reader: KeepPayload option

README.md

@@ -51,8 +51,9 @@ ndntdump emits two output files.
 
 The **packets** file is a [pcapng](https://datatracker.ietf.org/doc/draft-ietf-opsawg-pcapng/) file.
 It contains Ethernet packets that carry NDN traffic.
-IP anonymization has been performed on these packets.
+Address anonymization has been performed on these packets.
 When feasible, NDN packet payload, including Interest ApplicationParameters and Data Content, is zeroized, so that the output can be compressed effectively.
+Payload blanking may be disabled with `--keep-payload` flag.
 
 The **records** file is a [Newline delimited JSON (NDJSON)](https://github.com/ndjson/ndjson-spec) file.
 Each line in this file is a JSON object that describes a NDN packet, either layer 2 or layer 3.
@@ -71,6 +72,7 @@ This may be used with [logrotate](https://man7.org/linux/man-pages/man8/logrotat
 To ensure privacy compliance, ndntdump anonymizes IP and MAC addresses before output files are written.
 IPv4 address keeps its leading 24 bits; IPv6 address keeps its leading 48 bits; MAC address keeps its leading 24 bits.
 Lower bits are XOR'ed with a random value, which is consistent in each run, so that the same original address yields the same anonymized address.
+Notice that this is a very simple and limited anonymization procedure, and we will incorporate better anonymization techniques in the future.
 
 For WebSocket traffic, HTTP request header `X-Forwarded-For` may contain full client address.
 This address is anonymized by changing the lower bits to zeros.

cmd/ndntdump/main.go

@@ -69,6 +69,10 @@ var app = &cli.App{
 			Name:  "keep-mac",
 			Usage: "don't anonymize MAC addresses",
 		},
+		&cli.BoolFlag{
+			Name:  "keep-payload",
+			Usage: "don't zeroize payload",
+		},
 	},
 	Action: func(c *cli.Context) (e error) {
 		if input, e = pcapinput.Open(c.String("ifname"), c.String("input"), c.String("local")); e != nil {
@@ -82,6 +86,7 @@ var app = &cli.App{
 			TCPPort:       c.Int("tcp-port"),
 			WebSocketPort: c.Int("wss-port"),
 			Anonymizer:    ndntdump.NewAnonymizer(keepIPs, c.Bool("keep-mac")),
+			KeepPayload:   c.Bool("keep-payload"),
 		})
 
 		if output, e = fileoutput.Open(c.String("json"), c.String("pcapng")); e != nil {

reader.go

@@ -46,11 +46,12 @@ func saveFlowPorts[P ~uint8, N ~uint16](flow []byte, dir Direction, proto P, src
 
 // Reader reads NDN packets from ZeroCopyPacketDataSource.
 type Reader struct {
-	src     gopacket.ZeroCopyPacketDataSource
-	isLocal func(net.HardwareAddr) bool
-	tcpPort layers.TCPPort
-	wssPort layers.TCPPort
-	anon    *Anonymizer
+	src            gopacket.ZeroCopyPacketDataSource
+	isLocal        func(net.HardwareAddr) bool
+	tcpPort        layers.TCPPort
+	wssPort        layers.TCPPort
+	anon           *Anonymizer
+	zeroizePayload bool
 
 	dlp     *gopacket.DecodingLayerParser
 	dlpTLV  *gopacket.DecodingLayerParser
@@ -186,15 +187,21 @@ func (r *Reader) readPacket(rec *Record) bool {
 	case pkt.Interest != nil:
 		pktType = PktTypeInterest
 		rec.SaveInterest(*pkt.Interest, an.NackNone)
-		zeroizeInterestPayload(pkt.Interest)
+		if r.zeroizePayload {
+			zeroizeInterestPayload(pkt.Interest)
+		}
 	case pkt.Data != nil:
 		pktType = PktTypeData
 		rec.SaveData(*pkt.Data)
-		zeroizeDataPayload(pkt.Data)
+		if r.zeroizePayload {
+			zeroizeDataPayload(pkt.Data)
+		}
 	case pkt.Nack != nil:
 		pktType = PktTypeNack
 		rec.SaveInterest(pkt.Nack.Interest, pkt.Nack.Reason)
-		zeroizeInterestPayload(&pkt.Nack.Interest)
+		if r.zeroizePayload {
+			zeroizeInterestPayload(&pkt.Nack.Interest)
+		}
 	default:
 		return false
 	}
@@ -250,11 +257,12 @@ func (Reader) readFragment(lpl3 ndn.LpL3, frag ndn.LpFragment, rec *Record) {
 // NewReader creates Reader.
 func NewReader(src gopacket.ZeroCopyPacketDataSource, opts ReaderOptions) (r *Reader) {
 	r = &Reader{
-		src:     src,
-		isLocal: opts.IsLocal,
-		tcpPort: layers.TCPPort(opts.TCPPort),
-		wssPort: layers.TCPPort(opts.WebSocketPort),
-		anon:    opts.Anonymizer,
+		src:            src,
+		isLocal:        opts.IsLocal,
+		tcpPort:        layers.TCPPort(opts.TCPPort),
+		wssPort:        layers.TCPPort(opts.WebSocketPort),
+		anon:           opts.Anonymizer,
+		zeroizePayload: !opts.KeepPayload,
 	}
 	if r.wssPort == 0 {
 		r.wssPort = 9696
@@ -273,6 +281,7 @@ type ReaderOptions struct {
 	TCPPort       int
 	WebSocketPort int
 	Anonymizer    *Anonymizer
+	KeepPayload   bool
 }
 
 type incompleteTLV struct {