epan: Add BASE_SHOW_UTF_8_PRINTABLE

Add BASE_SHOW_UTF_8_PRINTABLE and related function tvb_utf_8_isprint
for supporting fields of bytes that are "maybe UTF-8" (default or
SHOULD be UTF-8 but could be something else, with no encoding indicator),
such as SSID fields in IEEE 802.11 (See #16208), certain OctetString
fields in Diameter or PFCP, and other places where
BASE_SHOW_ASCII_PRINTABLE is currently used. Fix #5307

2 files changed, 34 insertions, 8 deletions

diff --git a/wsutil/str_util.c b/wsutil/str_util.c
index bb0ab91a0a..d17f9a6d60 100644
--- a/wsutil/str_util.c
+++ b/wsutil/str_util.c
@@ -273,18 +273,24 @@ isprint_string(const gchar *str)
 
 /* Check if an entire UTF-8 string is printable. */
 gboolean
-isprint_utf8_string(const gchar *str, guint length)
+isprint_utf8_string(const gchar *str, const guint length)
 {
-    const char *c;
+    const gchar *strend = str + length;
 
-    if (!g_utf8_validate (str, length, NULL)) {
+    if (!g_utf8_validate_len(str, length, NULL)) {
         return FALSE;
     }
 
-    for (c = str; *c; c = g_utf8_next_char(c)) {
-        if (!g_unichar_isprint(g_utf8_get_char(c))) {
+    while (str < strend) {
+        /* This returns false for G_UNICODE_CONTROL | G_UNICODE_FORMAT |
+         * G_UNICODE_UNASSIGNED | G_UNICODE_SURROGATE
+         * XXX: Could it be ok to have certain format characters, e.g.
+         * U+00AD SOFT HYPHEN? If so, format_text() should be changed too.
+         */
+        if (!g_unichar_isprint(g_utf8_get_char(str))) {
             return FALSE;
         }
+        str = g_utf8_next_char(str);
     }
 
     return TRUE;
diff --git a/wsutil/str_util.h b/wsutil/str_util.h
index 3029753c26..be6ce12485 100644
--- a/wsutil/str_util.h
+++ b/wsutil/str_util.h
@@ -114,14 +114,34 @@ gchar *ascii_strup_inplace(gchar *str);
 WS_DLL_PUBLIC
 gboolean isprint_string(const gchar *str);
 
-/** Check if an entire UTF-8 string consists of printable characters
+/** Given a not-necessarily-null-terminated string, expected to be in
+ *  UTF-8 but possibly containing invalid sequences (as it may have come
+ *  from packet data), and the length of the string, deterimine if the
+ *  string is valid UTF-8 consisting entirely of printable characters.
+ *
+ *  This means that it:
+ *
+ *   does not contain an illegal UTF-8 sequence (including overlong encodings,
+ *   the sequences reserved for UTF-16 surrogate halves, and the values for
+ *   code points above U+10FFFF that are no longer in Unicode)
+ *
+ *   does not contain a non-printable Unicode character such as control
+ *   characters (including internal NULL bytes)
+ *
+ *   does not end in a partial sequence that could begin a valid character;
+ *
+ *   does not start with a partial sequence that could end a valid character;
+ *
+ * and thus guarantees that the result of format_text() would be the same as
+ * that of wmem_strndup() with the same parameters.
  *
  * @param str    The string to be checked
  * @param length The number of bytes to validate
- * @return       TRUE if the entire string is printable, otherwise FALSE
+ * @return       TRUE if the entire string is valid and printable UTF-8,
+ *               otherwise FALSE
  */
 WS_DLL_PUBLIC
-gboolean isprint_utf8_string(const gchar *str, guint length);
+gboolean isprint_utf8_string(const gchar *str, const guint length);
 
 /** Check if an entire string consists of digits
  *