osmith
/
wireshark
forked from osmocom/wireshark
1
0
Fork 0
Code Pull Requests Releases Wiki Activity

get_zonename(): don't convert _tzname[] values to UTF-8. 

Now that we're setting the C-language locale to use the UTF-8 code page,
they're already *in* UTF-8; g_locale_to_utf8() doesn't treat the
C-language locale's code page as the "locale" code page, it uses the
system code page, so it reads a UTF-8 string as being in some local code
page's encoding and proceeds to mangle it in the process of converting
it to UTF-8.

Closes #16811 (closed)
This commit is contained in:
Guy Harris 2020-09-01 23:52:35 -07:00
parent c702b44a01
commit a1762dcbb0
1 changed files with 20 additions and 28 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

48
epan/to_str.c
View File

@ -374,34 +374,26 @@ static const gchar *
get_zonename(struct tm *tmp)
{
#if defined(_WIN32)
/* Windows C Runtime: */
/* _tzname is encoded using the "system default ansi code page" */
/* ("which is not necessarily the same as the C library locale"). */
/* So: _tzname must be converted to UTF8 before use. */
/* Alternative: use Windows GetTimeZoneInformation() to get the */
/* timezone name in UTF16 and convert same to UTF8. */
/* XXX: the result is that the timezone name will be based upon the */
/* system code page (iow: the charset of the system). */
/* Since Wireshark is not internationalized, it would seem more */
/* correct to show the timezone name in English, no matter what */
/* the system code page, but I don't how to do that (or if it's */
/* really even possible). */
/* In any case converting to UTF8 presumably at least keeps GTK */
/* happy. (A bug was reported wherein Wireshark crashed in GDK */
/* on a "Japanese version of Windows XP" when trying to copy */
/* the date/time string (containing a copy of _tz_name) to the */
/* clipboard). */
static char *ws_tzname[2] = {NULL, NULL};
/* The g_malloc'd value returned from g_locale_to_utf8() is */
/* cached for all further use so there's no need to ever */
/* g_free() that value. */
if (ws_tzname[tmp->tm_isdst] == NULL) {
ws_tzname[tmp->tm_isdst] = g_locale_to_utf8(_tzname[tmp->tm_isdst], -1, NULL, NULL, NULL);
if (ws_tzname[tmp->tm_isdst] == NULL) {
ws_tzname[tmp->tm_isdst] = "???";
}
}
/*
* The strings in _tzname[] are encoded using the code page
* for the current C-language locale.
*
* On Windows, all Wireshark programs set that code page
* to the UTF-8 code page by calling
*
* setlocale(LC_ALL, ".UTF-8");
*
* so the strings in _tzname[] are UTF-8 strings, and we can
* just return them.
*
* (Note: the above does *not* mean we've set any code pages
* *other* than the one used by the Visual Studio C runtime
* to UTF-8, so don't assume, for example, that the "ANSI"
* versions of Windows APIs will take UTF-8 strings, or that
* non-UTF-16 output to the console will be treated as UTF-8.
* Setting those other code pages can cause problems, especially
* on pre-Windows 10 or older Windows 10 releases.)
*/
return ws_tzname[tmp->tm_isdst];
#else
/*