[pkg/ottl]: Add ParseXML converter

.chloggen/feat_ottl_xml-parse-function.yaml

@@ -0,0 +1,13 @@
+# Use this changelog template to create an entry for release notes.
+
+# One of 'breaking', 'deprecation', 'new_component', 'enhancement', 'bug_fix'
+change_type: "enhancement"
+
+# The name of the component, or a single word describing the area of concern, (e.g. filelogreceiver)
+component: pkg/ottl
+
+# A brief description of the change.  Surround your text with quotes ("") if it needs to start with a backtick (`).
+note: Add `ParseXML` function for parsing XML from a target string.
+
+# Mandatory: One or more tracking issues related to the change. You can use the PR number here if no issue exists.
+issues: [31133]

pkg/ottl/e2e/e2e_test.go

@@ -479,6 +479,22 @@ func Test_e2e_converters(t *testing.T) {
 				m.PutStr("k2", "v2__!__v2")
 			},
 		},
+		{
+			statement: `set(attributes["test"], ParseXML("<Log id=\"1\"><Message>This is a log message!</Message></Log>"))`,
+			want: func(tCtx ottllog.TransformContext) {
+				log := tCtx.GetLogRecord().Attributes().PutEmptyMap("test")
+				log.PutStr("tag", "Log")
+
+				attrs := log.PutEmptyMap("attributes")
+				attrs.PutStr("id", "1")
+
+				logChildren := log.PutEmptySlice("children")
+
+				message := logChildren.AppendEmpty().SetEmptyMap()
+				message.PutStr("tag", "Message")
+				message.PutStr("content", "This is a log message!")
+			},
+		},
 		{
 			statement: `set(attributes["test"], Seconds(Duration("1m")))`,
 			want: func(tCtx ottllog.TransformContext) {

pkg/ottl/ottlfuncs/README.md

@@ -402,6 +402,7 @@ Available Converters:
 - [ParseCSV](#parsecsv)
 - [ParseJSON](#parsejson)
 - [ParseKeyValue](#parsekeyvalue)
+- [ParseXML](#parsexml)
 - [Seconds](#seconds)
 - [SHA1](#sha1)
 - [SHA256](#sha256)
@@ -897,6 +898,78 @@ Examples:
 - `ParseKeyValue(attributes["pairs"])`
 
 
+### ParseXML
+
+`ParseXML(target)`
+
+The `ParseXML` Converter returns a `pcommon.Map` struct that is the result of parsing the target string as an XML document.
+
+`target` is a Getter that returns a string. This string should be in XML format.
+If `target` is not a string, nil, or cannot be parsed as XML, `ParseXML` will return an error.
+
+Unmarshalling XML is done using the following rules:
+1. All character data for an XML element is trimmed, joined, and placed into the `content` field.
+2. The tag for an XML element is trimmed, and placed into the `tag` field.
+3. The attributes for an XML element is placed as a `pcommon.Map` into the `attribute` field.
+4. Processing instructions, directives, and comments are ignored and not represented in the resultant map.
+5. All child elements are parsed as above, and placed in a `pcommon.Slice`, which is then placed into the `children` field.
+
+For example, the following XML document:
+```xml
+<?xml version="1.0" encoding="UTF-8" ?>
+<Log>
+  <User>
+    <ID>00001</ID>
+    <Name type="first">Joe</Name>
+    <Email>joe.smith@example.com</Email>
+  </User>
+  <Text>User fired alert A</Text>
+</Log>
+```
+
+will be parsed as:
+```json
+{
+  "tag": "Log",
+  "children": [
+    {
+      "tag": "User",
+      "children": [
+        {
+          "tag": "ID",
+          "content": "00001"
+        },
+        {
+          "tag": "Name",
+          "content": "Joe",
+          "attributes": {
+            "type": "first"
+          }
+        },
+        {
+          "tag": "Email",
+          "content": "joe.smith@example.com"
+        }
+      ]
+    },
+    {
+      "tag": "Text",
+      "content": "User fired alert A"
+    }
+  ]
+}
+```
+
+Examples:
+
+- `ParseXML(body)`
+
+- `ParseXML(attributes["xml"])`
+
+- `ParseXML("<HostInfo hostname=\"example.com\" zone=\"east-1\" cloudprovider=\"aws\" />")`
+
+
+
 ### Seconds
 
 `Seconds(value)`

pkg/ottl/ottlfuncs/func_parse_xml.go

@@ -0,0 +1,134 @@
+// Copyright The OpenTelemetry Authors
+// SPDX-License-Identifier: Apache-2.0
+
+package ottlfuncs // import "github.com/open-telemetry/opentelemetry-collector-contrib/pkg/ottl/ottlfuncs"
+
+import (
+	"bytes"
+	"context"
+	"encoding/xml"
+	"errors"
+	"fmt"
+	"strings"
+
+	"go.opentelemetry.io/collector/pdata/pcommon"
+
+	"github.com/open-telemetry/opentelemetry-collector-contrib/pkg/ottl"
+)
+
+type ParseXMLArguments[K any] struct {
+	Target ottl.StringGetter[K]
+}
+
+func NewParseXMLFactory[K any]() ottl.Factory[K] {
+	return ottl.NewFactory("ParseXML", &ParseXMLArguments[K]{}, createParseXMLFunction[K])
+}
+
+func createParseXMLFunction[K any](_ ottl.FunctionContext, oArgs ottl.Arguments) (ottl.ExprFunc[K], error) {
+	args, ok := oArgs.(*ParseXMLArguments[K])
+
+	if !ok {
+		return nil, fmt.Errorf("ParseXMLFactory args must be of type *ParseXMLArguments[K]")
+	}
+
+	return parseXML(args.Target), nil
+}
+
+// parseXML returns a `pcommon.Map` struct that is a result of parsing the target string as XML
+func parseXML[K any](target ottl.StringGetter[K]) ottl.ExprFunc[K] {
+	return func(ctx context.Context, tCtx K) (any, error) {
+		targetVal, err := target.Get(ctx, tCtx)
+		if err != nil {
+			return nil, err
+		}
+
+		parsedXML := xmlElement{}
+
+		decoder := xml.NewDecoder(strings.NewReader(targetVal))
+		err = decoder.Decode(&parsedXML)
+		if err != nil {
+			return nil, fmt.Errorf("unmarshal xml: %w", err)
+		}
+
+		if decoder.InputOffset() != int64(len(targetVal)) {
+			return nil, errors.New("trailing bytes after parsing xml")
+		}
+
+		parsedMap := pcommon.NewMap()
+		parsedXML.intoMap(parsedMap)
+
+		return parsedMap, nil
+	}
+}
+
+type xmlElement struct {
+	tag        string
+	attributes []xml.Attr
+	text       string
+	children   []xmlElement
+}
+
+// UnmarshalXML implements xml.Unmarshaler for xmlElement
+func (a *xmlElement) UnmarshalXML(d *xml.Decoder, start xml.StartElement) error {
+	a.tag = start.Name.Local
+	a.attributes = start.Attr
+
+	for {
+		tok, err := d.Token()
+		if err != nil {
+			return fmt.Errorf("decode next token: %w", err)
+		}
+
+		switch t := tok.(type) {
+		case xml.StartElement:
+			child := xmlElement{}
+			err := d.DecodeElement(&child, &t)
+			if err != nil {
+				return err
+			}
+
+			a.children = append(a.children, child)
+		case xml.EndElement:
+			// End element means we've reached the end of parsing
+			return nil
+		case xml.CharData:
+			// Strip leading/trailing spaces to ignore newlines and
+			// indentation in formatted XML
+			a.text += string(bytes.TrimSpace([]byte(t)))
+		case xml.Comment: // ignore comments
+		case xml.ProcInst: // ignore processing instructions
+		case xml.Directive: // ignore directives
+		default:
+			return fmt.Errorf("unexpected token type %T", t)
+		}
+	}
+}
+
+// intoMap converts and adds the xmlElement into the provided pcommon.Map.
+func (a xmlElement) intoMap(m pcommon.Map) {
+	m.EnsureCapacity(4)
+
+	m.PutStr("tag", a.tag)
+
+	if a.text != "" {
+		m.PutStr("content", a.text)
+	}
+
+	if len(a.attributes) > 0 {
+		attrs := m.PutEmptyMap("attributes")
+		attrs.EnsureCapacity(len(a.attributes))
+
+		for _, attr := range a.attributes {
+			attrs.PutStr(attr.Name.Local, attr.Value)
+		}
+	}
+
+	if len(a.children) > 0 {
+		children := m.PutEmptySlice("children")
+		children.EnsureCapacity(len(a.children))
+
+		for _, child := range a.children {
+			child.intoMap(children.AppendEmpty().SetEmptyMap())
+		}
+	}
+}