[PATCH 0/1] MR7506: kernel32/tests: Add broken in test_IdnToNameprepUnicode.
Currently test patterns show a systematic failure up to Windows 10 1607. https://test.winehq.org/data/patterns.html#kernel32:locale As far as I see do these older Windows versions not normalize out the codepoint `U+2066`, [which got added in Unicode 6.3 in 2013](https://codepoints.net/U+2066). A testbot run [with this patch is here](https://testbot.winehq.org/JobDetails.pl?Key=151649). In [this testbot run](https://testbot.winehq.org/JobDetails.pl?Key=151647) is another variant of the patch specifying the whole broken string, maybe this is to be preferred? -- https://gitlab.winehq.org/wine/wine/-/merge_requests/7506
From: Bernhard Übelacker <bernhardu(a)mailbox.org> --- dlls/kernel32/tests/locale.c | 14 +++++++++++++- 1 file changed, 13 insertions(+), 1 deletion(-) diff --git a/dlls/kernel32/tests/locale.c b/dlls/kernel32/tests/locale.c index f21fb14dbad..a2175942a68 100644 --- a/dlls/kernel32/tests/locale.c +++ b/dlls/kernel32/tests/locale.c @@ -5493,6 +5493,15 @@ static void test_GetStringTypeW(void) } } +/* Up to Windows 10 1607 */ +static int is_codepoint_2066_broken(void) +{ + WCHAR buf[10]; + int len = ARRAY_SIZE(buf); + pRtlNormalizeString( 13, L"\x2066", 1, buf, &len ); + return buf[0] != L'\0'; +} + static void test_IdnToNameprepUnicode(void) { struct { @@ -5606,7 +5615,10 @@ static void test_IdnToNameprepUnicode(void) status = pRtlNormalizeString( 13, test_data[i].in, test_data[i].in_len, buf, &len ); ok( status == test_data[i].status || broken(status == test_data[i].broken_status), "%ld: failed %lx\n", i, status ); - if (!status) ok( !wcsnicmp(test_data[i].out, buf, len), "%ld: buf = %s\n", i, wine_dbgstr_wn(buf, len)); + if (!status) + ok( !wcsnicmp(test_data[i].out, buf, len) || + broken(buf[1] == L'\x2066' && is_codepoint_2066_broken()), + "%ld: buf = %s\n", i, wine_dbgstr_wn(buf, len)); } } } -- GitLab https://gitlab.winehq.org/wine/wine/-/merge_requests/7506
participants (1)
-
Bernhard Übelacker