JSON converter: add option to convert octetArray fields always as hexa values

By default, if the options is not specified and octetArray fields have size
less or equal to 8 bytes, they are interpreted as unsigned integer.
This code modification preserves the previous behavior.

Author: Lukas955

include/libfds/converters.h

@@ -1388,7 +1388,7 @@ enum fds_convert_drec2json {
      */
     FDS_CD2J_FORMAT_PROTO    = (1U << 5),
     /**
-     * Ingore non-printable charactes (newline, tab, control charactes, etc.)
+     * Ignore non-printable characters (newline, tab, control characters, etc.)
      * in IPFIX string fields instead of escaping them.
      */
     FDS_CD2J_NON_PRINTABLE   = (1U << 6),
@@ -1402,7 +1402,13 @@ enum fds_convert_drec2json {
      * Convert all timestamps to ISO 8601 textual representation of UTC with
      * milliseconds, e.g. "2019-05-22T22:34:57.828Z"
      */
-    FDS_CD2J_TS_FORMAT_MSEC  = (1U << 8)
+    FDS_CD2J_TS_FORMAT_MSEC  = (1U << 8),
+    /**
+     * Always convert IPFIX octetArray field as hexadecimal value in network
+     * byte order i.e. never try to interpret is as unsigned integer.
+     * For example, "0x8BADF00D".
+     */
+    FDS_CD2J_OCTETS_NOINT    = (1U << 9)
 };
 
 /**
@@ -1421,7 +1427,11 @@ enum fds_convert_drec2json {
  *     format "enXX:idYY", where XX is an Enterprise Number and YY is
  *     an Information Element ID.
  *   - Fields with unknown definition of their Information Elements are
- *     interpreted as octetArray (i.e. hexadecimal value in network byte order).
+ *     interpreted as octetArray.
+ *   - Fields with octetArray type are interpreted as unsigned integer if their
+ *     size is less or equal to 8 bytes. Fields with the size above the limit
+ *     are formatted as hexadecimal value in network byte order. To change this
+ *     behavior, see the conversion flags.
  *   - If conversion of an IPFIX field of the Data Record fails (for example,
  *     due to invalid size or value), "null" is used as its value.
  *   - All timestamps are formatted as number of milliseconds from UNIX Epoch.
@@ -1446,7 +1456,7 @@ enum fds_convert_drec2json {
  * \return On success returns a number of characters (excluding the termination
  *   null byte) placed into the buffer \p str. Therefore, if the result is
  *   greater that or equal to zero, conversion was successful (Note: the null
- *   byte has been also successfuly added at the end of the JSON string)
+ *   byte has been also successfully added at the end of the JSON string)
  * \return #FDS_ERR_BUFFER if the length of the result string (including the
  *   termination null byte) would exceed \p str_size and #FDS_CD2J_ALLOW_REALLOC
  *   flag is not set. The context written to the output buffer \p str is

src/converters/json.c

@@ -227,7 +227,7 @@ to_uint(struct context *buffer, const struct fds_drec_field *field)
 static int
 to_octet(struct context *buffer, const struct fds_drec_field *field)
 {
-    if (field->size <= 8) {
+    if (field->size <= 8 && (buffer->flags & FDS_CD2J_OCTETS_NOINT) == 0) {
         // Print as unsigned integer
         return to_uint(buffer, field);
     }

tests/unit_tests/converters/data/iana.xml

@@ -1373,4 +1373,12 @@ and will be lost when the list is regenerated.
         <status>current</status>
     </element>
 
+    <element>
+        <id>1007</id>
+        <name>myOctetArray</name>
+        <dataType>octetArray</dataType>
+        <dataSemantics>default</dataSemantics>
+        <status>current</status>
+    </element>
+
 </ipfix-elements>

tests/unit_tests/converters/json.cpp

@@ -625,6 +625,7 @@ class Drec_extra : public Drec_base {
         trec.add_field(1005,8);             // myMInf
         trec.add_field(1006,8);             // myNan
         trec.add_field(56, 6);              // sourceMacAddress
+        trec.add_field(1007, ipfix_trec::SIZE_VAR); // myOctetArray
         trec.add_field(95,10);              // applicationId
 
         // Prepare an IPFIX Data Record
@@ -655,6 +656,7 @@ class Drec_extra : public Drec_base {
         drec.append_float(VALUE_MY_MINF,8);
         drec.append_float(VALUE_MY_NAN, 8);
         drec.append_mac(VALUE_SRC_MAC);
+        drec.append_octets(VALUE_MY_OCTETS.c_str(), (uint16_t) 6, true);
         drec.append_octets(VALUE_APP_ID.c_str(),(uint16_t)10, false);
 
         register_template(trec);
@@ -663,7 +665,7 @@ class Drec_extra : public Drec_base {
 
     std::string VALUE_SRC_IP4    = "127.0.0.1";
     std::string VALUE_SAMP_NAME1  = "\xc2\xa1\xc3\xbd";
-    std::string VALUE_SAMP_NAME2  = "\xFF\xEE"; // invalid characterss
+    std::string VALUE_SAMP_NAME2  = "\xFF\xEE"; // invalid characters
     std::string VALUE_SAMP_NAME3  = "\xef\xbf\xa6"; // FULLWIDTH WON SIGN (3 bytes)
     std::string VALUE_SAMP_NAME4  = "\xf0\x90\x8e\xa0"; // OLD PERSIAN SIGN A (4 bytes)
     std::string VALUE_DST_IP4    = "8.8.8.8";
@@ -686,11 +688,13 @@ class Drec_extra : public Drec_base {
     double      VALUE_MY_FLOAT32 = 0.5678;
     signed      VALUE_MY_INT     = 1006;
     std::string VALUE_SRC_MAC    = "01:12:1F:13:11:8A";
+    std::string VALUE_MY_OCTETS  = "\x1E\xA3\xAB\xAD\xC0\xDE"; // 33688308793566
+    uint64_t    VALUE_MY_OCTETS_NUM = 33688308793566ULL;
     std::string VALUE_APP_ID     = "\x33\x23\x24\x30\x31\x32\x34\x35\x36\x37"; // 3#$0124567
 
 };
 
-// Test for diferent data types
+// Test of different data types
 TEST_F(Drec_extra, testTypes)
 {
     constexpr size_t BSIZE = 10U;
@@ -706,12 +710,12 @@ TEST_F(Drec_extra, testTypes)
     EXPECT_EQ((double)cfg["iana:myFloat32"], VALUE_MY_FLOAT32);
     EXPECT_EQ(cfg["iana:myBool"], VALUE_MY_BOOL);
     EXPECT_EQ((signed)cfg["iana:myInt"], VALUE_MY_INT);
+    EXPECT_EQ(cfg["iana:myOctetArray"], VALUE_MY_OCTETS_NUM); // interpreted as number
     EXPECT_EQ(cfg["iana:sourceMacAddress"], VALUE_SRC_MAC);
-
     free(buff);
 }
 
-// Test for non printable characters (FDS_CD2J_NON_PRINTABLE)
+// Test of non printable characters (FDS_CD2J_NON_PRINTABLE)
 TEST_F(Drec_extra, nonPrintable)
 {
     constexpr size_t BSIZE = 10U;
@@ -728,7 +732,7 @@ TEST_F(Drec_extra, nonPrintable)
     free(buff);
 }
 
-// Test for eskaping characters
+// Test of escaping characters
 TEST_F(Drec_extra, printableChar)
 {
     constexpr size_t BSIZE = 10U;
@@ -746,7 +750,7 @@ TEST_F(Drec_extra, printableChar)
     free(buff);
 }
 
-// Test for NAN, +INF, -INF values
+// Test of NAN, +INF, -INF values
 TEST_F(Drec_extra, extraValue)
 {
     constexpr size_t BSIZE = 5U;
@@ -769,7 +773,7 @@ TEST_F(Drec_extra, extraValue)
     free(buff);
 }
 
-// Test for other ASCII characters
+// Test of other ASCII characters
 TEST_F(Drec_extra, otherChar)
 {
     constexpr size_t BSIZE = 5U;
@@ -795,7 +799,7 @@ TEST_F(Drec_extra, otherChar)
     free(buff);
 }
 
-// Test for other MAC adress
+// Test of other MAC address
 TEST_F(Drec_extra, macAdr)
 {
     constexpr size_t BSIZE = 5U;
@@ -813,19 +817,20 @@ TEST_F(Drec_extra, macAdr)
     free(buff);
 }
 
-// Test for octet values
+// Test of octet values (FDS_CD2J_OCTETS_NOINT)
 TEST_F(Drec_extra, octVal)
 {
     constexpr size_t BSIZE = 5U;
     char* buff = (char*) malloc(BSIZE);
-    uint32_t flags = FDS_CD2J_ALLOW_REALLOC;
+    uint32_t flags = FDS_CD2J_ALLOW_REALLOC | FDS_CD2J_OCTETS_NOINT;
     size_t buff_size = BSIZE;
 
     int rc = fds_drec2json(&m_drec, flags, m_iemgr.get(), &buff, &buff_size);
     ASSERT_GT(rc, 0);
     EXPECT_NE(buff_size, BSIZE);
     Config cfg = parse_string(buff, JSON, "drec2json");
     EXPECT_EQ((std::string)cfg["iana:applicationId"], "0x33232430313234353637");
+    EXPECT_EQ((std::string)cfg["iana:myOctetArray"], "0x1EA3ABADC0DE");
 
     free(buff);
 }