|
There is a slight discrepancy in your code. You wrote:
m_serial(new QSerialPort(this),
m_TAB_Ext(new A_BT_TAB_DIALOG_EXT)
Meanwhile the compiler error message is:
m_TAB_Ext(new A_BT_TAB_DIALOG_EXT()) (note the extra pair of parenthesis).
The code you posted is potentially correct, assuming m_TAB_Ext is declared as:
A_BT_TAB_DIALOG_EXT *m_TAB_Ext; However the compiler error message seems to indicate that you have those extra parenthesis and, in this case, the compiler interprets it as a function call.
Mircea
|
|
|
|
|
Good catch, but even after removing the () I get same error.
m_serial(new QSerialPort(this),
m_TAB_Ext(new A_BT_TAB_DIALOG_EXT)
|
|
|
|
|
Not sure I understand what you try to do. Maybe the code samle is a bit too short. Where is the `m_TAB_ext` initialization line? I don;t see a declaration for `m_TAB_ext`.
Mircea
|
|
|
|
|
Quote: m_serial(new QSerialPort(this),
// sequence 1` st add HERE
m_TAB_Ext(new A_BT_TAB_DIALOG_EXT)
Which of the overloaded constructors of the QSerialPort class are you trying to call?
Could you please show full details of m_TAB_Ext variable?
"In testa che avete, Signor di Ceprano?"
-- Rigoletto
|
|
|
|
|
SOLVED
THANKS,
you have a very nice , positive , constructive way to lead to solution.
|
|
|
|
|
What is the definition of m_TAB_Ext ? It looks like it, or m_serial , is looking for a function reference/pointer, but you are passing it a pointer to a A_BT_TAB_DIALOG_EXT object.
|
|
|
|
|
May I please inquire what is a Q_OBJECT .
|
|
|
|
|
Here are all parameters definitions (?)
PS
There are no other errors / issues .
So what is
QSerialPort *m_serial = nullptr;
contributing to the issue / error and why is it part of the discussion?
( just asking )
<pre> private:
public:
A_BT_TAB_DIALOG_EXT *m_TAB_Ext = nullptr;
private:
void showStatusMessage(const QString &message);
Ui::MainWindow_Bluetooth *m_ui = nullptr;
QMdiArea *m_mdiarea = nullptr;
QLabel *m_status = nullptr;
Console *m_console = nullptr;
SettingsDialog *m_settings = nullptr;
QSerialPort *m_serial = nullptr;
SettingsDialog_RFcomm *m_settings_RFcomm = nullptr;
|
|
|
|
|
I inquire re/ Q_OBJECT because I do not understand why the compiler does not report an error re/ its usage in the code as written as I assume it is either the name of a type perhaps a class type or an identifier of an object of some type. In either case it would of course be a syntax error exempli gratia as per below:
class cfoo
{
int
public:
void foo();
} Likewise it is a syntax error to write the identifier of an object without a type specifier exempli gratia as per below:
class cfoo
{
xyz
public:
void foo();
} Perhaps Q_OBJECT is a macro hiding these details.
In conclusion I do not see the answer to my inquiry.
|
|
|
|
|
Sorry , I got sidetracked fixing the issue.
I have been told , not just advised, NOT to post Qt questions.
Since I am not welcomed
to do so I , on my own, do not want to discuss Qt here.
I think that is fair.
Cheers
|
|
|
|
|
I am using makefile / make DEFINES as common resource for project.
DEFINES += define PORT_CONFIGURATION
DEFINES += undefine SERIAL_SETUP
DEFINES += define STATUS_BAR
DEFINES += define TRACE_MENU
DEFINES += define TEST_SERIAL_PORT
DEFINES += define SERIAL_PORT_SETUP
DEFINES
# add Apr15
DEFINES += define DEBUG_MDI
DEFINES += define TRACE_CONSOLE
Is there a simple way to ERASE all DEFINES and start fresh?
My code has few places of DEFINES and commenting them out is impractical.
I did ask Mr Google , and the above did not work.
Thanks
.
|
|
|
|
|
|
I have run out of options to resolve this in another forum.
The "conclusion " was - it is a linker error...
I am asking for help to actually correct the error.
There are no other errors posted.
This is ALL (of the errors ) I have to go by.
I will be happy to provide RELEVANT info , barring " what are you trying to do...".
I am deliberately NOT adding more info , for now, I just do not want to
derail / influence your ideas of fixing it.
lib/libQt5SerialPort.so /home/nov25-1/Qt/5.15.2/gcc_64/lib/libQt5Core.so -lGL -lpthread
/usr/bin/ld: /usr/bin/ld: DWARF error: invalid or unhandled FORM value: 0x23
mainwindow.o: in function `MainWindow::MainWindow()':
mainwindow.cpp:(.text+0x270): undefined reference to `ORIGINAL_TERMINAL_MainWindow::ORIGINAL_TERMINAL_MainWindow(QWidget*)'
clang: error: linker command failed with exit code 1 (use -v to see invocation)
make[1]: Leaving directory '/mnt/A_BT_DEC10/BT__PROGRAMS/A_APR9_MAR7_MAR19_CLEAN/A_BT_LIBRARY/mdi'
make[1]: *** [Makefile:293: mdi] Error 1
make: *** [Makefile:569: sub----A_BT_LIBRARY-mdi-make_first] Error 2
13:13:54: The process "/usr/bin/make" exited with code 2.
Error while building/deploying project A_Apr9_Mar7_MAR17_CLEAN (kit: Qt 5.15.2 (gcc_64))
When executing step "Make"
13:13:54: Elapsed time: 00:52.
|
|
|
|
|
Salvatore Terress wrote: I just do not want to
derail / influence your ideas of fixing it. Is this a quiz?
The difficult we do right away...
...the impossible takes slightly longer.
|
|
|
|
|
Look at the error message (again):
mainwindow.cpp:(.text+0x270): undefined reference to `ORIGINAL_TERMINAL_MainWindow::ORIGINAL_TERMINAL_MainWindow(QWidget*)
In your mainwindow.cpp source you are calling the constructor of the ORIGINAL_TERMINAL_MainWindow class. But the linker did not find a definition of that constructor anywhere in the modules and libraries that are being used to build your application. So you need to find out exactly where it is defined and why the linker cannot find it.
|
|
|
|
|
Calling break inside nested for loops breaks all loops or just the inner loop where break is found?
Do I need to do this to exit all loops?
bool breakouter = false;
for(int x=0;x<count;x++)
{
for(int y=0;y<count;y++)
{
breakouter = true;
break;
}
if(breakouter)
break;
}
modified 11-Apr-24 7:46am.
|
|
|
|
|
Only the inner loop.
If you want to break out of more loops you can do something like:
for(int x=0;x<count && !breakouter;x++)
{
for(int y=0;y<count;y++)
{
breakouter = true;
break;
}
}
Mircea
|
|
|
|
|
Probably one of the last valid usecase for goto.
Especially if more than 2 inner loops.
Obviously, if your loops are doing many resources allocations, you'll need to do manual cleaning.
CI/CD = Continuous Impediment/Continuous Despair
|
|
|
|
|
|
Nah, using C++ , that's horribile visu, but we could make an exception for you...
"In testa che avete, Signor di Ceprano?"
-- Rigoletto
|
|
|
|
|
Maximilien wrote: Probably one of the last valid usecase for goto. Especially if more than 2 inner loops. There were lots of valid 'goto' use cases cases, before for loops, while loops, and arbitrarily nested if-else, were invented.
Structured exit from multiply nested loops were invented 40+ years ago. In 1980 vintage CHILL, you could do
outermost: <loop control outermost>
<outermost statements>
intermediate: <loop control intermediate>
<intermediate statements>
inner: <loop control inner>
<statements>
exit intermediate;
<end inner>
<more intermediate statements>
<end intermediate>
<more outermost statements>
<end outermost> exit intermediate; would skip the "more intermediate statements" as well as further intermediate iterations, going directly to "more outermost statements" and continue with further outermost iterations.
If it rather said exit outermost;, the "more outermost statements" would be skipped as well, and the outermost loop would be left.
exit would be a short form of exit innermost;
This construct never made it into mainstream language such as the C derived ones. One partial explanation is differences in label semantics: In CHILL, a label identifies a block, not a point in the code. Furthermore, a statement is a block; it need not be embraced to be labeled.
I would be willing to trade a lot of other 'improvements' in more recent language extensions in favor of something like this.
Some concepts were not well developed in 1980. If a language of today were to introduce such a multi-level exit, it would probably come with something 'finally'-like mechanisms to do required cleanup operations regardless of loop exit method.
Religious freedom is the freedom to say that two plus two make five.
|
|
|
|
|
The proprietary language in which I worked for many years had exactly what you describe. A label was for a BLOCK /ENDBLOCK that you could EXIT . A label could also precede IF , DO , and the equivalents of switch ) without using the BLOCK keyword. The language was designed in the late 70s. This was in the telecom sector, so they probably took the design from CHILL.
|
|
|
|
|
Whenever I have such questions and I have them often I experiment w/ brief simple code and observe the behavior.
|
|
|
|
|
If my preprocessor define / undefine code hierarchy is in this sequence:
#define TEST
#undefine TEST
#define TEST
assuming it is read "top to bottom "
is "TEST" defined?
|
|
|
|
|
Salvatore Terress wrote: is "TEST" defined?
Most certainly
PS. By #undefine I assume you mean #undef . #undefine is not a standard preprocessor keyword
Mircea
|
|
|
|