Pack/unpack double/float in Little Endian and Big Endian

UPGRADING

@@ -112,6 +112,7 @@ PHP 7.2 UPGRADE NOTES
     ];
   . count() now raises a warning when an invalid parameter is passed.
     Only arrays and objects implementing the Countable interface should be passed.
+  . pack() and unpack() now support float and double in both little and big endian.
 
 - XML:
   . utf8_encode() and utf8_decode() have been moved to the Standard extension

ext/standard/pack.c

@@ -97,8 +97,132 @@ static void php_pack(zval *val, size_t size, int *map, char *output)
 }
 /* }}} */
 
+/* {{{ php_pack_reverse_int32
+ */
+inline uint32_t php_pack_reverse_int32(uint32_t arg)
+{
+    uint32_t result;
+    result = ((arg & 0xFF) << 24) | ((arg & 0xFF00) << 8) | ((arg >> 8) & 0xFF00) | ((arg >> 24) & 0xFF);
+
+	return result;
+}
+/* }}} */
+
+/* {{{ php_pack
+ */
+inline uint64_t php_pack_reverse_int64(uint64_t arg)
+{
+	union Swap64 {
+		uint64_t i;
+		uint32_t ul[2];
+	} tmp, result;
+	tmp.i = arg;
+	result.ul[0] = php_pack_reverse_int32(tmp.ul[1]);
+	result.ul[1] = php_pack_reverse_int32(tmp.ul[0]);
+
+	return result.i;
+}
+/* }}} */
+
+/* {{{ php_pack_copy_float
+ */
+static void php_pack_copy_float(int is_little_endian, void * dst, float f)
+{
+	union Copy32 {
+		float f;
+		uint32_t i;
+	} m;
+	m.f = f;
+
+#ifdef WORDS_BIGENDIAN
+	if (is_little_endian) {
+		m.i = php_pack_reverse_int32(m.i);
+	}
+#else /* WORDS_BIGENDIAN */
+	if (!is_little_endian) {
+		m.i = php_pack_reverse_int32(m.i);
+	}
+#endif /* WORDS_BIGENDIAN */
+
+	memcpy(dst, &m.f, sizeof(float));
+}
+/* }}} */
+
+/* {{{ php_pack_copy_double
+ */
+static void php_pack_copy_double(int is_little_endian, void * dst, double d)
+{
+	union Copy64 {
+		double d;
+		uint64_t i;
+	} m;
+	m.d = d;
+
+#ifdef WORDS_BIGENDIAN
+	if (is_little_endian) {
+		m.i = php_pack_reverse_int64(m.i);
+	}
+#else /* WORDS_BIGENDIAN */
+	if (!is_little_endian) {
+		m.i = php_pack_reverse_int64(m.i);
+	}
+#endif /* WORDS_BIGENDIAN */
+
+	memcpy(dst, &m.d, sizeof(double));
+}
+/* }}} */
+
+/* {{{ php_pack_parse_float
+ */
+static float php_pack_parse_float(int is_little_endian, void * src)
+{
+	union Copy32 {
+		float f;
+		uint32_t i;
+	} m;
+	memcpy(&m.i, src, sizeof(float));
+
+#ifdef WORDS_BIGENDIAN
+	if (is_little_endian) {
+		m.i = php_pack_reverse_int32(m.i);
+	}
+#else /* WORDS_BIGENDIAN */
+	if (!is_little_endian) {
+		m.i = php_pack_reverse_int32(m.i);
+	}
+#endif /* WORDS_BIGENDIAN */
+
+	return m.f;
+}
+/* }}} */
+
+/* {{{ php_pack_parse_double
+ */
+static double php_pack_parse_double(int is_little_endian, void * src)
+{
+	union Copy64 {
+		double d;
+		uint64_t i;
+	} m;
+	memcpy(&m.i, src, sizeof(double));
+
+#ifdef WORDS_BIGENDIAN
+	if (is_little_endian) {
+		m.i = php_pack_reverse_int64(m.i);
+	}
+#else /* WORDS_BIGENDIAN */
+	if (!is_little_endian) {
+		m.i = php_pack_reverse_int64(m.i);
+	}
+#endif /* WORDS_BIGENDIAN */
+
+	return m.d;
+}
+/* }}} */
+
 /* pack() idea stolen from Perl (implemented formats behave the same as there except J and P)
  * Implemented formats are Z, A, a, h, H, c, C, s, S, i, I, l, L, n, N, q, Q, J, P, f, d, x, X, @.
+ * Added g, G for little endian float and big endian float, added e, E for little endian double and big endian double.
  */
 /* {{{ proto string pack(string format, mixed arg1 [, mixed arg2 [, mixed ...]])
    Takes one or more arguments and packs them into a binary string according to the format argument */
@@ -210,8 +334,12 @@ PHP_FUNCTION(pack)
 			case 'N':
 			case 'v':
 			case 'V':
-			case 'f':
-			case 'd':
+			case 'f': /* float */
+			case 'g': /* little endian float */
+			case 'G': /* big endian float */
+			case 'd': /* double */
+			case 'e': /* little endian double */
+			case 'E': /* big endian double */
 				if (arg < 0) {
 					arg = num_args - currentarg;
 				}
@@ -289,11 +417,15 @@ PHP_FUNCTION(pack)
 				break;
 #endif
 
-			case 'f':
+			case 'f': /* float */
+			case 'g': /* little endian float */
+			case 'G': /* big endian float */
 				INC_OUTPUTPOS(arg,sizeof(float))
 				break;
 
-			case 'd':
+			case 'd': /* double */
+			case 'e': /* little endian double */
+			case 'E': /* big endian double */
 				INC_OUTPUTPOS(arg,sizeof(double))
 				break;
 
@@ -468,6 +600,26 @@ PHP_FUNCTION(pack)
 				}
 				break;
 			}
+
+			case 'g': {
+				/* pack little endian float */
+				while (arg-- > 0) {
+					float v = (float) zval_get_double(&argv[currentarg++]);
+					php_pack_copy_float(1, &ZSTR_VAL(output)[outputpos], v);
+					outputpos += sizeof(v);
+				}
+
+				break;
+			}
+			case 'G': {
+				/* pack big endian float */
+				while (arg-- > 0) {
+					float v = (float) zval_get_double(&argv[currentarg++]);
+					php_pack_copy_float(0, &ZSTR_VAL(output)[outputpos], v);
+					outputpos += sizeof(v);
+				}
+				break;
+			}
 
 			case 'd': {
 				while (arg-- > 0) {
@@ -477,6 +629,26 @@ PHP_FUNCTION(pack)
 				}
 				break;
 			}
+
+			case 'e': {
+				/* pack little endian double */
+				while (arg-- > 0) {
+					double v = (double) zval_get_double(&argv[currentarg++]);
+					php_pack_copy_double(1, &ZSTR_VAL(output)[outputpos], v);
+					outputpos += sizeof(v);
+				}
+				break;
+			}
+
+			case 'E': {
+				/* pack big endian double */
+				while (arg-- > 0) {
+					double v = (double) zval_get_double(&argv[currentarg++]);
+					php_pack_copy_double(0, &ZSTR_VAL(output)[outputpos], v);
+					outputpos += sizeof(v);
+				}
+				break;
+			}
 
 			case 'x':
 				memset(&ZSTR_VAL(output)[outputpos], '\0', arg);
@@ -537,6 +709,7 @@ static zend_long php_unpack(char *data, size_t size, int issigned, int *map)
  * Numeric pack types will return numbers, a and A will return strings,
  * f and d will return doubles.
  * Implemented formats are Z, A, a, h, H, c, C, s, S, i, I, l, L, n, N, q, Q, J, P, f, d, x, X, @.
+ * Added g, G for little endian float and big endian float, added e, E for little endian double and big endian double.
  */
 /* {{{ proto array unpack(string format, string input)
    Unpack binary string into named array elements according to format argument */
@@ -679,11 +852,15 @@ PHP_FUNCTION(unpack)
 
 			/* Use sizeof(float) bytes of input */
 			case 'f':
+			case 'g':
+			case 'G':
 				size = sizeof(float);
 				break;
 
 			/* Use sizeof(double) bytes of input */
 			case 'd':
+			case 'e':
+			case 'E':
 				size = sizeof(double);
 				break;
 
@@ -939,18 +1116,37 @@ PHP_FUNCTION(unpack)
 					}
 #endif
 
-					case 'f': {
+					case 'f': /* float */ 
+					case 'g': /* little endian float*/
+					case 'G': /* big endian float*/
+					{
 						float v;
 
-						memcpy(&v, &input[inputpos], sizeof(float));
+						if (type == 'g') {
+							v = php_pack_parse_float(1, &input[inputpos]);
+						} else if (type == 'G') {
+							v = php_pack_parse_float(0, &input[inputpos]);
+						} else {
+							memcpy(&v, &input[inputpos], sizeof(float));
+						}
+
 						add_assoc_double(return_value, n, (double)v);
 						break;
 					}
+
 
-					case 'd': {
+					case 'd': /* double */
+					case 'e': /* little endian float */
+					case 'E': /* big endian float */ 
+					{
 						double v;
-
-						memcpy(&v, &input[inputpos], sizeof(double));
+						if (type == 'e') {
+							v = php_pack_parse_double(1, &input[inputpos]);
+						} else if (type == 'E') {
+							v = php_pack_parse_double(0, &input[inputpos]);
+						} else {
+							memcpy(&v, &input[inputpos], sizeof(double));
+						}
 						add_assoc_double(return_value, n, v);
 						break;
 					}