Document the release scripts really used for making releases.
End the polite fiction that build/tools/create-archive.py is used to build the
release because it isn't and all the last releases were built using different
scripts. Document their use in more details.
Use "INNO" env var in Inno Setup script only if it's defined.
Just use the sources directory itself if %INNO% is not defined. In fact, I
think we could always use the source directory, actually, but just in case
there was some reason to do it like this, keep using %INNO% if it is defined.
Also check that WXW_VER environment variable that we use is, in fact, defined.
FIX: Use relative path, not %WXWIN% in ISS script.
Add manually created VC10 project and solution files for the library.
These files were obtained by importing the VC9 files and then setting up the
dependencies correctly in the solution. This is just a temporary solution and
we'll hopefully generate them with bakefile-1 for the 3.0 release but for now
this is better than nothing.
Fix insertion of radio menu items in wxOSX wxMenu.
Deal correctly with updating the indices when a radio item is inserted into an
existing radio group (which wasn't done previously and resulted in a unit test
failure in MenuTestCase::RadioItems()) and also with inserting the normal
items before an existing radio group as the stored indices were not updated
correctly.
The code is still ugly and it probably wouldn't be a bad idea to reuse
wxMenuRadioItemsData used in wxMSW for similar purposes, but at least the unit
tests pass now.
This code was probably copied from wxUniv but was wrong as we can't rely on
the items being always inserted in order. This commit on its own fixes
removing the first radio group menu item but it also makes possible to
properly implement the insertion of new items in the middle of an existing
radio group which couldn't be done with m_startRadioGroup approach at all.
Use [DOMRange markupString] to get selection source.
This DOMRange method, previously used in GetSelectedText(), seems to provide
exactly what we need so there doesn't seem to be any reason to use JS to get
the selection text, especially as it didn't even work under OS X 10.8 and
returned an empty string in the unit test.
The unit test still needs adjustment to pass because we don't get back exactly
the same HTML as we used originally, but with more relaxed matching it does
pass now.
The changes of r74532 were incomplete/wrong and the fix of r74536 didn't
entirely fix the logic here, so try to do it again, hopefully correctly this
time.
Move HAVE_WCSXXX from wx/osx/config_xcode.h to wx/osx/chkconf.h.
The latter file is also used when building from the command line while the
former is only used in Xcode build, as its name indicates (which didn't help
me to notice it, however).
Also restore the test for wcsftime() in configure as it is available even on
older systems but do not test for strnlen() -- as it is not.
Support using GetTextExtent() with empty string to get descent in wxOSX.
Allow measuring the descent and external leading of an empty string by
measuring just a space instead in wxOSX. This makes the behaviour more
consistent with wxMSW and makes the unit test added in r74464 pass under OS X
too.
Don't test for wide character functions in configure under OS X.
The tests for these functions may succeed on the system we're running under
but the functions may not be actually available on the system the application
is going to run under (if it's < 10.7).
So use our own replacements for these functions unless 10.7 was explicitly
chosen as the minimal possible version.
An alternative solution could have been to use weak linking but this is more
difficult and there is no real gain (performance would need to be measured
first).
Fix totally broken LocaleSetter class in the test suite.
It was based on a completely erroneous assumption that setlocale() returns the
locale that had been previously active when it actually returns the newly set
locale.
This fixes unit test failures in StringTestCase under OS X, as the locale
wasn't correctly restored by DateTimeTestCase that used this class.
Suppress harmless clang 3.3 warning about unused wxMessageOutputBest field.
Clang detects that a class member is unused (under non-Windows systems) which
is quite impressive but not really useful in this particular case, so suppress
this warning by "using" it.
Make wxMSW wxSpinCtrl "not enough space" messages more helpful.
And also less annoying: remove the messages from DoMoveWindow() which could be
given during resizing but not necessarily corresponded to the final control
size.
And give more details about which control is not being given enough space when
a too small size is given in the ctor.
Fix crash in docview code if view initialization failed.
The change of r71371 resulted in using the already destroyed (implicitly, done
as part of destroying the wxView that failed to initialize) wxDocument in
wxScopeGuard destructor.
Avoid this and make the old (i.e. pre-r71371) code exception safe while
keeping its semantics exactly, i.e. only delete all views if the document
still exists.
Also add a comment explaining what's going on here because the ownership rules
in docview code are frankly crazy.
The changes of r73483 broke inclusion of the manifest in the programs built
using MinGW because wxUSE_RC_MANIFEST was never defined. Somehow nobody
complained about it but this resulted in using comctl32.dll 5.80 instead of
6.0 and e.g. problems with toolbar appearance (see #13512).
Do use the manifest by default with MinGW and, in fact, all the other
compilers if any other ones still work, as only MSVC is known to embed the
manifest automatically and we take care of it separately.
Move wx/msw/gccpriv.h inclusion back to wx/platform.h from wx/compiler.h.
Do keep compiler-specific wxCHECK_MINGW32_VERSION() in wx/compiler.h as it's
needed by wx/cpp.h which includes wx/compiler.h only. But put the rest of the
stuff in gccpriv.h in its old place as including it before wx/setup.h didn't
work correctly and was unnecessary anyhow.
Use wxNotebook background colour for the tab row background in wxMSW.
The free space near the tabs was always filled with the default light grey
colour, use wxNotebook background colour itself to fill it now.
Do it by changing the code in wxNotebook::OnPaint() to use the background
colour brush and ExtFloodFill() to paint over the default grey, instead of the
patterned background brush which is supposed to be used for the pages area
only and not for the tabs and which was not even taken into account anyhow as
the default window proc erases the entire contents of the DC we pass to it
anyhow (at least in the default top-aligned tabs case).
Also inherit the background colour from parent by default now, this ensures
that wxNotebook blends with its parent when its background colour has been
explicitly set by default, without the need to call its SetBackgroundColour()
explicitly.
Updated the Xcode projects to include wakeuppipe.cpp.
Also fixes the wxiphone project which appeared to have 2 source file references concatenated (stattext.mm and slider.mm) since the last commit. This unfortunately happens sometimes when running makeprojects.applescript.
Use "MSW" and "Windows" more accurately to avoid confusion.
"MSW" is used as the name of toolkit using the standard Windows UI while
"Windows" is the name of the OS under which wxGTK can be now used too (in
addition to wxMSW).
This terminology is not perfect but at least let's keep it unambiguous.
In the majority of cases set the svn:eol-style property (mostly to native for sources and LF for m4 files) and svn:keywords to Id. Applied for files that were added since r72503.
The samples makefiles were not regenerated after the recent Scintilla changes
(see r74425), do it now. There are no real changes for most of them (except
stc sample), but it ensures that rerunning bakefile doesn't change anything.
Extract compiler-specific macro definitions in a new wx/compiler.h.
This solves the problem with wx/defs.h -> wx/platform.h -> wx/setup.h which
resolves to msvc/wx/setup.h -> wx/version.h -> wx/cpp.h include path which
resulted in __VISUALC__ not being defined in wx/cpp.h.
This problem was not new but went unnoticed for a long time and was only
discovered when wxCHECK_VISUALC_VERSION() started being used in wx/cpp.h too
as now the compiler started warning about wrong #if syntax due to it being
undefined.
Putting the compiler-specific definitions in a separate file allows this file
to be included from wx/cpp.h to ensure that these symbols are always defined
in it and also makes things a little better organized.
Fix wrong return value from wxWebViewIE::Find() in 64 bit build.
Using wxNOT_FOUND and an unsigned size() return value in the same operator ?:
resulted in wxNOT_FOUND being converted to an unsigned size_t type and while
converting it back to (signed) long worked in 32 bit builds where long and
size_t have the same size, it was broken in 64 bit builds where their sizes
are different.
Make destructors of COM interface classes virtual.
This is needed to at least silence g++ warnings but may actually even fix a
real problem as these classes can be inherited from and are deleted via a
pointer (to "this" itself, in Release() implementation).
Reconcile HAVE_VARIADIC_MACROS and wxHAS_VARIADIC_MACROS definitions.
The main user-visible effect of this change is that giving configure
--disable-vararg-macros argument now really disables the use of variadic
macros whereas it didn't disable them in wx/cpp.h before.
It is, of course, also less confusing to not have two very similar but
different symbols.
Václav Slavík [Thu, 11 Jul 2013 07:53:35 +0000 (07:53 +0000)]
Use int instead of wxWindowID in wxNewId() and friends.
The functions are available in wxBase builds too, but wx/windowid.h
isn't. Rather than always including that header, just use int, for which
wxWindowID is a typedef. This keeps the functions available in wxBase
for compatibility and is consistent with how IDs are handled in other
parts of wxBase, particularly wxEvent.
Václav Slavík [Thu, 11 Jul 2013 06:58:35 +0000 (06:58 +0000)]
Use wxWindowID in wxNewId() and related functions.
wxNewId(), wxRegisterId() and wxGetCurrentId() functions all work with
window IDs, so they should use the dedicated type. Previously, they
worked with long, which is not even the same type (wxWindowID is int),
causing implicit type conversion warnings.
Fix restoring the status bar help message after closing MSW menu from keyboard.
Add a hack to ignore WM_MENUSELECT messages we get for the top level menu
items: we must not clear the status bar help message when we get those because
it had just been restored to the original message that was there before the
menu was opened from the base class wxEVT_MENU_CLOSE handler, but this message
comes after it when keyboard is used to close the menu.
Allow retrieving the descent and external leading of empty strings.
This used to work before wxTextMeasure changes so make it work again instead
of optimizing the case of empty string away and not returning anything in the
descent and externalLeading output parameters in this case.
Václav Slavík [Tue, 9 Jul 2013 17:29:02 +0000 (17:29 +0000)]
Fix VC++ warnings about __has_include().
The compiler warns about content after defined(__has_included), so avoid
the issue by putting these Clang tests inside one big #ifdef
__has_included block.
Václav Slavík [Tue, 9 Jul 2013 15:44:01 +0000 (15:44 +0000)]
Work around wxFinite() definition conflict with <cmath>.
Using the GNU libstdc++ or Clang's libc++ implementations of the
standard library, <cmath> inclusion undefines the isfinite macro,
resulting in compilation errors when wxFinite() is used. <cmath> can be
included unintentionally, e.g. wx/hashmap.h may result in its inclusion
when using STD containers.
Work around this mess by checking if one of these <cmath>
implementations were already included and using std::isfinite() in that
case.
projects / wxWidgets.git / log
