C / C++ / MFC

Randor 7-Jun-10 20:44

Randor

7-Jun-10 20:44

Hi,

First you need to call FindResource[^] to find the resource string then you can call SizeofResource[^] to get the size of the string so you can resize your std::string and finally call LoadString[^] to copy it.

Best Wishes,
-David Delaune

Randor 7-Jun-10 21:21

Randor

7-Jun-10 21:21

Hi,

Not sure who gave me a one vote... but the solution I presented is a secure and safe method for loading a std::string from a resource.

The method presented by Aescleal will fail for a string greater than the 4096 constant. The wstring solution presented by Stephen Hewitt just flat out doesn't work for std::string

Best Wishes,
-David Delaune

Stephen Hewitt7-Jun-10 21:28

7-Jun-10 21:28

It does for a wstring (which was all it was intended to, it wasn't a reply to the OP but to Aescleal).

It's documented. MSDN (on LoadString):

nBufferMax [in]
int

The size of the buffer, in characters. The string is truncated and null-terminated if it is longer than the number of characters specified. If this parameter is 0, then lpBuffer receives a read-only pointer to the resource itself.

Secondly, I tried it (after reading the documentation thoroughly first) and verified that it did work.

Finally, your technique will not work. LoadResource returns nothing that LoadString can consume.

Steve

Randor 7-Jun-10 22:34

Randor

7-Jun-10 22:34

Stephen Hewitt wrote:
Finally, your technique will not work. LoadResource returns nothing that LoadString can consume.

Hi Stephen,

You may be correct. I was thinking that you could just do something like this:

HRSRC hres = FindResource(NULL, MAKEINTRESOURCE(IDS_SOMESTRING), RT_STRING);
DWORD dwSizeRes = SizeofResource(NULL,hres);
char *p = NULL;
int result = LoadString((HINSTANCE)&__ImageBase,AFX_IDS_APP_TITLE,p,0);
std::string s(p,dwSizeRes);

Which is not much different than what you posted.

With the method that you posted...I am thinking that the std::string would need a NULL terminator in the string resource. I admit that I have not tried either method (including the one I just typed above). I will test both methods later when I get some free time.

Best Wishes,
-David Delaune

Randor 8-Jun-10 0:29

Randor

8-Jun-10 0:29

Hey Steve

I just spent 30 minutes testing both of our code samples. I could not get the code you wrote working for std::string probably because the resource strings are actually stored in Unicode within the PE files. However for wstrings your code is absolutely perfect. I suspect... that the brilliant code you posted could be actually be modified to utilize wcstombs and work for both ANSI and Unicode builds.

As for my code sample... I had even less success although I was eventually able to get it working. I had to re-read some of the MSDN docs and ended up using FindResourceEx with the language identifier. To make things worse the address I was recieving was the start of the string resource block. I had to load the string block into an HGLOBAL and walk the string table before I could even read the string. D'Oh! | :doh:

Anyway I just wanted to let you know that you were correct. Smile | :)

Best Wishes,
-David Delaune

Stephen Hewitt8-Jun-10 14:16

8-Jun-10 14:16

Thanks.

Steve

Aescleal7-Jun-10 21:36

7-Jun-10 21:36

I didn't give you a 1 vote but the method I outlined will work for strings greater than 4096 characters, it just won't load the whole string. If you're really worried about that you have to use the OS/2 vintage "try it once to fail, once to succeed" method of sizing buffers.

And it'd be fairly easy to change Stephen's method to work for std::string - depending on how much you know about the resource and the characters in it it's either trivial (use the two iterator string constructor) or slightly harder (use wcstombs).

Ash

Aescleal7-Jun-10 20:46

7-Jun-10 20:46

Load it into a character array first then assign the result to the std::string.

std::string load_string_from_resource( HINSTANCE app_instance, unsigned string_ID )
{
    char buffer[ 4096 ];
    unsigned bytes_copied = LoadString( app_instance, string_ID, buffer, sizeof( buffer ) );
    if( !bytes_copied )
        throw std::runtime_error( "Resource not found!" );

    return std::string( buffer, bytes_copied );

}

(Code not tested, Errors and Omissions Excluded)

Cheers,

Ash

Stephen Hewitt7-Jun-10 20:52

7-Jun-10 20:52

If you're using Unicode it can be done more efficiently than that:

std::wstring LoadString(HINSTANCE hInstance, UINT uID)
{
	LPCWSTR pString;
	int res = LoadStringW(hInstance, uID, (LPWSTR)(&pString), 0);
	if (res == 0)
		return std::wstring(); // No such string.
	return std::wstring(pString, res);
}

MSDN on LoadString[^]:

If this parameter is 0, then lpBuffer receives a read-only pointer to the resource itself.

NOTE: Raw string resources are not NULL terminated.

Steve

Aescleal7-Jun-10 21:14

7-Jun-10 21:14

Cool, I'd either forgotten about or never known the behaviour of LoadString when you pass zero as the buffer size.

What's the reason this wouldn't work on ANSI as well?

Stephen Hewitt7-Jun-10 21:17

7-Jun-10 21:17

I'm not positive, but it doesn't seem to. Note that the documentation says it returns a pointer to the raw resource and string resources are always Unicode, so perhaps this is a reason. In fact, in the ANSI case 0xFFFFffff was returned to indicate an error (when 0 is passed as the buffer size), which isn't mentioned in the documentation.

Steve

Aescleal7-Jun-10 21:30

7-Jun-10 21:30

Thanks for that, I'll stick that on the list as the first new thing I've learnt today.

Ash

Alain Rist8-Jun-10 7:22

Alain Rist

8-Jun-10 7:22

See my answer including Stephen Hewitt contribution at http://www.codeproject.com/Tips/86496/Load-a-Windows-string-resource-into-a-std-string-o.aspx[^]
cheers,
AR

When the wise (person) points at the moon the fool looks at the finger (Chinese proverb)

gmallax8-Jul-10 1:18

gmallax

8-Jul-10 1:18

Very nice and useful..Thanx Smile | :)

Alain Rist8-Jul-10 3:42

Alain Rist

8-Jul-10 3:42

gmallax wrote:
Very nice and useful

Then please upvote. Thanks, AR

When the wise (person) points at the moon the fool looks at the finger (Chinese proverb)

gmallax9-Jul-10 0:42

gmallax

9-Jul-10 0:42

hi,

here we declare char buffer as char buffer[ 1024 ] = { '\0' }..

How to release this?

fatal error C1001: INTERNAL COMPILER ERROR

Alain Rist9-Jul-10 2:42

Alain Rist

9-Jul-10 2:42

It is released at the next }

ReturnType SomeClass::SomeFunction(SomeParams)
{
    char buffer[1024] = {0}; // this a stack allocation
    // any code ...
    return ReturnType::SomeValue;
} // buffer is deallocated here

cheers,
AR

When the wise (person) points at the moon the fool looks at the finger (Chinese proverb)

archana.srikar7-Jun-10 20:04

archana.srikar

7-Jun-10 20:04

Hi..
can u please help me to solve this error... i'm getting the following error

"fatal error C1001: INTERNAL COMPILER ERROR
(compiler file 'msc1.cpp', line 1786) "

I'm using vc6.0 and in my project settings options - there are no such options like /Og , /Oa .. but still i'm getting the error.

can anyone suggest me some solution asap..

Thanks in advance...
Archana

Re: fatal error C1001: INTERNAL COMPILER ERROR

Niklas L7-Jun-10 20:34

Niklas L

7-Jun-10 20:34

I had a few of those. I think I rearranged some of the code to get it to work. But it was a long time ago. Use #error directive at the top of your source file that fails, move it downwards and see where it breaks. Then, try rearranging your code at that point.

home

Re: fatal error C1001: INTERNAL COMPILER ERROR

arun_pk7-Jun-10 20:40

arun_pk

7-Jun-10 20:40

try rebuild/clean the solution once before running

Unable to print micron character µ (mu) in japanese OS

arun_pk7-Jun-10 19:38

arun_pk

7-Jun-10 19:38

I have a vc++ application that has been showing odd characters on Japanese Windows installations.

It shows up an japanese katakana letter O (U + 30AA) instead of the micron character µ (mu)

i am using
sprintf(Buf, "%d µV", Values[i]);

My very simple task is to embed or force the "µ" symbol to remain as-is despite language settings.
Surely there is an easy method that I am missing somewhere!? Does anyone know an answer?

Much appreciation for any help.

Re: Unable to print micron character µ (mu) in japanese OS

GDavy7-Jun-10 21:17

GDavy

7-Jun-10 21:17

Don't use ANSI strings, use UNICODE strings, ANSI strings will use the local Codepage for character conversions wich causes your problem.

Windows message

Pryabu7-Jun-10 19:27

Pryabu

7-Jun-10 19:27

Hi,

What windows message will be send to the application while closing the applcation[End Process] via task manager?

Thanks,

Re: Windows message

Stephen Hewitt7-Jun-10 19:37