Fix the printf family for non-ASCII.

The bug here turned out to be that we hadn't increased the constant
corresponding to the maximum number of bytes in a character to match
our new implementation, so any character requiring more than a byte
in UTF-8 would break our printf family.

Bug: 15439554
Change-Id: I693e5e6eb11c640b5886e848502908ec5fff53b1
This commit is contained in:
Elliott Hughes
2014-06-05 20:10:09 -07:00
parent 8aabecf35c
commit 69f05d291d
4 changed files with 50 additions and 2 deletions

View File

@@ -449,3 +449,32 @@ TEST(wchar, wmemmove) {
wmemmove(wstr+5, wstr, sizeof(const_wstr)/sizeof(wchar_t) - 6);
EXPECT_STREQ(L"This This is a test of something or other", wstr);
}
TEST(wchar, mbrtowc_15439554) {
// http://b/15439554
ASSERT_STREQ("C.UTF-8", setlocale(LC_CTYPE, "C.UTF-8"));
uselocale(LC_GLOBAL_LOCALE);
ASSERT_GE(static_cast<size_t>(MB_LEN_MAX), MB_CUR_MAX);
ASSERT_GE(MB_CUR_MAX, 4U);
wchar_t wc;
size_t n;
// 1-byte character.
n = mbrtowc(&wc, "x", MB_CUR_MAX, NULL);
EXPECT_EQ(1U, n);
EXPECT_EQ(L'x', wc);
// 2-byte character.
n = mbrtowc(&wc, "\xc2\xa2", MB_CUR_MAX, NULL);
EXPECT_EQ(2U, n);
EXPECT_EQ(L'¢', wc);
// 3-byte character.
n = mbrtowc(&wc, "\xe2\x82\xac", MB_CUR_MAX, NULL);
EXPECT_EQ(3U, n);
EXPECT_EQ(L'', wc);
// 4-byte character.
n = mbrtowc(&wc, "\xf0\xa4\xad\xa2", MB_CUR_MAX, NULL);
EXPECT_EQ(4U, n);
EXPECT_EQ(L'𤭢', wc);
}