June 2008 - Wine-devel - list.winehq.org

Re: ntdll: rtlstr.c: Implement checking for control characters in RtlIsTextUnicode [try 2]
by John Klehm 19 Jun '08

19 Jun '08

On Wed, Jun 18, 2008 at 7:02 PM, Zac Brown <zac(a)zacbrown.org> wrote: > Implement setting of IS_TEXT_UNICODE_CONTROLS and > IS_TEXT_UNICODE_REVERSE_CONTROLS when control characters are present > in RtlIsTextUnicode. > > Just a minor point: AFAIK the rule is to use C style /* comments */ only. --John Klehm

2 1

Re: user32:msg: Test for SetFocus
by Dmitry Timoshkov 19 Jun '08

19 Jun '08

"Anatoly Lyutin" <vostok(a)etersoft.ru> wrote: > I try to make this test (in test_ShowWindow : "/* 32 */ { SW_HIDE, TRUE, > 0, WmHide_3, TRUE }, ") passes clear and after some changes I observe > that SetFocus() is working strange for some case. > After this I wrote a test case for SetFocus(). SetFocus() have one > incorrect behavior for working of combination ( visible-invisible parent > or child). The test actually doesn't do anything useful. It doesn't check any focus windows and many vital message parameters. Besides we already have test_SetFocus() in win.c. -- Dmitry.

2 5

Re: urlmon: Fix a test that fails on all platforms [RESEND]
by Alexandre Julliard 19 Jun '08

19 Jun '08

"James Hawkins" <jhawkins(a)codeweavers.com> writes: > Anything wrong with this patch? ../../../tools/runtest -q -P wine -M urlmon.dll -T ../../.. -p urlmon_test.exe.so url.c && touch url.ok url.c:2309: Test succeeded inside todo block: bctx should not be destroyed here make: *** [url.ok] Error 1 -- Alexandre Julliard julliard(a)winehq.org

1 0

Re: Development plans
by Alexandre Julliard 19 Jun '08

19 Jun '08

Eric Pouech <eric.pouech(a)orange.fr> writes: > how to you plan to handle patches that don't commit cleanly ? > do you, by design of the process, intend to drop them (id est: drop > every bug fixes that come after a functional change in the same area > of code), or do you plan to handle yourself the task of retrofit ? If a patch can't be made to apply then I'll ask the author to merge it, but for 1.0.1 that shouldn't happen, we won't be diverging that fast. And once backporting starts to become too painful then it will be time to make a new stable release. -- Alexandre Julliard julliard(a)winehq.org

1 0

Re: mapi32: stop crash under win98
by Paul Vriens 19 Jun '08

19 Jun '08

Alistair Leslie-Hughes wrote: > Hi, > I just added a check for a null pointer since we deference it. > > Changelog: > mapi32: stop crash under win98 > > Best Regards > Alistair Leslie-Hughes > > > > ------------------------------------------------------------------------ > > + if(lpResProp->Value.MVszA.lppszA) + { + ok(lpResProp->Value.MVszA.lppszA[0] == buffer + sizeof(SPropValue) + sizeof(char*), + "wrong lppszA[0] %p\n",lpResProp->Value.MVszA.lppszA[0]); + ok(!strcmp(lpResProp->Value.MVszA.lppszA[0], szTestA), + "wrong string '%s'\n", lpResProp->Value.MVszA.lppszA[0]); + } + else + { + skip("lpResProp->Value.MVszA.lppszA is NULL"); + } I'm not sure if skip() is appropriate here as we are not skipping a function call and we're not skipping subsequent function calls because of this. If Win98 is plainly wrong, and the test maybe has to be changed, here it should probably be: ok(0, "lpResProp->Value.MVszA.lppszA is NULL"); If it's perfectly valid for win98 to return that NULL value is should probably be a trace() as we still have to overcome the dereferencing. -- Cheers, Paul.

1 0

Re: gdi32: stop crash in font tests (try 3)
by Dmitry Timoshkov 19 Jun '08

19 Jun '08

"Alistair Leslie-Hughes" <leslie_alistair(a)hotmail.com> wrote: > - if (!*font_name && font_charset == DEFAULT_CHARSET) /* do it only once */ > + /* EnumFontFamilies with the NULL parameter crashes win98 */ > + if (pGetGlyphIndicesW && !*font_name && font_charset == DEFAULT_CHARSET) /* do it only once */ You are excluding NT4 again. Why don't you simply use unicode version of EnumFontFamilies in this if() block? -- Dmitry.

1 0

Re: ntdll: rtlstr.c: Implement checking for control characters in RtlIsTextUnicode
by Zac Brown 18 Jun '08

18 Jun '08

Zac Brown wrote: > Implement setting of IS_TEXT_UNICODE_CONTROLS and > IS_TEXT_UNICODE_REVERSE_CONTROLS when control characters are present > in RtlIsTextUnicode. > > > ------------------------------------------------------------------------ > > --- > dlls/ntdll/rtlstr.c | 54 ++++++++++++++++++++++++++++++++++++++++++++- > dlls/ntdll/tests/rtlstr.c | 2 - > 2 files changed, 53 insertions(+), 3 deletions(-) > > diff --git a/dlls/ntdll/rtlstr.c b/dlls/ntdll/rtlstr.c > index deec931..2191825 100644 > --- a/dlls/ntdll/rtlstr.c > +++ b/dlls/ntdll/rtlstr.c > @@ -1591,8 +1591,26 @@ NTSTATUS WINAPI RtlFindCharInUnicodeString( > */ > BOOLEAN WINAPI RtlIsTextUnicode( LPCVOID buf, INT len, INT *pf ) > { > + /* lookup table for Unicode values of control chars */ > + static const WCHAR control_chars[] = { > + '\t', > + '\r', > + '\n', > + ' ', > + 0x3000 // CJK space > + }; > + > + /* lookup table for byte-reversed Unicode values of control chars */ > + static const WCHAR reversed_control_chars[] = { > + 0x0900, // tab > + 0x0d00, // carriage return > + 0x0a00, // new line > + 0x2000, // space > + 0x0030 // CJK space > + }; > + > const WCHAR *s = buf; > - int i; > + int i, j; > unsigned int flags = ~0U, out_flags = 0; > > if (len < sizeof(WCHAR)) > @@ -1650,6 +1668,40 @@ BOOLEAN WINAPI RtlIsTextUnicode( LPCVOID buf, INT len, INT *pf ) > } > } > > + /* > + * Check for unicode control characters, both standard and byte reversed. > + * ie: (space, tab, carriage return, line feed, CJK space. > + */ > + > + for (i = 0; i < len; i++) > + { > + if ((flags & IS_TEXT_UNICODE_CONTROLS)) > + { > + if (s[i] == control_chars[0] || > + s[i] == control_chars[1] || > + s[i] == control_chars[2] || > + s[i] == control_chars[3] || > + s[i] == control_chars[4]) > + { > + out_flags |= IS_TEXT_UNICODE_CONTROLS; > + break; > + } > + } > + > + if ((flags & IS_TEXT_UNICODE_REVERSE_CONTROLS) && (out_flags & IS_TEXT_UNICODE_REVERSE_SIGNATURE)) > + { > + if (s[i] == reversed_control_chars[0] || > + s[i] == reversed_control_chars[1] || > + s[i] == reversed_control_chars[2] || > + s[i] == reversed_control_chars[3] || > + s[i] == reversed_control_chars[4]) > + { > + out_flags |= IS_TEXT_UNICODE_REVERSE_CONTROLS; > + break; > + } > + } > + } > + > if (pf) > { > out_flags &= *pf; > diff --git a/dlls/ntdll/tests/rtlstr.c b/dlls/ntdll/tests/rtlstr.c > index aa4a391..5d77f9a 100644 > --- a/dlls/ntdll/tests/rtlstr.c > +++ b/dlls/ntdll/tests/rtlstr.c > @@ -1689,7 +1689,6 @@ static void test_RtlIsTextUnicode(void) > > flags = IS_TEXT_UNICODE_UNICODE_MASK; > ok(pRtlIsTextUnicode(unicode, sizeof(unicode), &flags), "Text should not pass a Unicode\n"); > - todo_wine > ok(flags == (IS_TEXT_UNICODE_STATISTICS | IS_TEXT_UNICODE_CONTROLS), > "Expected flags 0x6, obtained %x\n", flags); > > @@ -1718,7 +1717,6 @@ static void test_RtlIsTextUnicode(void) > > flags = IS_TEXT_UNICODE_REVERSE_MASK; > ok(!pRtlIsTextUnicode(be_unicode, sizeof(unicode) + 2, &flags), "Reverse endian should be Unicode\n"); > - todo_wine > ok(flags == (IS_TEXT_UNICODE_REVERSE_CONTROLS | IS_TEXT_UNICODE_REVERSE_SIGNATURE), > "Expected flags 0xc0, obtained %x\n", flags); > HeapFree(GetProcessHeap(), 0, be_unicode); > > > ------------------------------------------------------------------------ > > Ignore this patch, aside from tabs, it needs something changed. Thanks, Zac

1 0

Re: Development plans
by Steven Edwards 18 Jun '08

18 Jun '08

On Wed, Jun 18, 2008 at 3:31 PM, Eric Pouech <eric.pouech(a)orange.fr> wrote: > how to you plan to handle patches that don't commit cleanly ? > do you, by design of the process, intend to drop them (id est: drop > every bug fixes that come after a functional change in the same area of > code), or do you plan to handle yourself the task of retrofit ? > A+ I think the person that submits the patch should take the effort to make sure the fix applies to 1.0.x if they care about it and post a patch for that to the bug also. If there is backporting work involved in getting it to apply I don't think its reasonable to expect that to be multiplied by hundreds or thousands of patches. Perhaps we could switch to something like the Linux kernel maintenance system where after a certain point, maybe when 1.1 is tagged, Alexandre turns the 1.0 branch over to someone, so that if say a third party vendor has built an application around 1.0.x, there is a public place where such patches can be maintained. Rather than maintenance of the branch just abruptly ending when 1.1 is tagged it would gradually tapper off as those vendors switch to 1.1. Assuming a larger winelib application, some vendor may want to continue to share patches for 1.0 for 3 to 5 years. I don't know if Alexandre will want to maintain 1.0.x for that long assuming 1.1 1.2 1.3 have been tagged or we've moved on to 2.0 -- Steven Edwards "There is one thing stronger than all the armies in the world, and that is an idea whose time has come." - Victor Hugo

1 1

Re: ntdll: rtlstr.c: Implement checking for control characters in RtlIsTextUnicode
by James Hawkins 18 Jun '08

18 Jun '08

On Wed, Jun 18, 2008 at 5:10 PM, Zac Brown <zac(a)zacbrown.org> wrote: > Implement setting of IS_TEXT_UNICODE_CONTROLS and > IS_TEXT_UNICODE_REVERSE_CONTROLS when control characters are present > in RtlIsTextUnicode. > You've mixed tabs and spaces in this patch. -- James Hawkins

1 0

Re: Wine release 1.0
by Peter Bortas 18 Jun '08

18 Jun '08

Thanks for the code thus far! Wine has helped me many times throughout the years. -- Peter Bortas

1 0