Am 16.03.2014 10:04, schrieb Akihiro Sagawa:
On Sat, 15 Mar 2014 00:15:54 +0100, André Hentschel wrote:
{JOHAB_CHARSET, 0x8446, 0x3135,
{JOHAB_CHARSET, 0x8446, 0x3135, TRUE /* Win8.1 */, {TRUE, TRUE, FALSE, FALSE, TRUE, FALSE, FALSE, FALSE, FALSE, FALSE, FALSE, TRUE}},(snip)
ok((pGetCharABCWidthsA(hdc, c[i].a, c[i].a + 1, a) &&pGetCharABCWidthsW(hdc, c[i].w, c[i].w + 1, w) &&memcmp(a, w, sizeof a) == 0) || broken(c[i].broken), "GetCharABCWidthsA and GetCharABCWidthsW should return same widths. charset = %u\n", c[i].cs);Regarding this test about Johab charset, you are missing the point. Which version of Windows does succeed in? I suspect that the test is broken. Don't you think?
As far as I know, other wine testbots just skip the test due to missing Johab (=130) charset font. Moreover, my Windows 7 with Arial Unicode MS or YOzFont[1] doesn't pass because GetCharABCWidthsA returns 0. But I've not tested on Korean version Windows.
If you have further information, please let me know.
[1] YOzFont apparently supports Johab charset. It's available from http://yozvox.web.fc2.com/446F776E6C6F6164.html
Thx for clarification, after testing a lot i finally removed it: https://source.winehq.org/patches/data/103264
After hours of testing, this is the best solution i found. On Windows with fonts that support johab charset, GetCharABCWidthsA fails for characters >= 0xff I see no need to replicate that in Wine and with this behavior there's nothing to test...