| Commit message (Collapse) | Author | Age | Files | Lines |
|---|
| |
|
|
| |
environment.
|
| | |
|
| | |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
* unix/Makefile.in: Updates to account for new and deleted files
* win/Makefile.in: tkStubImg.c and tkOldTest.c.
* win/makefile.bc:
* win/makefile.vc:
* generic/tkOldTest.c (new): New file used to create testing
* generic/tkTest.c: commands for testing various Tk
* tests/constraints.tcl: legacy interfaces where a separate
* tests/image.test: compilation unit is needed in order
to #define suitable macros during compilation. Only the effect of
USE_OLD_IMAGE on Tk_CreateImageType() is currently tested, but more
similar testing commands can be added to this same file. New
constraint defined to detect presence of the image type provided by
the new testing code, and a few tests added to exercise it. Having
USE_OLD_IMAGE support tested by the default test suite should reduce
chance of a recurrence of this bug.
* doc/CrtImgType.3: Revised docs to better indicate the legacy
* doc/CrtPhImgFmt.3: nature of the interfaces supported by
USE_OLD_IMAGE.
* generic/tkDecls.h: make genstubs
* generic/tkStubInit.c:
* generic/tk.decls: Reworked USE_OLD_IMAGE support to use
* generic/tk.h: the same support mechanisms both with
* generic/tkStubImg.c (deleted):and without a stub-enabled build. In
each case, route the legacy calls to Tk_CreateImageType and
Tk_CreatePhotoImageFormat through the Tk_CreateOldImageType and
Tk_CreateOldPhotoImageFormat routines. Add those routines to the
public stub table so they're available to a stub-enabled extension.
Remove the definition of Tk_InitImageArgs() and use a macro to
convert any calls to it in source code into a comment.
* generic/tkImage.c: Removed the MODULE_SCOPE declarations that
* generic/tkImgPhoto.c: broke USE_OLD_IMAGE support.
|
| | |
|
| |
|
|
|
|
|
|
| |
* tests/image.test: where Tk has been [load]ed, there's no
* tests/select.test: guarantee that child processes created by
* tests/unixWm.test: [exec [interpreter]] will have Tk in them.
* tests/window.test: Made modifications to force a [load] of Tk
in those situations.
|
| |
|
|
|
| |
* tests/image.test: image command name that would overwrite an
existing command name. Thanks to Michael Schlenker. [Bug 808039].
|
| |
|
|
| |
Can't delete early though as that removes the link with widgets and image names.
|
| |
|
|
|
| |
* tests/constraints.tcl: -loadfile to hold definitions
* tests/*.test: common to all test files.
|
| |
|
|
|
|
|
|
|
|
|
|
| |
that you're not creating an image named the same as .'s command,
which refixes 220891, even when the name of the command has been
changed with 'rename'. The error message is better too.
* generic/tkFrame.c (TkToplevelWindowForCommand): Added way of
mapping from command names to tkwins-for-toplevels.
* tests/image.test (image-1.10,image-1.11): Updated to match new
error message and added test for the rename case.
* generic/tclInt.decls: Made TkToplevelWindowForCommand exported
privately.
|
| | |
|
| |
|
|
|
| |
which also requires no documentation changes - the behaviour is now
what you would predict from existing docs (as opposed to a crash!)
|
| |
|
|
|
|
|
|
|
| |
I'm still not quite sure what exactly was going wrong in the bug, but
I suspect the problem might have actually been due to the deletion of
all images while in the midst of creating an image. The restriction
is pretty light when it comes to existing code, since it previously
caused a core dump due to an image typePtr field being set to random
garbage, an indication of over-eager use of free()...
|
| |
|
|
|
|
|
|
| |
* tests/image.test: Added tests for [image inuse] command.
* generic/tkImage.c (Tk_ImageObjCmd): Added [image inuse] command,
which provides a means for programmers to determine if a given
image is in use by any widgets. [RFE: 3327].
|
| | |
|
| | |
|
| |
|
