("Neither ipString nor inetAddress is set");
+    }
+
+    @Override
+    @SuppressForbidden(
+        reason = "Calling InetAddress#getHostAddress to mimic what InetAddressPoint does. "
+            + "Some tests depend on the exact string representation."
+    )
+    public String toString() {
+        StringBuilder result = new StringBuilder();
+        result.append(getClass().getSimpleName());
+        result.append(" <");
+        result.append(name);
+        result.append(':');
+
+        // IPv6 addresses are bracketed, to not cause confusion with historic field:value representation
+        BytesRef bytes = (BytesRef) fieldsData;
+        InetAddress address = InetAddressPoint.decode(BytesRef.deepCopyOf(bytes).bytes);
+        if (address.getAddress().length == 16) {
+            result.append('[');
+            result.append(address.getHostAddress());
+            result.append(']');
+        } else {
+            result.append(address.getHostAddress());
+        }
+
+        result.append('>');
+        return result.toString();
+    }
+}

server/src/main/java/org/elasticsearch/index/mapper/IpFieldMapper.java

@@ -9,7 +9,6 @@
 
 package org.elasticsearch.index.mapper;
 
-import org.apache.lucene.document.Field;
 import org.apache.lucene.document.InetAddressPoint;
 import org.apache.lucene.document.SortedSetDocValuesField;
 import org.apache.lucene.document.StoredField;
@@ -44,6 +43,7 @@
 import org.elasticsearch.search.lookup.FieldValues;
 import org.elasticsearch.search.lookup.SearchLookup;
 import org.elasticsearch.xcontent.XContentParser;
+import org.elasticsearch.xcontent.XContentString;
 
 import java.io.IOException;
 import java.net.InetAddress;
@@ -644,10 +644,12 @@ protected String contentType() {
 
     @Override
     protected void parseCreateField(DocumentParserContext context) throws IOException {
-        InetAddress address;
-        String value = context.parser().textOrNull();
+        ESInetAddressPoint address;
+        XContentString value = context.parser().optimizedTextOrNull();
         try {
-            address = value == null ? nullValue : InetAddresses.forString(value);
+            address = value == null
+                ? nullValue == null ? null : new ESInetAddressPoint(fieldType().name(), nullValue)
+                : new ESInetAddressPoint(fieldType().name(), value);
         } catch (IllegalArgumentException e) {
             if (ignoreMalformed) {
                 context.addIgnoredField(fieldType().name());
@@ -665,29 +667,29 @@ protected void parseCreateField(DocumentParserContext context) throws IOExceptio
         }
         if (offsetsFieldName != null && context.isImmediateParentAnArray() && context.canAddIgnoredField()) {
             if (address != null) {
-                BytesRef sortableValue = new BytesRef(InetAddressPoint.encode(address));
+                BytesRef sortableValue = address.binaryValue();
                 context.getOffSetContext().recordOffset(offsetsFieldName, sortableValue);
             } else {
                 context.getOffSetContext().recordNull(offsetsFieldName);
             }
         }
     }
 
-    private void indexValue(DocumentParserContext context, InetAddress address) {
+    private void indexValue(DocumentParserContext context, ESInetAddressPoint address) {
         if (dimension) {
-            context.getRoutingFields().addIp(fieldType().name(), address);
+            context.getRoutingFields().addIp(fieldType().name(), address.getInetAddress());
         }
+        LuceneDocument doc = context.doc();
         if (indexed) {
-            Field field = new InetAddressPoint(fieldType().name(), address);
-            context.doc().add(field);
+            doc.add(address);
         }
         if (hasDocValues) {
-            context.doc().add(new SortedSetDocValuesField(fieldType().name(), new BytesRef(InetAddressPoint.encode(address))));
+            doc.add(new SortedSetDocValuesField(fieldType().name(), address.binaryValue()));
         } else if (stored || indexed) {
             context.addToFieldNames(fieldType().name());
         }
         if (stored) {
-            context.doc().add(new StoredField(fieldType().name(), new BytesRef(InetAddressPoint.encode(address))));
+            doc.add(new StoredField(fieldType().name(), address.binaryValue()));
         }
     }
 
@@ -698,7 +700,12 @@ protected void indexScriptValues(
         int doc,
         DocumentParserContext documentParserContext
     ) {
-        this.scriptValues.valuesForDoc(searchLookup, readerContext, doc, value -> indexValue(documentParserContext, value));
+        this.scriptValues.valuesForDoc(
+            searchLookup,
+            readerContext,
+            doc,
+            value -> indexValue(documentParserContext, new ESInetAddressPoint(fieldType().name(), value))
+        );
     }
 
     @Override

server/src/main/java/org/elasticsearch/index/mapper/RangeType.java

@@ -54,14 +54,14 @@ public Field getRangeField(String name, RangeFieldMapper.Range r) {
         @Override
         public InetAddress parseFrom(RangeFieldMapper.RangeFieldType fieldType, XContentParser parser, boolean coerce, boolean included)
             throws IOException {
-            InetAddress address = InetAddresses.forString(parser.text());
+            InetAddress address = InetAddresses.forString(parser.optimizedText().bytes());
             return included ? address : nextUp(address);
         }
 
         @Override
         public InetAddress parseTo(RangeFieldMapper.RangeFieldType fieldType, XContentParser parser, boolean coerce, boolean included)
             throws IOException {
-            InetAddress address = InetAddresses.forString(parser.text());
+            InetAddress address = InetAddresses.forString(parser.optimizedText().bytes());
             return included ? address : nextDown(address);
         }
 

server/src/test/java/org/elasticsearch/common/network/InetAddressesTests.java

@@ -19,11 +19,13 @@
 
 import org.elasticsearch.core.Tuple;
 import org.elasticsearch.test.ESTestCase;
+import org.elasticsearch.xcontent.Text;
 import org.hamcrest.Matchers;
 
 import java.net.InetAddress;
 import java.net.NetworkInterface;
 import java.net.UnknownHostException;
+import java.nio.charset.StandardCharsets;
 import java.util.Enumeration;
 
 import static org.hamcrest.Matchers.equalTo;
@@ -171,12 +173,16 @@ public void testForStringIPv6EightColons() throws UnknownHostException {
     }
 
     public void testConvertDottedQuadToHex() throws UnknownHostException {
-        String[] ipStrings = { "7::0.128.0.127", "7::0.128.0.128", "7::128.128.0.127", "7::0.128.128.127" };
+        String[] ipStrings = { "7::0.128.0.127", "7::0.128.0.128", "7::128.128.0.127", "7::0.128.128.127", "::ffff:10.10.1.1" };
 
         for (String ipString : ipStrings) {
             // Shouldn't hit DNS, because it's an IP string literal.
             InetAddress ipv6Addr = InetAddress.getByName(ipString);
             assertEquals(ipv6Addr, InetAddresses.forString(ipString));
+            byte[] asBytes = ipString.getBytes(StandardCharsets.UTF_8);
+            byte[] bytes = new byte[32];
+            System.arraycopy(asBytes, 0, bytes, 8, asBytes.length);
+            assertEquals(ipv6Addr, InetAddresses.forString(bytes, 8, asBytes.length));
             assertTrue(InetAddresses.isInetAddress(ipString));
         }
     }
@@ -244,4 +250,13 @@ public void testParseCidr() {
         assertEquals(InetAddresses.forString("::fffe:0:0"), cidr.v1());
         assertEquals(Integer.valueOf(128), cidr.v2());
     }
+
+    public void testEncodeAsIpv6() throws Exception {
+        assertEquals(16, InetAddresses.encodeAsIpv6(new Text("::1")).length);
+        assertEquals(16, InetAddresses.encodeAsIpv6(new Text("192.168.0.0")).length);
+        assertEquals(
+            "192.168.0.0",
+            InetAddresses.toAddrString(InetAddress.getByAddress(InetAddresses.encodeAsIpv6(new Text("192.168.0.0"))))
+        );
+    }
 }

server/src/test/java/org/elasticsearch/index/mapper/DynamicTemplatesTests.java

@@ -9,7 +9,6 @@
 
 package org.elasticsearch.index.mapper;
 
-import org.apache.lucene.document.InetAddressPoint;
 import org.apache.lucene.document.IntField;
 import org.apache.lucene.document.LongField;
 import org.apache.lucene.index.IndexOptions;
@@ -2205,22 +2204,22 @@ public void testMatchAndUnmatchWithArrayOfFieldNamesMapToIpType() throws IOExcep
         merge(mapperService, dynamicMapping(parsedDoc.dynamicMappingsUpdate()));
         LuceneDocument doc = parsedDoc.rootDoc();
 
-        assertNotEquals(InetAddressPoint.class, doc.getField("one_ip").getClass());
+        assertNotEquals(ESInetAddressPoint.class, doc.getField("one_ip").getClass());
         Mapper fieldMapper = mapperService.documentMapper().mappers().getMapper("one_ip");
         assertNotNull(fieldMapper);
         assertEquals("text", fieldMapper.typeName());
 
-        assertNotEquals(InetAddressPoint.class, doc.getField("ip_two").getClass());
+        assertNotEquals(ESInetAddressPoint.class, doc.getField("ip_two").getClass());
         fieldMapper = mapperService.documentMapper().mappers().getMapper("ip_two");
         assertNotNull(fieldMapper);
         assertEquals("text", fieldMapper.typeName());
 
-        assertEquals(InetAddressPoint.class, doc.getField("three_ip").getClass());
+        assertEquals(ESInetAddressPoint.class, doc.getField("three_ip").getClass());
         fieldMapper = mapperService.documentMapper().mappers().getMapper("three_ip");
         assertNotNull(fieldMapper);
         assertEquals("ip", fieldMapper.typeName());
 
-        assertEquals(InetAddressPoint.class, doc.getField("ip_four").getClass());
+        assertEquals(ESInetAddressPoint.class, doc.getField("ip_four").getClass());
         fieldMapper = mapperService.documentMapper().mappers().getMapper("ip_four");
         assertNotNull(fieldMapper);
         assertEquals("ip", fieldMapper.typeName());
@@ -2257,17 +2256,17 @@ public void testMatchWithArrayOfFieldNamesUsingRegex() throws IOException {
         merge(mapperService, dynamicMapping(parsedDoc.dynamicMappingsUpdate()));
         LuceneDocument doc = parsedDoc.rootDoc();
 
-        assertEquals(InetAddressPoint.class, doc.getField("one100_ip").getClass());
+        assertEquals(ESInetAddressPoint.class, doc.getField("one100_ip").getClass());
         Mapper fieldMapper = mapperService.documentMapper().mappers().getMapper("one100_ip");
         assertNotNull(fieldMapper);
         assertEquals("ip", fieldMapper.typeName());
 
-        assertEquals(InetAddressPoint.class, doc.getField("iptwo").getClass());
+        assertEquals(ESInetAddressPoint.class, doc.getField("iptwo").getClass());
         fieldMapper = mapperService.documentMapper().mappers().getMapper("iptwo");
         assertNotNull(fieldMapper);
         assertEquals("ip", fieldMapper.typeName());
 
-        assertNotEquals(InetAddressPoint.class, doc.getField("threeip").getClass());
+        assertNotEquals(ESInetAddressPoint.class, doc.getField("threeip").getClass());
         fieldMapper = mapperService.documentMapper().mappers().getMapper("threeip");
         assertNotNull(fieldMapper);
         assertEquals("text", fieldMapper.typeName());
@@ -2303,18 +2302,18 @@ public void testSimpleMatchWithArrayOfFieldNamesMixingGlobsAndRegex() throws IOE
         merge(mapperService, dynamicMapping(parsedDoc.dynamicMappingsUpdate()));
         LuceneDocument doc = parsedDoc.rootDoc();
 
-        assertEquals(InetAddressPoint.class, doc.getField("oneip").getClass());
+        assertEquals(ESInetAddressPoint.class, doc.getField("oneip").getClass());
         Mapper fieldMapper = mapperService.documentMapper().mappers().getMapper("oneip");
         assertNotNull(fieldMapper);
         assertEquals("ip", fieldMapper.typeName());
 
         // this one will not match and be an IP field because it was specified with a regex but match_pattern is implicit "simple"
-        assertNotEquals(InetAddressPoint.class, doc.getField("iptwo").getClass());
+        assertNotEquals(ESInetAddressPoint.class, doc.getField("iptwo").getClass());
         fieldMapper = mapperService.documentMapper().mappers().getMapper("iptwo");
         assertNotNull(fieldMapper);
         assertEquals("text", fieldMapper.typeName());
 
-        assertNotEquals(InetAddressPoint.class, doc.getField("threeip").getClass());
+        assertNotEquals(ESInetAddressPoint.class, doc.getField("threeip").getClass());
         fieldMapper = mapperService.documentMapper().mappers().getMapper("threeip");
         assertNotNull(fieldMapper);
         assertEquals("text", fieldMapper.typeName());
@@ -2362,18 +2361,18 @@ public void testDefaultMatchTypeWithArrayOfFieldNamesMixingGlobsAndRegexInPathMa
 
         LuceneDocument doc = parsedDoc.rootDoc();
 
-        assertEquals(InetAddressPoint.class, doc.getField("outer.oneip").getClass());
+        assertEquals(ESInetAddressPoint.class, doc.getField("outer.oneip").getClass());
         Mapper fieldMapper = mapperService.documentMapper().mappers().getMapper("outer.oneip");
         assertNotNull(fieldMapper);
         assertEquals("ip", fieldMapper.typeName());
 
         // this one will not match and be an IP field because it was specified with a regex but match_pattern is implicit "simple"
-        assertNotEquals(InetAddressPoint.class, doc.getField("outer.iptwo").getClass());
+        assertNotEquals(ESInetAddressPoint.class, doc.getField("outer.iptwo").getClass());
         fieldMapper = mapperService.documentMapper().mappers().getMapper("outer.iptwo");
         assertNotNull(fieldMapper);
         assertEquals("text", fieldMapper.typeName());
 
-        assertEquals(InetAddressPoint.class, doc.getField("outer.threeip").getClass());
+        assertEquals(ESInetAddressPoint.class, doc.getField("outer.threeip").getClass());
         fieldMapper = mapperService.documentMapper().mappers().getMapper("outer.threeip");
         assertNotNull(fieldMapper);
         assertEquals("ip", fieldMapper.typeName());
@@ -2421,18 +2420,18 @@ public void testSimpleMatchTypeWithArrayOfFieldNamesMixingGlobsAndRegexInPathMat
 
         LuceneDocument doc = parsedDoc.rootDoc();
 
-        assertEquals(InetAddressPoint.class, doc.getField("outer.oneip").getClass());
+        assertEquals(ESInetAddressPoint.class, doc.getField("outer.oneip").getClass());
         Mapper fieldMapper = mapperService.documentMapper().mappers().getMapper("outer.oneip");
         assertNotNull(fieldMapper);
         assertEquals("ip", fieldMapper.typeName());
 
         // this one will not match and be an IP field because it was specified with a regex but match_pattern is implicit "simple"
-        assertNotEquals(InetAddressPoint.class, doc.getField("outer.iptwo").getClass());
+        assertNotEquals(ESInetAddressPoint.class, doc.getField("outer.iptwo").getClass());
         fieldMapper = mapperService.documentMapper().mappers().getMapper("outer.iptwo");
         assertNotNull(fieldMapper);
         assertEquals("text", fieldMapper.typeName());
 
-        assertEquals(InetAddressPoint.class, doc.getField("outer.threeip").getClass());
+        assertEquals(ESInetAddressPoint.class, doc.getField("outer.threeip").getClass());
         fieldMapper = mapperService.documentMapper().mappers().getMapper("outer.threeip");
         assertNotNull(fieldMapper);
         assertEquals("ip", fieldMapper.typeName());

server/src/test/java/org/elasticsearch/index/mapper/IpScriptMapperTests.java

@@ -75,16 +75,16 @@ protected IpFieldScript.Factory multipleValuesScript() {
     @Override
     protected void assertMultipleValues(List<IndexableField> fields) {
         assertEquals(4, fields.size());
-        assertEquals("InetAddressPoint <field:[0:0:0:0:0:0:0:1]>", fields.get(0).toString());
+        assertEquals("ESInetAddressPoint <field:[0:0:0:0:0:0:0:1]>", fields.get(0).toString());
         assertEquals("docValuesType=SORTED_SET<field:[0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1]>", fields.get(1).toString());
-        assertEquals("InetAddressPoint <field:[0:0:0:0:0:0:0:2]>", fields.get(2).toString());
+        assertEquals("ESInetAddressPoint <field:[0:0:0:0:0:0:0:2]>", fields.get(2).toString());
         assertEquals("docValuesType=SORTED_SET<field:[0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2]>", fields.get(3).toString());
     }
 
     @Override
     protected void assertDocValuesDisabled(List<IndexableField> fields) {
         assertEquals(1, fields.size());
-        assertEquals("InetAddressPoint <field:[0:0:0:0:0:0:0:1]>", fields.get(0).toString());
+        assertEquals("ESInetAddressPoint <field:[0:0:0:0:0:0:0:1]>", fields.get(0).toString());
     }
 
     @Override