* contrib/hbgt/doc/en/hbgt.txt
* doc/en/gtslang.txt
* extras/gfspell/spell.prg
* tests/mousetst.prg
* tests/testdbf.prg
* tests/testrdd2.prg
* contrib/hbmisc/numtxthu.prg
* contrib/hbqt/tests/wvtqt.prg
* contrib/hbxbp/tests/wvtqt.prg
* include/hbapigt.h
* tests/db_brows.prg
* tests/gtchars.prg
* tests/tb1.prg
* tests/testbrw.prg
* tests/wcecon.prg
! fixing high 8-bit chars (using various methods)
(finished)
; NOTE: Some sources have been converted to UTF-8,
so from now on make sure to use an UTF-8 enabled
editor and use only UTF-8 chars when typing
non-ASCII (7-bit) characters.
* contrib/hbqt/qtgui/qth/QTreeWidgetItem.qth
! Fixed: regression as per prev commit.
QTreeWidgetItem() can receive another QTreeWidgetItem
as its child and hence is subject to not destroy Qt
object by Harbour.
* contrib/hbqt/tests/browqt.prg
+ Enabled: Build_TreeView() which was causing GPF on exit,
a result of above fix.
* contrib/hbqt/qtgui/qth/QBoxLayout.qth
* contrib/hbqt/qtgui/qth/QComboBox.qth
* contrib/hbqt/qtgui/qth/QDirModel.qth
* contrib/hbqt/qtgui/qth/QDockWidget.qth
* contrib/hbqt/qtgui/qth/QDrag.qth
* contrib/hbqt/qtgui/qth/QGraphicsScene.qth
* contrib/hbqt/qtgui/qth/QGridLayout.qth
* contrib/hbqt/qtgui/qth/QLayout.qth
* contrib/hbqt/qtgui/qth/QMainWindow.qth
* contrib/hbqt/qtgui/qth/QMenu.qth
* contrib/hbqt/qtgui/qth/QMenuBar.qth
* contrib/hbqt/qtgui/qth/QScrollArea.qth
* contrib/hbqt/qtgui/qth/QStandardItemModel.qth
* contrib/hbqt/qtgui/qth/QStatusBar.qth
* contrib/hbqt/qtgui/qth/QTabWidget.qth
* contrib/hbqt/qtgui/qth/QToolBar.qth
* contrib/hbqt/qtgui/qth/QTreeWidgetItem.qth
* contrib/hbqt/qtgui/qth/QWidget.qth
% Replaced: [*D=n*] => [*X=n*] except for
QTreeWidget/QListWidget/QTableWidget as per the change
in core engine where deletion of a Qt object is entirely
thought of on new grounds.
* contrib/hbqt/tests/demoqt.prg
! Optimizations and removal of redundant code.
* contrib/hbqt/qtcore/hbqt.h
* contrib/hbqt/qtcore/hbqt_bind.cpp
* contrib/hbqt/qtcore/hbqt_destroyer.cpp
* contrib/hbqt/qtcore/hbqt_destroyer.h
+ Overhauled: the Qt object destruction protocol.
This commit is the result of two weeks intensive debugging
spree. This commits now handles MT applications better
then before. This commit also covers the mem leaks evident
for certain widgets.
WARNING: regression is possible, please report.
* contrib/hbqt/qtcore/hbqt_bind.cpp
+ Implemented: Events, Slots, Destroyers on thread level.
Previously those were created per object level. This greatly
decreased the os-resource count and also simplifies the code.
Thanks to Przemek for the tip.
+ Implemented: release of all objects per thread level.
hbQT employs internal counter to identify threads.
All widgets are created in one global list which has
made it possible to use Qt's blocking widgets, like
QMessageBox, in Harbour MT applications. It seems that
Qt looks for the event processing only in main appln
thread and hence these widgets were not usable in
other threads. The testbed is hbqt/tests/demoqt.prg
which is now multithreaded. You check it with
menu option <Dialog><Another Dialog>.
+ Formatting: in tracelog entries.
* contrib/hbqt/qtcore/hbqt_hbqevents.cpp
- Commented out: a hack where QClosEvent was being treated
differently then other events.
* contrib/hbqt/tests/demoqt.prg
* Parented: QMessageBox() ( Read below )
* contrib/hbide/idedocks.prg
- Removed: one hack which at begining was there to activate
QMainWindow's docking area resizable.
* contrib/hbide/idemain.prg
+ Changed: ::cWrkCodec to be "EN" by default.
It fixes recent regression in hbIDE for certain users.
* contrib/hbide/idemisc.prg
* Parented: QMessageBox() should always be constructed with a
parent because of special handelling of blocking dialog
where parent is treated slightly different than other
Qt widgets. This fixes a long-standing bug in hbIDE where
appnl was rendered in indefinite loop when "X" button
was employed to terminate and confirm dialog would
not be terminated internally.
So for QMessageBox() and, to be on the safer side, all
widets derived from QDialog() should be constructed
with a parent and at the end be reparented with a
just-to-be-deleted-QWidget(), like:
oMB := QMessageBox( oParent )
...
oMB:setParent( QWidget() )
RETURN something
* contrib/hbqt/qtcore/hbqt_pointer.cpp
* contrib/hbqt/qtcore/hbqt_hbqslots.cpp
* contrib/hbqt/qtcore/hbqt_misc.prg
* Moved: all slots related functions from hbqt_pointer to hbqt_hbqslots
.cpp.
- Removed: __pSlots variable and use of harbour class HBQSlots
Partially implemented hbqt_disconnect function
hbqt_connect now also accepts harbour style
(qtObject, signal, codeblock).
It is the preferred way of making a signal/slot connection...
- contrib/hbqt/qtcore/qth/HBQSlots.qth
* contrib/hbqt/qtcore/qth/filelist.hbm
* removed generated harbour HBQSlots class
* contrib/hbqt/qtcore/hbqt_hbqevents.cpp
! Fix to prev.
* contrib/hbqt/tests/browqt.prg
* contrib/hbqt/tests/dbfbrowser.prg
* contrib/hbqt/tests/demoqt.prg
* contrib/hbqt/tests/draggable.prg
* contrib/hbqt/tests/inherit.prg
* contrib/hbqt/tests/qtrevamp.prg
* contrib/hbqt/tests/testbrow.prg
- Removed: :disconnect() calls. Now all is working like
the way we are accustomed to.
* contrib/hbide/hbqreportsmanager.prg
* contrib/hbide/ideharbourhelp.prg
* contrib/hbide/ideskeletons.prg
* contrib/hbide/ideuisrcmanager.prg
- Removed: all calls to :hasAValidPointer().
* contrib/hbqt/qtcore/hbqt_bind.cpp
+ Added: more debug code to view what happens when
appln exits.
* contrib/hbqt/qtcore/hbqt_hbqevents.cpp
* contrib/hbqt/qtcore/hbqt_hbqevents.h
* contrib/hbqt/qtcore/hbqt_hbqslots.cpp
* contrib/hbqt/qtcore/qth/HBQEvents.qth
! Synchronized: per QObject standards.
* contrib/hbqt/qtgui/qth/QStringListModel.qth
! Fixed: wrong if/else if constructs.
* contrib/hbqt/qtcore/hbqt_misc.prg
* contrib/hbqt/qtgui/THbQtUI.prg
- Removed: the use of :pPtr variable plus method :hasValidPointer().
* contrib/hbqt/tests/demoqt.prg
! Reworked: to adapt to revamped hbQT where variable is not needed
to hold the built widget if parent is provided or is :connect()ed.
* contrib/hbqt/gtqtc/gtqtc.cpp
* contrib/hbqt/hbmk2_qt.hb
* contrib/hbqt/qtcore/hbqt.h
* contrib/hbqt/qtcore/hbqt_bind.cpp
* contrib/hbqt/qtcore/hbqt_hbqevents.cpp
* contrib/hbqt/qtcore/hbqt_hbqslots.cpp
* contrib/hbqt/qtcore/hbqt_init.cpp
* contrib/hbqt/qtcore/hbqt_misc.prg
* contrib/hbqt/qtcore/hbqt_pointer.cpp
* contrib/hbqt/qtcore/qth/HBQEvents.qth
* contrib/hbqt/qtcore/qth/HBQSlots.qth
* contrib/hbqt/qtcore/qth/QLibraryInfo.qth
* contrib/hbqt/qtcore/qth/QMimeData.qth
* contrib/hbqt/qtgui/hbqt_hbqabstractitemmodel.cpp
* contrib/hbqt/qtgui/hbqt_hbqgraphicsitem.cpp
* contrib/hbqt/qtgui/hbqt_hbqgraphicsscene.cpp
* contrib/hbqt/qtgui/hbqt_hbqplaintextedit.cpp
* contrib/hbqt/qtgui/hbqt_hbqtableview.cpp
* contrib/hbqt/qtgui/hbqt_init.cpp
* contrib/hbqt/qtgui/qth/HBQAbstractItemModel.qth
* contrib/hbqt/qtgui/qth/HBQGraphicsItem.qth
* contrib/hbqt/qtgui/qth/HBQGraphicsScene.qth
* contrib/hbqt/qtgui/qth/HBQPlainTextEdit.qth
* contrib/hbqt/qtgui/qth/HBQTableView.qth
* contrib/hbqt/qtgui/qth/QApplication.qth
* contrib/hbqt/qtgui/qth/QClipboard.qth
* contrib/hbqt/qtgui/qth/QDirModel.qth
* contrib/hbqt/qtgui/qth/QDropEvent.qth
* contrib/hbqt/qtgui/qth/QFileSystemModel.qth
* contrib/hbqt/qtgui/qth/QGraphicsSceneDragDropEvent.qth
* contrib/hbqt/qtnetwork/hbqt_init.cpp
* contrib/hbqt/tests/draggable.prg
* contrib/hbxbp/xbpbrowse.prg
* contrib/hbide/ideedit.prg
* contrib/hbide/idefindreplace.prg
+ Implemented: Revamped HBQT.
The complete code base has been cleaned as per new API
and there is no leftover is kept inside the sources.
All demos and hbIDE seems TO be working fine with few
exceptions here and there, but usually all seems fine.
The only issue left, so far I can assess is that no object
is subject to destrution if a connect is issued onto that.
And I have no idea how TO handle this situation. It is the
programmers responsibility to disconnect whenever an object
is supposed to go out of scope.
TODO: still debug tracelog entries ( under HB_TR_DEBUG ) are
retained in hbqt_bind.c which are schduled TO be
removed once code base stabilizes enough.
+ contrib/hbqt/tests/dbfbrowser.prg
+ contrib/hbqt/tests/dbfbrowserclass.prg
+ Added: Table browser demo code by Bacco, thank you.
This code works absolutely flawless both with current and
__HBQT_REVAMP__ protocols.
* contrib/hbqt/hbmk2_qt.hb
! Fixed: some formatting in generated source.
* contrib/hbqt/qtcore/hbqt.h
+ Added: HB_EXPORT PHB_ITEM hbqt_bindSetHbObject( PHB_ITEM pItem, void * qtObject, const char * szClassName, PHBQT_DEL_FUNC pDelFunc, int iFlags );
* contrib/hbqt/qtcore/hbqt_bind.cpp
+ Added: static PHB_DYNS s_dynsym_SETSLOTS = NULL;
Used to initiate __Slots hash to contain code block for slots.
+ Added: PHB_ITEM hbqt_bindSetHbObject( PHB_ITEM pItem, void * qtObject, const char * szClassName, PHBQT_DEL_FUNC pDelFunc, int iFlags )
Used for class inheritance with :new() method.
void hbqt_bindDelSlots( PHB_ITEM pSenderObject );
Used to delete all slots for an object.
+ Added: HB_FUNC( HBQT_PROMOTEWIDGET2 )
Provided by Francesco.
+ Added: HB_FUNC( HBQT_ITEMSONGLOBALLIST )
Provided by Francesco.
- Transferred: some HB_EXPORT prototypes in hbqt.h from hbqt_bind.cpp.
* contrib/hbqt/qtcore/hbqt_hbqslots.cpp
+ Transferred: code blocks exploiting slots to PRG level from C++ level.
* contrib/hbqt/qtcore/hbqt_init.cpp
* contrib/hbqt/qtgui/hbqt_init.cpp
* Formatting.
* contrib/hbqt/qtcore/hbqt_misc.prg
+ Added: VAR __Slots
+ Added: METHOD setSlots()
* contrib/hbqt/qtcore/qth/QObject.qth
+ Uncommented: metaObject() method.
* contrib/hbqt/tests/inherit.prg
+ Added: demo code to exploit inheritance of classes.
This implementation is aimed at transferring signal/slot codeblocks from C++
level to PRG level. This to happen, I am driffting from Przemek's orginal idea
of a separate class, encapsulating the same in exsting superclass HbqtObjectHandeller.
So far has been successful but the concept has introduced a GPF on exit.
* contrib/hbqt/tests/draggable.prg
+ Added: a draggable Harbour Logo without titlebar.
* contrib/hbqt/tests/testres.prg
! Rewritten: per current hbQT implementation.
; Both are supplied by Bacco, thankyou.
* contrib/hbqt/tests/qtrevamp.prg
- Removed: note on returned QAction if it is not of same type
as stated in previous commit. New hbQT is always returning
NIL if a valid Qt pointer could not been obtained. This holds
true for all type of objects. Actually I did not test it
along these lines. So with __HBQT_REVAMP__ there is no need
of :isValidPointer() any more, just test a returned object
against NIL.
* contrib/hbqt/qtcore/hbqt_bind.cpp
* Struggled: to set pointer arithmatic to be used properly.
Also at times Qt returns different type instead of one
defined in documentation, but referred in description part only.
For example, a popup menu returns null instead of QAction if
no item is selected or user pressed escape. This is not
possible to manage such instances in present engine. For
this reason only :isValidPointer() method was introduced.
In REVAMPED hbQT this is different. So to test it, the only
way to ascertain is if the Harbour class name is of same type.
* contrib/hbqt/tests/qtrevamp.prg
+ Added: popup menu option to demonstrate what happens
at the background, if tracelog is activated.
; NOTE: I am not from C or C++ background at all, so some things
may not be as elegant as those should be when viewed from
source point of view. Also trace entries are left in the
code, though under DEBUG mode, which will be cleaned once
code settles down.
* contrib/hbqt/qtcore/hbqt_bind.cpp
+ Implemented: Harbour way of looking at hbQT. QAction has been controlled,
which has facilitated to achieve the desired behavior.
+ contrib/hbqt/tests/qtrevamp.prg
+ Added: demo source to review if __HBQT_REVAMP__ protocol is
upto Harbour expectations. I have added few notes here and there
so that user is aware of some sublities.
; NOTE: please add to qtrevamp.prg any functionality to test all
parts of Qt which may not be confirming to Harbour vision,
and report back on the list.
* contrib/hbqt/qtcore/hbqt_pointer.cpp
+ Added: HB_FUNC( SIGNAL2SLOT ) which simulates exactly Qt's way
of signal/slot mechanism. Now hbQT has yet another way to
mimic signal/slots.
+ contrib/hbqt/tests/signalslots.prg
+ Added: deno program to demonstrate Qt's way of signal/slot
affinity in Harbour.
NOTE: above contribution is made by Francessco Perrilo, thanks.
* contrib/hbqt/hbmk2_qt.hb
! Fix to last fix.
* contrib/hbqt/qtgui/qth/QListWidget.qth
* contrib/hbqt/qtgui/qth/QMainWindow.qth
+ Added: attach/detach/return_not_new tags.
[*D=n;A=n;R=n*]
* contrib/hbqt/tests/demoqt.prg
! On path to prevent memory leaks.
Please investigate the code changes and you will know
how many objects were been around without any mechanism
to free them. You can take help of Qt documentation.
; NOTE: currently demoqt.prg is the test bed for oncoming
GC related changes.
* contrib/hbide/hbqreportsmanager.prg
* contrib/hbide/hbqtoolbar.prg
* contrib/hbide/idebrowse.prg
* contrib/hbide/ideeditor.prg
* contrib/hbide/ideharbourhelp.prg
! On path to preparation for memory leakage.
; WARNING: hbIDE is not at stable stage present.
; I always wondered why memory consumption goes on increasing
with time application written with hbQT. Now I have the answer,
and hopefully in few coming days a great amount of them will be
covered.
* contrib/hbqt/qtcore/hbqt_hbqevents.cpp
* contrib/hbqt/qtcore/hbqt_hbqslots.cpp
* contrib/hbqt/qtcore/hbqt_misc.prg
* Optimizations.
* More VM protections.
* contrib/hbxbp/xbplistbox.prg
! Fixed: a nasty construct in :clear() method where slots were
being disconnected and connected again increasing the unnecessary
object destruction/creation.
* contrib/hbqt/tests/demoqt.prg
* contrib/hbqt/tests/testbrow.prg
! Minor guards.
; NOTE: all the hbQT related test programmes are working flawlessly.
hbIDE still reports "reference to freed block" error but
this is not due to hbQT, instead, the complex object
relationship of hbXBP/hbQT/hbIDE classes. This is only
visible when hbIDE is linked without :destroy() mechanism.
BUT no GPF whatsoever. With :destroy() calls hbIDE EXITS
normally.
* contrib/hbide/ideuisrcmanager.prg
* Completed: minimum required components to materialize some
project based on this protocol.
+ contrib/hbqt/tests/cls_dbstruct.prg
+ contrib/hbqt/tests/dbstruct.prg
+ contrib/hbqt/tests/dbstruct.ui
+ contrib/hbqt/tests/ideui.hbp
+ Added: a small project where cls_dbstruct.prg is created by
hbIDE's UI Src Manager.
; Note: event is define for only first button. Try with other
buttons yourselves and come-out with suggessions.
* contrib/hbqt/tests/browqt.prg
* contrib/hbqt/tests/wvtqt.prg
* Refined: to execute a Qt dialog from within GTWVT application.
This implementation assumes that GTWVT is the main application
window and Qt dialogs are opened in separate thread.
Will look into how the reverse can be achieved.
* contrib/hbqt/qtcore/hbqt_misc.prg
+ Embedded: oQt:hbSetEventBlock( QEvent_Paint, bBlock )
=>
oQt:connect( QEvent_Paint, bBlock )
NOTE: the difference is that in all other events only one
parameter is passed to bBlock, i.e., an object of
type relevant QEvent, whereas, for paintEvent two
parameters are passed, first as usual qPaintEvent,
the other qPainter.
* contrib/hbqt/tests/demoqt.prg
! Replaced: oBtn:hbSetEventBlock( QEvent_Paint, bBlock )
=>
oBtn:connect( QEvent_Paint, bBlock )
* contrib/hbqt/qtgui/qth/QPushButton.qth
* contrib/hbqt/qtgui/qth/QWidget.qth
+ Added: PaintEvent = YES to check if engine is working ok.
Yes, it is working.
* contrib/hbqt/tests/demoqt.prg
+ Added: functionality to paint a PushButton via user code.
Still looking for the how to determine state of the object,
in this case, pushbutton, whether it is being pushed or
otherwise.
* contrib/hbqt/hbmk2_qt.hbs
+ Implemented: engine to receive protected methods, i.e., events
and a way to exploit them. The code needs some more make-over
but the proof of concept is working. Lookfor hbqt/tests/demoqt.prg
for a live example.
NOTE: there is only a minimal need to add few keywords, rest all
is handelled by the engine. The next step is to synchronize
this methodology with :connect() / :disconnect() protocol.
* contrib/hbqt/qtcore/hbqt_pointer.cpp
* contrib/hbqt/qtcore/hbqtcore.hbx
- deleted HBQT_FINDCHILD(). it used raw pointer. was used by hbqtuitools.
% deleted unused extern.
* package/winuni/mpkg_win_uni.bat
* contrib/hbnetio/utils/hbnetioq/hbnetioq.hbp
* contrib/hbqt/hbqt.hbc
* contrib/hbqt/hbqt_all.hbp
- contrib/hbqt/tests/QtUiTools.txt
- contrib/hbqt/qtuitools
- contrib/hbqt/hbqtuitools.hbc
* contrib/hbide/hbide.hbp
- deleted dynamic UI loading. it had too many hopeless TOFIXes
and it required raw pointers. and a lot of bloat.
Should be reimplemented by loading and processing .ui files
manually.
* contrib/hbxbp/xbpfiledialog.prg
* contrib/hbxbp/tests/xbpqtc.prg
* contrib/hbxbp/xbpbrowse.prg
* contrib/hbxbp/xbpgeneric.prg
* contrib/hbqt/gtqtc/gtqtc.cpp
* contrib/hbqt/qtcore/hbqt_pointer.cpp
* contrib/hbqt/qtcore/qth/HBQSlots.qth
* contrib/hbqt/qtcore/qth/HBQEvents.qth
* contrib/hbqt/qtcore/hbqt_hbqslots.cpp
* contrib/hbqt/tests/demoqt.prg
* contrib/hbide/hbqreportsmanager.prg
* contrib/hbide/idemain.prg
* contrib/hbide/ideeditor.prg
! HB_TR_ALWAYS -> HB_TR_DEBUG.
Neverending battle with abused debugging facilities continue.
Please _never_ use HB_TR_ALWAYS in committed code, not even
in commented lines because it's confusing, it trashes console
on *nix systems, it creates a rubbish/unfinished feel for
anyone trying out these components and it creates unnecessary
noise on lists in the form of problem reports.
For those few developers who actually need these lines:
_Set HB_TR_LEVEL on your development machine_ to HB_TR_DEBUG
to get those debug lines!!
Also: Never use HB_TR_ALWAYS as a mean of generating
test output meant for users, I've found some of these in
tests/ .prgs. Use something native instead, like
a text box, msgbox, log file, OutStd(), QOut(), etc.
If this won't change in the future, the best will be to
simply remove HB_TR_ALWAYS from hbtrace.ch, thus making
it unsuited for purposes it wasn't designed for.
Thank you.
* config/win/mingw.mk
+ Added commented -Wextra warning option which will be particularly
interesting with the release of gcc 4.6, since they've finally
implemented the last useful feature which was unique to BCC:
-Wunused-but-set-variable and -Wunused-but-set-parameter
See:
http://gcc.gnu.org/gcc-4.6/changes.html
* contrib/hbqt/hbmk2_qt.hbs
* contrib/hbqt/qtcore/hbqt.h
* contrib/hbqt/qtcore/hbqt_hbqevents.cpp
* contrib/hbqt/qtcore/hbqt_hbqslots.cpp
* contrib/hbqt/qtcore/hbqt_pointer.cpp
* contrib/hbqt/qtcore/qth/HBQEvents.qth
* contrib/hbqt/qtcore/qth/HBQSlots.qth
* contrib/hbqt/qtgui/hbqt_hbqabstractitemmodel.cpp
* contrib/hbqt/qtgui/hbqt_hbqgraphicsitem.cpp
* contrib/hbqt/qtgui/hbqt_hbqgraphicsscene.cpp
* contrib/hbqt/qtgui/hbqt_hbqplaintextedit.cpp
* contrib/hbqt/qtgui/hbqt_hbqtableview.cpp
* contrib/hbqt/qtgui/qth/HBQAbstractItemModel.qth
* contrib/hbqt/qtgui/qth/HBQGraphicsItem.qth
* contrib/hbqt/qtgui/qth/HBQGraphicsScene.qth
* contrib/hbqt/qtgui/qth/HBQPlainTextEdit.qth
* contrib/hbqt/qtgui/qth/HBQTableView.qth
! Fixed: implemented correct GC memory collection. Thanks to Przemek
to have shown me the way.
! Fixed: some code has been changed
PLEASE be sure to do a full "make clean" before compiling
