Add a function to dissect an E.164 (MSISDN) number in UTF8 format and use it

in the Diameter dissector.

This new API adds a filter for the MSISDN as well as a subtree and filter for
the Country Code.

Change-Id: Ibcbf4b5f72178b7e4af63efa7496188d608a9de7
Reviewed-on: https://code.wireshark.org/review/3760
Petri-Dish: Jeff Morriss <jeff.morriss.ws@gmail.com>
Tested-by: Petri Dish Buildbot <buildbot-no-reply@wireshark.org>
Reviewed-by: Anders Broman <a.broman58@gmail.com>
Reviewed-by: Jeff Morriss <jeff.morriss.ws@gmail.com>

1 files changed, 163 insertions, 102 deletions

diff --git a/epan/dissectors/packet-e164.c b/epan/dissectors/packet-e164.c
index 93118e30dd..155408430f 100644
--- a/epan/dissectors/packet-e164.c
+++ b/epan/dissectors/packet-e164.c
@@ -345,6 +345,7 @@ const value_string E164_country_code_value[] = {
 	{ 0,	NULL }
 };
 static value_string_ext E164_country_code_value_ext = VALUE_STRING_EXT_INIT(E164_country_code_value);
+
 const value_string E164_GMSS_vals[] = {
 	{ 0x6, "Iridium Satellite LLC"},
 	{ 0x7, "Iridium Satellite LLC"},
@@ -352,6 +353,7 @@ const value_string E164_GMSS_vals[] = {
 	{ 0x9, "Globalstar"},
 	{ 0,	NULL }
 };
+
 const value_string E164_International_Networks_882_vals[] = {
 	{ 0x10, "Global Office Application"},
 	{ 0x12, "HyperStream International (HSI) Data Network"},
@@ -384,6 +386,7 @@ const value_string E164_International_Networks_882_vals[] = {
 	{ 0,	NULL }
 };
 static value_string_ext E164_International_Networks_882_vals_ext = VALUE_STRING_EXT_INIT(E164_International_Networks_882_vals);
+
 const value_string E164_International_Networks_883_vals[] = {
 	{ 0x100, "MediaLincc Ltd"},
 	{ 0x110, "Aicent Inc"},
@@ -397,24 +400,25 @@ const value_string E164_International_Networks_883_vals[] = {
 	{ 0,	NULL }
 };
 
-static int proto_e164						= -1;
+static int proto_e164				= -1;
 static int hf_E164_calling_party_number		= -1;
 static int hf_E164_called_party_number		= -1;
+static int hf_E164_number			= -1;
 static int hf_E164_identification_code		= -1;
-static int hf_E164_country_code				= -1;
-
+static int hf_E164_country_code			= -1;
 
+static int ett_e164_msisdn = -1;
 
 void
 dissect_e164_number(tvbuff_t *tvb, proto_tree *tree, int offset, int length,e164_info_t e164_info)
 {
-	switch (e164_info.e164_number_type){
-	case CALLING_PARTY_NUMBER :
+	switch (e164_info.e164_number_type) {
+	case CALLING_PARTY_NUMBER:
 		proto_tree_add_string(tree, hf_E164_calling_party_number, tvb, offset,
 				length, e164_info.E164_number_str);
 		break;
 
-	case CALLED_PARTY_NUMBER :
+	case CALLED_PARTY_NUMBER:
 		proto_tree_add_string(tree, hf_E164_called_party_number, tvb, offset,
 				length, e164_info.E164_number_str);
 		break;
@@ -424,206 +428,247 @@ dissect_e164_number(tvbuff_t *tvb, proto_tree *tree, int offset, int length,e164
 	}
 
 }
-void
-dissect_e164_cc(tvbuff_t *tvb, proto_tree *tree, int offset, gboolean bcd_coded){
 
+void
+dissect_e164_cc(tvbuff_t *tvb, proto_tree *tree, int offset, e164_encoding_t encoding)
+{
 	int	cc_offset;
 	guint8	address_digit_pair;
-	guint16	id_code;
+	guint16	id_code = 0;
 	guint8	cc_length;
 	guint8	length;
-	guint16 cc;
+	guint16 cc = 0;
 
 	cc_offset = offset;
 	address_digit_pair = tvb_get_guint8(tvb, cc_offset);
 
-	if(!bcd_coded){
+	/* Get the first 3 digits of the MSISDN */
+	switch (encoding) {
+	    case E164_ENC_BINARY:
 		/* Dissect country code after removing non significant zeros */
-		while ( address_digit_pair == 0 ) {
+		while (address_digit_pair == 0) {
 			cc_offset = cc_offset + 1;
 			address_digit_pair = tvb_get_guint8(tvb, cc_offset);
 		}
 		cc = tvb_get_ntohs(tvb, cc_offset);
-		if (( address_digit_pair & 0xf0 ) != 0 ){
+		if ((address_digit_pair & 0xf0) != 0) {
 			cc = cc >> 4;
 		}
-	}else{
+		break;
+	    case E164_ENC_BCD:
 		cc = address_digit_pair &0x0f;
 		cc = cc << 4;
 		cc = cc | (address_digit_pair &0xf0)>>4;
 		cc = cc << 4;
-		if (tvb_bytes_exist(tvb, cc_offset+1, 1)){
+		if (tvb_bytes_exist(tvb, cc_offset+1, 1)) {
 			address_digit_pair = tvb_get_guint8(tvb, cc_offset+1);
 			cc = cc | (address_digit_pair &0x0f);
 		}
-
+		break;
+	    case E164_ENC_UTF8:
+		/* XXX - do we need to worry about leading 0s? */
+		cc  = (tvb_get_guint8(tvb, cc_offset)   - '0') << 8;
+		cc |= (tvb_get_guint8(tvb, cc_offset+1) - '0') << 4;
+		cc |= (tvb_get_guint8(tvb, cc_offset+2) - '0');
+		break;
 	}
 
-	switch ( cc & 0x0f00 ) {
-
+	/* Determine how many of those digits are the Country Code */
+	switch (cc & 0x0f00) {
 	case 0x0:
 		cc_length = 1;
 		break;
-
 	case 0x0100:
 		cc_length = 1;
 		break;
-
 	case 0x0200:
-		switch ( cc & 0x00f0 ) {
+		switch (cc & 0x00f0) {
 		case 0:
-		case 0x70 :
+		case 0x70:
 			cc_length = 2;
 			break;
-		default :
+		default:
 			cc_length = 3;
 			break;
 		}
 		break;
-
-	case 0x0300 :
-		switch ( cc & 0x00f0 ) {
-		case 0 :
-		case 0x10 :
-		case 0x20 :
-		case 0x30 :
-		case 0x40 :
-		case 0x60 :
-		case 0x90 :
+	case 0x0300:
+		switch (cc & 0x00f0) {
+		case 0:
+		case 0x10:
+		case 0x20:
+		case 0x30:
+		case 0x40:
+		case 0x60:
+		case 0x90:
 			cc_length = 2;
 			break;
-		default :
+		default:
 			cc_length = 3;
 			break;
 		}
 		break;
-	case 0x0400 :
-		switch ( cc & 0x00f0 ) {
-		case 0x20 :
+	case 0x0400:
+		switch (cc & 0x00f0) {
+		case 0x20:
 			cc_length = 3;
 			break;
-		default :
+		default:
 			cc_length = 2;
 			break;
 		}
 		break;
-
-	case 0x0500 :
-		switch ( cc & 0x00f0 ) {
-		case 0 :
-		case 0x90 :
+	case 0x0500:
+		switch (cc & 0x00f0) {
+		case 0:
+		case 0x90:
 			cc_length = 3;
 			break;
-		default :
+		default:
 			cc_length = 2;
 			break;
 		}
 		break;
-
-	case 0x0600 :
-		switch ( cc & 0x00f0 ) {
-		case 0x70 :
-		case 0x80 :
-		case 0x90 :
+	case 0x0600:
+		switch (cc & 0x00f0) {
+		case 0x70:
+		case 0x80:
+		case 0x90:
 			cc_length = 3;
 			break;
-		default :
+		default:
 			cc_length = 2;
 			break;
 		}
 		break;
-
-	case 0x0700 :
+	case 0x0700:
 		cc_length = 1;
 		break;
-
-	case 0x0800 :
-		switch ( cc & 0x00f0 ) {
+	case 0x0800:
+		switch (cc & 0x00f0) {
 		case 0x10:
 		case 0x20:
 		case 0x40:
 		case 0x60:
 			cc_length = 2;
 			break;
-		default :
+		default:
 			cc_length = 3;
 			break;
 		}
 		break;
-
-	case 0x0900 :
-		switch ( cc & 0x00f0 ) {
-		case 0 :
-		case 0x10 :
-		case 0x20 :
-		case 0x30 :
-		case 0x40 :
-		case 0x50 :
-		case 0x80 :
+	case 0x0900:
+		switch (cc & 0x00f0) {
+		case 0:
+		case 0x10:
+		case 0x20:
+		case 0x30:
+		case 0x40:
+		case 0x50:
+		case 0x80:
 			cc_length = 2;
 			break;
-		default :
+		default:
 			cc_length = 3;
 			break;
 		}
 		break;
 
-	default :
+	default:
 		cc_length = 0;
 		break;
-	}/* End switch cc */
+	} /* End switch cc */
 
-	switch ( cc_length ) {
-	case 1 :
+	/* (Now that we know how long the CC is,)
+	 * shift off any extra digits we got.
+	 */
+	switch (cc_length) {
+	case 1:
 		cc = cc >> 8;
 		length = 1;
 		break;
-	case 2 :
+	case 2:
 		cc = cc >> 4;
 		length = 1;
 		break;
 	default:
 		length = 2;
 		break;
-	}/* end switch cc_length */
-
-	proto_tree_add_uint(tree, hf_E164_country_code, tvb, cc_offset, length, cc);
-
-	switch ( cc ) {
-	case 0x881 :
-		if (!bcd_coded) {
+	} /* end switch cc_length */
+
+	/* Display the CC */
+	if (encoding == E164_ENC_UTF8)
+	    proto_tree_add_uint(tree, hf_E164_country_code, tvb, cc_offset, cc_length, cc);
+	else
+	    proto_tree_add_uint(tree, hf_E164_country_code, tvb, cc_offset, length, cc);
+
+	/* Handle special Country Codes */
+	switch (cc) {
+	case 0x881:
+		/* Get the 1-digit ID code */
+		switch (encoding) {
+		    case E164_ENC_BINARY:
 			id_code = tvb_get_guint8(tvb, cc_offset + 1) & 0x0f;
-		} else {
+			break;
+		    case E164_ENC_BCD:
 			id_code = (tvb_get_guint8(tvb, cc_offset + 1) & 0xf0) >> 4;
+			break;
+		    case E164_ENC_UTF8:
+			id_code = tvb_get_guint8(tvb, cc_offset + cc_length) - '0';
+			break;
 		}
 		proto_tree_add_uint_format_value(tree, hf_E164_identification_code, tvb, (cc_offset + 1), 1,
 						id_code, "%x %s", id_code, val_to_str_const(id_code, E164_GMSS_vals, "Unknown"));
 		break;
-	case 0x882 :
-		if (!bcd_coded) {
+	case 0x882:
+		/* Get the 2-digit ID code */
+		switch (encoding) {
+		    case E164_ENC_BINARY:
 			id_code = tvb_get_ntohs(tvb, cc_offset + 1);
 			id_code = (id_code & 0x0ff0) >> 4;
-		} else {
+			break;
+		    case E164_ENC_BCD:
 			id_code  = tvb_get_guint8(tvb, cc_offset + 1) & 0xf0;
 			id_code |= tvb_get_guint8(tvb, cc_offset + 2) & 0x0f;
+			break;
+		    case E164_ENC_UTF8:
+			id_code  = (tvb_get_guint8(tvb, cc_offset+cc_length)   - '0') << 4;
+			id_code |= (tvb_get_guint8(tvb, cc_offset+cc_length+1) - '0');
+			break;
 		}
 		proto_tree_add_uint_format_value(tree, hf_E164_identification_code, tvb, (cc_offset + 1), 2,
 						id_code, "%x %s", id_code, val_to_str_ext_const(id_code, &E164_International_Networks_882_vals_ext, "Unknown"));
 		break;
-	case 0x883 :
-		if (!bcd_coded) {
+	case 0x883:
+		/* Get the 3-digit ID code */
+		switch (encoding) {
+		    case E164_ENC_BINARY:
 			id_code = tvb_get_ntohs(tvb, cc_offset + 1);
 			id_code = id_code & 0x0fff;
-		} else {
+			break;
+		    case E164_ENC_BCD:
 			id_code  = (tvb_get_guint8(tvb, cc_offset + 1) & 0xf0) << 4;
 			id_code |= (tvb_get_guint8(tvb, cc_offset + 2) & 0x0f) << 4;
 			id_code |= (tvb_get_guint8(tvb, cc_offset + 2) & 0xf0) >> 4;
+			break;
+		    case E164_ENC_UTF8:
+			id_code  = (tvb_get_guint8(tvb, cc_offset+cc_length)   - '0') << 8;
+			id_code |= (tvb_get_guint8(tvb, cc_offset+cc_length+1) - '0') << 4;
+			id_code |= (tvb_get_guint8(tvb, cc_offset+cc_length+2) - '0');
+			break;
 		}
 		if ((id_code & 0x0ff0) == 0x510) {
-			if (!bcd_coded) {
+			/* Get the 4th digit of the ID code */
+			switch (encoding) {
+			    case E164_ENC_BINARY:
 				id_code = (id_code << 4) | ((tvb_get_guint8(tvb, cc_offset + 3) & 0xf0) >> 4);
-			} else {
+				break;
+			    case E164_ENC_BCD:
 				id_code = (id_code << 4) | (tvb_get_guint8(tvb, cc_offset + 3) & 0x0f);
+				break;
+			    case E164_ENC_UTF8:
+				id_code = (id_code << 4) | (tvb_get_guint8(tvb, cc_offset + cc_length + 3) - '0');
+				break;
 			}
 			proto_tree_add_uint_format_value(tree, hf_E164_identification_code, tvb, (cc_offset + 1), 3,
 					id_code, "%x %s", id_code, val_to_str_const(id_code, E164_International_Networks_883_vals, "Unknown"));
@@ -638,6 +683,23 @@ dissect_e164_cc(tvbuff_t *tvb, proto_tree *tree, int offset, gboolean bcd_coded)
 
 }
 
+const gchar *
+dissect_e164_utf8_number(tvbuff_t *tvb, proto_tree *tree, int offset, int length)
+{
+	proto_item *pi;
+	proto_tree *subtree;
+	gchar *msisdn_str;
+
+	msisdn_str = tvb_get_string_enc(wmem_packet_scope(), tvb, offset, length, ENC_UTF_8);
+	pi = proto_tree_add_string(tree, hf_E164_number, tvb, offset, length, msisdn_str);
+
+	subtree = proto_item_add_subtree(pi, ett_e164_msisdn);
+
+	dissect_e164_cc(tvb, subtree, offset, E164_ENC_UTF8);
+
+	return msisdn_str;
+}
+
 /*
  * Register the protocol with Wireshark.
  *
@@ -661,6 +723,11 @@ proto_register_e164(void)
 		  FT_STRING, BASE_NONE, NULL, 0x0,
 			NULL, HFILL }},
 
+		{ &hf_E164_number,
+		  { "E.164 number (MSISDN)", "e164.msisdn",
+		  FT_STRING, BASE_NONE, NULL, 0x0,
+			NULL, HFILL }},
+
 		{ &hf_E164_identification_code,
 		  { "Identification Code", "e164.identification_code",
 		  FT_UINT32, BASE_HEX, NULL, 0x0,
@@ -672,18 +739,12 @@ proto_register_e164(void)
 			NULL, HFILL }},
 	};
 
-	/*
-	 * Register the protocol name and description
-	 */
-	proto_e164 = proto_register_protocol(
-			"ITU-T E.164 number",
-			"E.164",
-			"e164");
-
-	/*
-	 * Required function calls to register
-	 * the header fields and subtrees used.
-	 */
-	proto_register_field_array(proto_e164, hf, array_length(hf));
+	static gint *ett_e164_array[] = {
+	    &ett_e164_msisdn,
+	};
+
+	proto_e164 = proto_register_protocol("ITU-T E.164 number", "E.164", "e164");
 
+	proto_register_field_array(proto_e164, hf, array_length(hf));
+	proto_register_subtree_array(ett_e164_array, array_length(ett_e164_array));
 }