[pkg/ottl] add User Agent parsing

cmd/otelcontribcol/go.mod

@@ -738,6 +738,7 @@ require (
 	github.com/tinylib/msgp v1.2.0 // indirect
 	github.com/tklauser/go-sysconf v0.3.14 // indirect
 	github.com/tklauser/numcpus v0.8.0 // indirect
+	github.com/ua-parser/uap-go v0.0.0-20240611065828-3a4781585db6 // indirect
 	github.com/valyala/fastjson v1.6.4 // indirect
 	github.com/vincent-petithory/dataurl v1.0.0 // indirect
 	github.com/vishvananda/netlink v1.1.1-0.20201029203352-d40f9887b852 // indirect

pkg/ottl/e2e/e2e_test.go

@@ -774,6 +774,15 @@ func Test_e2e_converters(t *testing.T) {
 				m.PutStr("bar", "pass")
 			},
 		},
+		{
+			statement: `set(attributes["test"], UserAgent("curl/7.81.0"))`,
+			want: func(tCtx ottllog.TransformContext) {
+				m := tCtx.GetLogRecord().Attributes().PutEmptyMap("test")
+				m.PutStr("user_agent.original", "curl/7.81.0")
+				m.PutStr("user_agent.name", "curl")
+				m.PutStr("user_agent.version", "7.81.0")
+			},
+		},
 	}
 
 	for _, tt := range tests {

pkg/ottl/go.mod

@@ -12,6 +12,7 @@ require (
 	github.com/open-telemetry/opentelemetry-collector-contrib/internal/coreinternal v0.107.0
 	github.com/open-telemetry/opentelemetry-collector-contrib/pkg/pdatatest v0.107.0
 	github.com/stretchr/testify v1.9.0
+	github.com/ua-parser/uap-go v0.0.0-20240611065828-3a4781585db6
 	go.opentelemetry.io/collector/component v0.107.0
 	go.opentelemetry.io/collector/pdata v1.13.0
 	go.opentelemetry.io/collector/semconv v0.107.0
@@ -31,6 +32,7 @@ require (
 	github.com/go-logr/logr v1.4.2 // indirect
 	github.com/go-logr/stdr v1.2.2 // indirect
 	github.com/gogo/protobuf v1.3.2 // indirect
+	github.com/hashicorp/golang-lru v0.5.4 // indirect
 	github.com/magefile/mage v1.15.0 // indirect
 	github.com/modern-go/concurrent v0.0.0-20180306012644-bacd9c7ef1dd // indirect
 	github.com/modern-go/reflect2 v1.0.2 // indirect
@@ -51,6 +53,7 @@ require (
 	google.golang.org/genproto/googleapis/rpc v0.0.0-20240701130421-f6361c86f094 // indirect
 	google.golang.org/grpc v1.65.0 // indirect
 	google.golang.org/protobuf v1.34.2 // indirect
+	gopkg.in/yaml.v2 v2.4.0 // indirect
 	gopkg.in/yaml.v3 v3.0.1 // indirect
 )
 

pkg/ottl/ottlfuncs/README.md

@@ -461,6 +461,7 @@ Available Converters:
 - [UnixMilli](#unixmilli)
 - [UnixNano](#unixnano)
 - [UnixSeconds](#unixseconds)
+- [UserAgent](#useragent)
 - [UUID](#UUID)
 - [Year](#year)
 
@@ -1568,6 +1569,34 @@ Examples:
 
 - `UnixSeconds(Time("02/04/2023", "%m/%d/%Y"))`
 
+### UserAgent
+
+`UserAgent(value)`
+
+The `UserAgent` Converter parses the string argument trying to match it against well-known user-agent strings.
+
+`value` is a string or a path to a string.  If `value` is not a string an error is returned.
+
+The results of the parsing are returned as a map containing `user_agent.name`, `user_agent.version` and `user_agent.original`
+as defined in semconv v1.25.0.
+
+Parsing is done using the [uap-go package](https://github.com/ua-parser/uap-go). The specific formats it recognizes can be found [here](https://github.com/ua-parser/uap-core/blob/master/regexes.yaml).
+
+Examples:
+
+- `UserAgent("curl/7.81.0")`
+  ```yaml
+  "user_agent.name": "curl"
+  "user_agent.version": "7.81.0"
+  "user_agent.original": "curl/7.81.0"
+  ```
+- `Mozilla/5.0 (X11; Linux x86_64; rv:126.0) Gecko/20100101 Firefox/126.0`
+  ```yaml
+  "user_agent.name": "Firefox"
+  "user_agent.version": "126.0"
+  "user_agent.original": "Mozilla/5.0 (X11; Linux x86_64; rv:126.0) Gecko/20100101 Firefox/126.0"
+  ```
+
 ### URL
 
 `URL(url_string)`

pkg/ottl/ottlfuncs/func_useragent.go

@@ -0,0 +1,46 @@
+// Copyright The OpenTelemetry Authors
+// SPDX-License-Identifier: Apache-2.0
+
+package ottlfuncs // import "github.com/open-telemetry/opentelemetry-collector-contrib/pkg/ottl/ottlfuncs"
+import (
+	"context"
+	"fmt"
+
+	"github.com/open-telemetry/opentelemetry-collector-contrib/pkg/ottl"
+	"github.com/ua-parser/uap-go/uaparser"
+	semconv "go.opentelemetry.io/collector/semconv/v1.25.0"
+)
+
+type UserAgentArguments[K any] struct {
+	UserAgent ottl.StringGetter[K]
+}
+
+func NewUserAgentFactory[K any]() ottl.Factory[K] {
+	return ottl.NewFactory("UserAgent", &UserAgentArguments[K]{}, createUserAgentFunction[K])
+}
+
+func createUserAgentFunction[K any](_ ottl.FunctionContext, oArgs ottl.Arguments) (ottl.ExprFunc[K], error) {
+	args, ok := oArgs.(*UserAgentArguments[K])
+	if !ok {
+		return nil, fmt.Errorf("URLFactory args must be of type *URLArguments[K]")
+	}
+
+	return userAgent[K](args.UserAgent), nil
+}
+
+func userAgent[K any](userAgentSource ottl.StringGetter[K]) ottl.ExprFunc[K] { //revive:disable-line:var-naming
+	parser := uaparser.NewFromSaved()
+
+	return func(ctx context.Context, tCtx K) (any, error) {
+		userAgentString, err := userAgentSource.Get(ctx, tCtx)
+		if err != nil {
+			return nil, err
+		}
+		parsedUserAgent := parser.ParseUserAgent(userAgentString)
+		return map[string]any{
+			semconv.AttributeUserAgentName:     parsedUserAgent.Family,
+			semconv.AttributeUserAgentOriginal: userAgentString,
+			semconv.AttributeUserAgentVersion:  parsedUserAgent.ToVersionString(),
+		}, nil
+	}
+}

pkg/ottl/ottlfuncs/func_useragent_test.go

@@ -0,0 +1,117 @@
+// Copyright The OpenTelemetry Authors
+// SPDX-License-Identifier: Apache-2.0
+
+package ottlfuncs // import "github.com/open-telemetry/opentelemetry-collector-contrib/pkg/ottl/ottlfuncs"
+import (
+	"context"
+	"testing"
+
+	"github.com/open-telemetry/opentelemetry-collector-contrib/pkg/ottl"
+	"github.com/stretchr/testify/assert"
+	"github.com/stretchr/testify/require"
+	semconv "go.opentelemetry.io/collector/semconv/v1.25.0"
+)
+
+func TestUserAgentParser(t *testing.T) {
+	testCases := []struct {
+		Name        string
+		UAString    string
+		ExpectedMap map[string]any
+	}{
+		{
+			Name:     "Firefox",
+			UAString: "Mozilla/5.0 (X11; Linux x86_64; rv:126.0) Gecko/20100101 Firefox/126.0",
+			ExpectedMap: map[string]any{
+				semconv.AttributeUserAgentOriginal: "Mozilla/5.0 (X11; Linux x86_64; rv:126.0) Gecko/20100101 Firefox/126.0",
+				semconv.AttributeUserAgentName:     "Firefox",
+				semconv.AttributeUserAgentVersion:  "126.0",
+			},
+		},
+		{
+			Name:     "Chrome",
+			UAString: "Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/51.0.2704.103 Safari/537.36",
+			ExpectedMap: map[string]any{
+				semconv.AttributeUserAgentOriginal: "Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/51.0.2704.103 Safari/537.36",
+				semconv.AttributeUserAgentName:     "Chrome",
+				semconv.AttributeUserAgentVersion:  "51.0.2704",
+			},
+		},
+		{
+			Name:     "Mobile Safari",
+			UAString: "Mozilla/5.0 (iPhone; CPU iPhone OS 13_5_1 like Mac OS X) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/13.1.1 Mobile/15E148 Safari/604.1",
+			ExpectedMap: map[string]any{
+				semconv.AttributeUserAgentOriginal: "Mozilla/5.0 (iPhone; CPU iPhone OS 13_5_1 like Mac OS X) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/13.1.1 Mobile/15E148 Safari/604.1",
+				semconv.AttributeUserAgentName:     "Mobile Safari",
+				semconv.AttributeUserAgentVersion:  "13.1.1",
+			},
+		},
+		{
+			Name:     "Edge",
+			UAString: "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/91.0.4472.124 Safari/537.36 Edg/91.0.864.59",
+			ExpectedMap: map[string]any{
+				semconv.AttributeUserAgentOriginal: "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/91.0.4472.124 Safari/537.36 Edg/91.0.864.59",
+				semconv.AttributeUserAgentName:     "Edge",
+				semconv.AttributeUserAgentVersion:  "91.0.864",
+			},
+		},
+		{
+			Name:     "Opera",
+			UAString: "Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/51.0.2704.106 Safari/537.36 OPR/38.0.2220.41",
+			ExpectedMap: map[string]any{
+				semconv.AttributeUserAgentOriginal: "Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/51.0.2704.106 Safari/537.36 OPR/38.0.2220.41",
+				semconv.AttributeUserAgentName:     "Opera",
+				semconv.AttributeUserAgentVersion:  "38.0.2220",
+			},
+		},
+		{
+			Name:     "curl",
+			UAString: "curl/7.81.0",
+			ExpectedMap: map[string]any{
+				semconv.AttributeUserAgentOriginal: "curl/7.81.0",
+				semconv.AttributeUserAgentName:     "curl",
+				semconv.AttributeUserAgentVersion:  "7.81.0",
+			},
+		},
+		{
+			Name:     "Unknown user agent",
+			UAString: "foobar/1.2.3 (foo; bar baz)",
+			ExpectedMap: map[string]any{
+				semconv.AttributeUserAgentOriginal: "foobar/1.2.3 (foo; bar baz)",
+				semconv.AttributeUserAgentName:     "Other",
+				semconv.AttributeUserAgentVersion:  "",
+			},
+		},
+		{
+			Name:     "Otel collector 0.106.1 linux/amd64 user agent",
+			UAString: "OpenTelemetry Collector Contrib/0.106.1 (linux/amd64)",
+			ExpectedMap: map[string]any{
+				semconv.AttributeUserAgentOriginal: "OpenTelemetry Collector Contrib/0.106.1 (linux/amd64)",
+				semconv.AttributeUserAgentName:     "Other",
+				semconv.AttributeUserAgentVersion:  "",
+			},
+		},
+	}
+
+	for _, tt := range testCases {
+		t.Run(tt.Name, func(t *testing.T) {
+			source := &ottl.StandardStringGetter[any]{
+				Getter: func(_ context.Context, _ any) (any, error) {
+					return tt.UAString, nil
+				},
+			}
+
+			exprFunc := userAgent[any](source) //revive:disable-line:var-naming
+			res, err := exprFunc(context.Background(), nil)
+			require.NoError(t, err)
+			require.IsType(t, map[string]any{}, res)
+			resMap := res.(map[string]any)
+			assert.Equal(t, tt.ExpectedMap, resMap)
+			assert.Len(t, resMap, len(tt.ExpectedMap))
+			for k, v := range tt.ExpectedMap {
+				if assert.Containsf(t, resMap, k, "key not found %q", k) {
+					assert.Equal(t, v, resMap[k])
+				}
+			}
+		})
+	}
+}

pkg/ottl/ottlfuncs/functions.go

@@ -88,6 +88,7 @@ func converters[K any]() []ottl.Factory[K] {
 		NewUnixSecondsFactory[K](),
 		NewUUIDFactory[K](),
 		NewURLFactory[K](),
+		NewUserAgentFactory[K](),
 		NewAppendFactory[K](),
 		NewYearFactory[K](),
 		NewHexFactory[K](),