C / C++ / MFC

[MISC] - Operators - Meaning of operator &

1-Aug-09 16:56

See here.

"Old age is like a bank account. You withdraw later in life what you have deposited along the way." - Unknown

"Fireproof doesn't mean the fire will never come. It means when the fire comes that you will be able to withstand it." - Michael Simmons

Thang29-Jul-09 6:59

Thang

29-Jul-09 6:59

Dear everybody,
I have a question about the meaning of & operator in the following expression:
((WORD)(((DWORD)(l) >> 16) & 0xFFFF))

1) What role does the & operator do here ?
2) Why do programmers often use & with hexadecimal, not integer or long integer or double ?

I'm looking forward to seeing your reply soon.

Best regards,
Jetflower

David Crow29-Jul-09 7:05

29-Jul-09 7:05

Thang wrote:
1) What role does the & operator do here ?

Bitwise AND.

Thang wrote:
2) Why do programmers often use & with hexadecimal, not integer or long integer or double ?

Bits, bytes, and words are easier to visualize in base-16.

enhzflep29-Jul-09 8:27

enhzflep

29-Jul-09 8:27

Okay, lets say that l = 3131961357

This is the same as 0xBAADF00D
Which is the same as 10111010101011011111000000001101

Which means your expression looks a little something like this:

((WORD)(((DWORD)(0xBAADF00D) >> 16) & 0xFFFF))

3131961357 >> 16 = 47789
0xBAADF00D >> 16 = 0x0000BAAD
10111010101011011111000000001101 >> 16 = 00000000000000001011101010101101

47789 & 65535 = 47789
0x0000BAAD & 0x0000FFFF = 0x0000BAAD
00000000000000001011101010101101 & 00000000000000001111111111111111 = 00000000000000001011101010101101

Basically, the code simply takes a 32 bit number and returns the top (most significant) 16 bits of them.
The >> 16 followed by the & 0xFFFF is unnecessary - the bitshift right 16 bits automatically fills the top 16 bits with 0s.

This code is functionally the same:

(WORD)((DWORD)(l) >> 16)

See how much easier things are when we use hex/binary?

Edited && --> &

Maximilien29-Jul-09 9:48

Maximilien

29-Jul-09 9:48

We're lucky that

0XDEADBEEF != 0XBAADFOOD

This signature was proudly tested on animals.

enhzflep29-Jul-09 10:00

enhzflep

29-Jul-09 10:00

Slider control on the status bar (where does WM_HSCROLL go?)

Thang30-Jul-09 6:10

Thang

30-Jul-09 6:10

Dear Mr enhzflep and Mr DavidCrow,
First of all, thank you so much about your clearly explanation.
I just still have a consideration, why the small expression (& 0xFFFF) has been still used in the expression ((l) >> 16) & 0xFFFF) here .
And this expression origins from window library, hix.
Why they still have used this expression if it was not necessary in here ?

Once again, thanks for your help, Mr enhzflep and Mr DavidCrow.

Best regards,
Duong Quoc Thang

BabakTaati29-Jul-09 5:46

BabakTaati

29-Jul-09 5:46

Hi,

I'm trying to place a slider control (CSliderCtrl) on the status bar. The slider bar appears fine and I can move it around (either manually or by calling SetPos). However, I can't find where the WM_HSCROLL messages from the slider bar are sent to! I set the parent of the slider control to the status bar, but no WM_HSCROLL is sent to that window! Any ideas?

Here's what my code looks like. First off, I derived a class from CStatusBar to handle WM_HSCROLL:

class CMyStatusBar : public CStatusBar 
{
public:
	DECLARE_MESSAGE_MAP()
	afx_msg void OnHScroll(UINT nSBCode, UINT nPos, CScrollBar* pScrollBar);
};

In my view class (e.g. CMySimpleProjctView) I added two members:

CSliderCtrl videoPosSlider;
CMyStatusBar statusBar;

and also a function called InitSliderBar() which is something like this:

void CMySimpleProjctView::InitSliderBar()
{
	CMainFrame* frame = (CMainFrame*) AfxGetApp()->m_pMainWnd; // get the main frame
	statusBar = (CMyStatusBar*)(frame->GetStatusBar()); // get the status bar

	// place a slide bar on the status bar, similar to http://support.microsoft.com/kb/142202
	RECT rc;
	statusBar->GetItemRect(0, &rc);
	VERIFY (videoPosSlider.Create(WS_CHILD | WS_VISIBLE, rc, statusBar, 1)); // parent of the slider = status bar

	videoPosSlider.SetRange(0, 50, 1);
	videoPosSlider.SetPos(5);
}

When I run the program, the slider bar appears and I can move it around, but CMyStatusBar::OnHScroll() is not called when I move the slider. To test if everything else is fine, I changed the parent of the slider control to the view window in the code as

VERIFY (videoPosSlider.Create(WS_CHILD | WS_VISIBLE, rc, m_hWnd, 1)); // parent of the slider = view window

and now CMySimpleProjctView::OnHScroll *is* called (but of course the slider bar appears at the wrong place, not on the status bar). I tried Spy++ to see where the WM_HSCROLL is sent to (when parent=status bar) but no WM_HSCROLL is sent at all! In fact no messages are passed when I move the slider around!!

Any ideas? thanks.

-Babak

(by the way, videoPosSlider.GetOwner() and videoPosSlider.GetParent() both return pointers that are identical to statusBar)

Reading from and writing to the registry in windows Vista

Sternocera29-Jul-09 5:00

Sternocera

29-Jul-09 5:00

Hello,

I'm maintaining an MFC application that reads and writes from/to the registry. Previously, it used the HKEY_LOCAL_MACHINE key. However, this meant that Vista users needed full OS administrative privileges, (and perhaps non-admin XP users, but there aren't a whole lot of those around) which isn't desirable. So, I changed the key used to HKEY_CURRENT_USER. This just appears to not work at all on Vista as a regular user, though it does work elsewhere.

How can I store values in the registry without administrative privileges?

Regards,
Sternocera

Re: Reading from and writing to the registry in windows Vista

David Crow29-Jul-09 5:47

Re: Reading from and writing to the registry in windows Vista

29-Jul-09 5:47

Sternocera wrote:
This just appears to not work at all on Vista as a regular user...

Do the registry functions fail?

bob1697229-Jul-09 5:53

bob16972

29-Jul-09 5:53

You need to include a security section in your manifest file to prevent Registry and File Virtualization for your application (or have the user turn User Access Control [UAC] off).

I would recommend the manifest resource file with the security section since it allows you to use registry and file functions without getting false positives from virtualization. I use "AsInvoker" for user apps.

"Windows via C/C++, Fifth Edition" by Richter has some good information on dealing with Vista and using manifests correctly.

The Microsoft endorsed way of properly designing an application for UAC is to use a security manifest section in your manifest file/resource.

Windows Vista Application Development Requirements for User Account Control Compatibility[^]

Create and Embed an application manifest[^]

Certified for Logo Test Case[^]

Re: Reading from and writing to the registry in windows Vista

Sternocera29-Jul-09 23:38

Sternocera

29-Jul-09 23:38

Bob,

Thanks for that, but all I want to do is stored some key:value pair application settings! Surely there is a simpler way?

bob16972 wrote:
You need to include a security section in your manifest file to prevent Registry and File Virtualization for your application (or have the user turn User Access Control [UAC] off).

Why? I thought that registry and file virtualisation where ways in which
Vista supported applications that used the registry.

Regards,
Sternocera

References

EliottA29-Jul-09 4:13

EliottA

29-Jul-09 4:13

Just brushing up my C++ here, and a quick question.

I see people pass by reference using the following methods

&int x;
int & x;
int &x;

Difference? Which is better practice? When declaring a function definition and you want that function to accept a reference, how do you define that function heading?

double myFunction(&int);

?

or do you include the variable name?

double myFunction(int &x);

Sorry to make you go over the basics, my book isn't very clear...

Check out the CodeProject forum Guidelines[^]
The original soapbox 1.0 is back![^]

image processing toolkits | batch image processing

C^hris L^osinger29-Jul-09 4:28

C^hris L^osinger

29-Jul-09 4:28

AFAIK, this: &int x; isn't valid C++.

the other two are identical.

double myFunction(&int);

that's not legal, either.

i always include the variable name, since the header is the first place i look when looking for documentation.

EliottA29-Jul-09 4:35

EliottA

29-Jul-09 4:35

Hey, thanks for the response.

So you include the variable name in the function definition? Seems like the smart thing to do I guess, I'll try to pick up on that habit.

Thanks for the reply. 5'ed.

Check out the CodeProject forum Guidelines[^]
The original soapbox 1.0 is back![^]

image processing toolkits | batch image processing

C^hris L^osinger29-Jul-09 4:40

C^hris L^osinger

29-Jul-09 4:40

EliottA wrote:

So you include the variable name in the function definition?

yeah. it's not strictly required, but it's really handy to people who want to use the function.

Maximilien29-Jul-09 7:59

Maximilien

29-Jul-09 7:59

Chris Losinger wrote:
but it's really handy to people who want to use the function.

especially when you have more than one parameter with the same type (as seen in our code)

void f(int, int, double, double, int, int& );
void f(int a, int b, double aa, double bb, int g, int& r){}

This signature was proudly tested on animals.

TrimRight() and TrimLeft() for unicode characters.

Rajesh R Subramanian29-Jul-09 4:50

Rajesh R Subramanian

29-Jul-09 4:50

Hello,

EliottA wrote:

& int x;
int & x;
int &x;

The first one is plainly illegal.

The second and third have no difference between them. Ideally, I would prefer the third one to avoid any confusion while declaring multiple variables in the same statement. But while you are declaring such a reference variable, you *must* initialize it properly by specifying which variable is it referring to. Unless which, you won't get it to compile.

If a function takes a parameter in by reference (as opposed to 'by value'), any changes made to the variable passed in will reflect on the original variable passed while making the function call (similar to pointers, but the syntax is much cleaner).

This might help you:

void FuncRef(int &n)
{
	n = 0;
}

void FuncNormal(int n)
{
	n = 4;
}

void _tmain(int argc, _TCHAR* argv[])
{
	int i = 10;
	int &Ref = i; //Ref now refers to the variable i. If you change the value of Ref, i will change as well (and vice versa).
        //int &Ref2; //This line, if enabled, won't compile. Because Ref2 is not initialised properly.
	cout<<Ref<<endl; //Will display 10. Because Ref is a reference to i.
	Ref = 2;
	cout<<i<<endl; //Will display 2. Because Ref was changed to 2, but Ref itself is referring to i.
	FuncRef(i);
	cout<<i<<endl; //Will display 0. Because the function takes the parameter as a reference and has modified it!
	FuncNormal(i);
	cout<<i<<endl; //Will still display 0. Because FuncNormal does not take the parameter as reference.
	
}

It is a crappy thing, but it's life -^ Carlo Pallini

Rakesh529-Jul-09 3:40

Rakesh5

29-Jul-09 3:40

Hi,

I have two values say "rakesh,gender" parsed into a cstring object(say obj = "rakesh,gender") from a file..(say notepad)..

In my application, they have used Trimright() and Trimleft() functions for this obj;

I changed my notepad file as unicode (before it was ANSI) now..

After passing those two functions (trimright and trimleft), its showing the obj with one square box at the beginning..("(square box)rakesh,gender");

How to remove the square box??

Can anyone guide me how to solve this issue?

Thanks,
Rakesh

Re: TrimRight() and TrimLeft() for unicode characters.

Emilio Garavaglia30-Jul-09 1:42

Emilio Garavaglia

30-Jul-09 1:42

It seems you trimmed using ANSI functions over a UNICODE string or viceversa

If Trimleft and Trimright are from CString, verify your project has been compiled for UNICODE (see the project properties). Also, verify that no ANSI to UNICODE translations are attempred while reading the file by the fucntions.

To have a check, just do a per-step debug and look how the string is represented in meory just after the reading is finished, with a ASCII only text (saved as UNICODE).

It should be "xx 00 xx 00 xx 00 xx ..... " where xx a two hex digits.

If it is "00 xx 00 xx 00 xx" probably you saved also a BOM or you saved with the wrong endianess (it is an option of the format, during "Save As" - something like UTF-16LE or UTF-16BE: 80x86 use LE format).

If it is "xx xx xx xx xx" it has been read as ANSI.

2 bugs found.
> recompile ...
65534 bugs found.
D'Oh! | :doh:

Font

kumar sanghvi29-Jul-09 3:28

kumar sanghvi

29-Jul-09 3:28

Hi,
im drawing a char inside Crect variable...and if the Font size of the char is increased then how can i increase the rect size based on the change in Font size of char ...

David Crow29-Jul-09 3:55

29-Jul-09 3:55

kumar sanghvi wrote:
...how can i increase the rect size...

Have you looked at the CRect methods such as InflateRect()?

CPallini29-Jul-09 3:57

CPallini

29-Jul-09 3:57

You may use, for instance, GetTextExtentPoint32 [^].
Smile | :)

If the Lord God Almighty had consulted me before embarking upon the Creation, I would have recommended something simpler.
-- Alfonso the Wise, 13th Century King of Castile.

This is going on my arrogant assumptions. You may have a superb reason why I'm completely wrong.
-- Iain Clarke

[My articles]

kumar sanghvi29-Jul-09 4:05

kumar sanghvi

29-Jul-09 4:05

ya..using GetTextExtentPoint32 ill get the width and height of the char..but how can i inflate the rect proportionally...may be should i inflate by the diff by which the char is changed...
Please correct me if im wrong..