|
I would like to know how do draw a Transparent Bitmap using Memory Dc Throught the entire code...And not screen dc at all ...........Then later when the Transparent Bitmap is in the Memory dc ...I am interested in doing this cox i would like to store the transparent BItmpa for late use.....Transfer it to The Screen Dc............
Is It possible to do This..........???
|
|
|
|
|
I think I've seen this before... Please only ask your questions once.
YOu can store your bitmap with a colour key ( presumably it comes this way ), and then draw it when you need to. You can use TransparentBlt if W95 compatibility does not matter to you, if it does I have code at www.wdj.com, Sept 2001, I'm sure a variation on it exists here somewhere as well.
Christian
After all, there's nothing wrong with an elite as long as I'm allowed to be part of it!! - Mike Burston Oct 23, 2001
Sonork ID 100.10002:MeanManOzI live in Bob's HungOut now
|
|
|
|
|
I would like to know how do draw a Transparent Bitmap using Memory Dc Throught the entire code...And not screen dc at all ...........Then later when the Transparent Bitmap is in the Memory dc ...I am interested in doing this cox i would like to store the transparent BItmpa for late use.....Transfer it to The Screen Dc............
Is It possible to do This..........???
|
|
|
|
|
The only way to draw a transparent image is to merge it with the image it is being drawn onto, using masks or similar to fill in the 'transparent' area with the image it is being drawn over.
But until you get to have a 'screenDC' ( i.e. a DC you're drawing to ), you're not going to draw it at all in any case. Until that point, why would you want to keep it in a DC ? ( You can do, of course, as a colour keyed image, but how can you draw it transparently until you have something to draw it onto ? ).
Christian
After all, there's nothing wrong with an elite as long as I'm allowed to be part of it!! - Mike Burston Oct 23, 2001
Sonork ID 100.10002:MeanManOzI live in Bob's HungOut now
|
|
|
|
|
The reason y i want to store it in Memory DC...Is becoz I wnat to Store the Bitmap for later use ...........This was I will not have to go thru the same code and perform the mask again and again evrytime the user wants to display the transparent image ..Hence it speed things up..............Yes later while displaying the image i will Just StretchBLT this Transparent imgae to the Screen DC ............Is there a way i can get my transparent Image in a CBitmap Object ......and Letr can copy it to the Screeen Dc ???
|
|
|
|
|
Is English not your first language ? Your posts are very annoying to try and read.
There is no such *thing* as a transparent bitmap in GDI, only a bitmap with a mask color which will be used to define transparency. Have you done any transparent drawing yet ? If so, how ? Both transparentBlt and the Win95 compatible solutions around the place use a mask color, which your stored bitmap quite plainly can ( and will ) have.
Christian
After all, there's nothing wrong with an elite as long as I'm allowed to be part of it!! - Mike Burston Oct 23, 2001
Sonork ID 100.10002:MeanManOzI live in Bob's HungOut now
|
|
|
|
|
Hi.
Under what situation should one use auto_ptr instead of plain pointer? I am in a situation where I cannot decide on using auto_ptr or pointer. I like auto_ptr because I trust that if something goes wrong at run-time, auto_ptr will help eliminate possible memory leaks. When do you recommend using auto_ptr?
Thanks,
Kuphryn
P.S. auto_ptr declaration cannot be in a switch:
switch (variable)
{
case 1 : auto_ptr<class> temp(new class); // illegal
...
}
|
|
|
|
|
>Under what situation should one use auto_ptr instead of plain pointer?
> [...] I like auto_ptr because I trust that if something goes wrong at
> run-time, auto_ptr will help eliminate possible memory leaks.
You generally use auto_ptr objects when using exception handling, so that if/when an exception is thrown, it will free it's contained pointer (the auto_ptr gets destructed as the stack is unwinding).
A little known/respected fact about auto_ptr implementations: they normally call "delete" on the contained pointer, when they are destructed. But what if the contained pointer should have been deleted with "delete[]" ? Depending on if your RTL/compiler treats "delete" and "delete[]" differently, you may or may not have a problem.
> P.S. auto_ptr declaration cannot be in a switch:
Perhaps your syntax is incorrect. This ia a bad example, but compiles for me (barring warnings):
<br />
switch( DWORD dwValue = 1 )<br />
{<br />
case 1:<br />
{<br />
std::auto_ptr< char > cpChar( new char );<br />
}<br />
}<br />
Peace!
-=- James.
|
|
|
|
|
Okay. Thanks.
I think I figure out the essense of auto_ptr. auto_ptr is a quick and easy exception handling for class objects.
Here is a performance question. I always make sure all classes I use are secure from memory leak. I implement exception handling when dealing with many standard library functions that could return errors, i.e. new. Anyways, is it better to use auto_ptr and let the class handles unexpected run-time errors, or is it better to make sure the class is completely error proof (as much as possible)?
Kuphryn
|
|
|
|
|
i set a breakpoint in the first line of OnDraw Function
but the program won't step into it
one is the other who absorbs the thoughts of others
|
|
|
|
|
I'm just starting to use GDI+...
In the "good old days" I always created a HDC in memory, did all the drawing on it, anf finally BitBlt'ed it to the "real" HDC that actually shows on the screen. This way there was never any flickering when repainting the screen.
Now I have a good question: How do I do that with GDI+?
Do I really have to make a HDC in memory, create a Graphics object that uses my memory-HDC, and then finally BitBlt() from my memdc to my screendc, or is there a smarter way to do it in GDI+?
I really don't want to draw directly on the screen, I need to draw all my graphics in memory, and then copy it all to the screen at once (no flicker).
- Anders
Money talks, but all mine ever says is "Goodbye!"
|
|
|
|
|
You'd associate a graphics object with your screen DC, create another to contain a Bitmap in order to draw on it, and then use DrawImage ( I'm going by memory here, I might be getting names wrong ) to draw that bitmap onto the Graphics object containing your screen DC.
Christian
After all, there's nothing wrong with an elite as long as I'm allowed to be part of it!! - Mike Burston Oct 23, 2001
Sonork ID 100.10002:MeanManOzI live in Bob's HungOut now
|
|
|
|
|
Thanks a lot Christian,
I'l try that  (why didn't I think of that my self  )
- Anders
Money talks, but all mine ever says is "Goodbye!"
|
|
|
|
|
hi to all the great developers out there!
what could be the reason for a program to crash at the last statement i=e, return 0?
the message is something like the instruction at location xxxx tried to reference memory at location xxxx. the memory could not be written.
please help, i'm stuck!
imran.
|
|
|
|
|
Probably something in a destructor that is being called as scope ends.
Christian
After all, there's nothing wrong with an elite as long as I'm allowed to be part of it!! - Mike Burston Oct 23, 2001
Sonork ID 100.10002:MeanManOzI live in Bob's HungOut now
|
|
|
|
|
It's straight C ? If it were me, I'd start commenting things out until I found which bit is causing the problem. How big is the function ? Can you post some code ?
Christian
After all, there's nothing wrong with an elite as long as I'm allowed to be part of it!! - Mike Burston Oct 23, 2001
Sonork ID 100.10002:MeanManOzI live in Bob's HungOut now
|
|
|
|
|
You're *sure* it's not a stack overflow then ?
Christian
After all, there's nothing wrong with an elite as long as I'm allowed to be part of it!! - Mike Burston Oct 23, 2001
Sonork ID 100.10002:MeanManOzI live in Bob's HungOut now
|
|
|
|
|
If you can localize the problem to a specific function (by selectively commenting out other stuff), go ahead and post the code (and the sample data you feed it). It should be relatively easy to spot the problem.
/ravi
"There is always one more bug..."
ravib@ravib.com
http://www.ravib.com
|
|
|
|
|
// ahum, sorry!
/*
ok ravi, YOU asked for it!!! nah, looks scary but ain't. just look at the function tryRule, or, the one that calls it, readStrings. rest is just file reading and formatting stuff. i'm sure you'll see the cause of the problem.
************************************
these four lines have to be placed as a text file, say grammar.txt
S -> Saabb | ccc | $
S -> kTk
T -> pppter
T -> pppalter
*************************************
these two lines may be put in another text file, say strings.txt
gyrwweweweerrw
kpppalterkaabb
******************************************
this is the code. command line parameters are
1. the grammar file,
2. the strings file
3. any file for getting the results.
******************************************
*/
#include <iostream.h>
#include <fstream.h>
#include <string.h>
#include <stdlib.h>
int countGrammarLines(char*);
void readGrammar(int, char**, char*);
void readStrings(char**, char*, char*);
bool tryRule (int, char**, char *, char* &);
int countGrammarLines(char* grammarFile)
{
int count=0;
char line[100];
char *token;
token = new (char) (100);
ifstream fin(grammarFile, ios::nocreate);
if (!fin) {
cout << grammarFile << ": File not found" << endl
<< "Exiting the application!" << endl;
exit(0);
}
while (!fin.eof()) {
fin.getline(line, 100, '\n');
count++;
token = strtok(line, "|");
while (token != NULL) {
count++;
token = strtok(NULL, "|");
}
--count;
}
fin.close();
return count;
}
void readGrammar(int count, char** rules, char* grammarFile)
{
char line[100];
char *token;
ifstream fin(grammarFile, ios::nocreate);
if (!fin) {
cout << grammarFile << ": File not found" << endl
<< "Exiting the application!" << endl;
exit(0);
}
for (int i=0; i<count; i++) {
fin.getline(line, 100, '\n');
token = strtok(line, "|");
*(rules[i]) = *token;
strcpy (rules[i]+1, token+5);
if (*(rules[i] + strlen(rules[i])-1) == ' ')
*(rules[i] + strlen(rules[i])-1) = '\0';
token = strtok(NULL, "|");
while (token != NULL) {
i++;
*(rules[i]) = line[0];
strcpy (rules[i]+1, token+1);
if (*(rules[i] + strlen(rules[i])-1) == ' ')
*(rules[i] + strlen(rules[i])-1) = '\0';
token = strtok(NULL, "|");
}
}
fin.close();
}
void readStrings(int count, char** rules, char *stringFile, char *resultFile)
{
char *line = new (char) (100);
int strCount = 1;
bool result = false;
ifstream fin(stringFile, ios::nocreate);
ofstream fout(resultFile);
if (!(fin)) {
cout << stringFile << " : File not found" << endl
<< "Exiting the application!" << endl;
exit(0);
}
while (!fin.eof()) {
fin.getline(line, 100, '\n');
for (int i=0; i<count; i++) {
fout << strCount++ << ". ";
if (*rules[i] == 'S') {
if (tryRule (count, rules, (rules[i]+1), line)) {
result = true;
break;
}
}
}
fout << (result ? "accepted" : "rejected") << endl;
cout << (result ? "accepted" : "rejected") << endl;
}
fin.close();
fout.close();
}
bool tryRule (int count, char** rules, char *rule, char* &str)
{
char *temp;
char *newrule = new (char) (2 * strlen(str));
bool result = false;
while (*rule == *str && *rule != 0 && *str != 0) {
rule++;
str++;
}
temp = str;
if (*rule == 0 && *str == 0)
return true;
else if (*rule == '$')
return tryRule (count, rules, ++rule, str);
else if (*rule >= 'A' && *rule <= 'Z') {
for (int i=0; i<count; i++) {
if (*rules[i] == *rule && (strlen(rules[i]) + strlen(rule)) <= 3 + strlen(str)) {
strcpy (newrule, rules[i]+1);
strcat (newrule, rule+1);
result = result || tryRule (count, rules, newrule, str);
if (result)
break;
else
str = temp;
}
else
result = false;
}
}
return result;
}
int main(int argc, char *argv[])
{
char **rules;
int count;
count = countGrammarLines(argv[1]);
rules = new (char*[count]);
for (int i=0; i<count; i++)
rules[i] = new (char) (100);
readGrammar(count, rules, argv[1]);
readStrings(count, rules, argv[2], argv[3]);
return 0;
}
|
|
|
|
|
hhh wrote:
ok ravi, YOU asked for it!!!
One request - can you repost (or just modify your message) and check the "Display the message as-is (no HTML)" box? That will include the #includes and will show the indenting clearer. Thanks!
/ravi
"There is always one more bug..."
ravib@ravib.com
http://www.ravib.com
|
|
|
|
|
|
When you post an #include, if it uses the < and > symbols, the contents of the include statment are lost unless you check 'display as is'. Obviousl you have such an include and because you didn't check it, Ravi can't tell what you're including.
Christian
After all, there's nothing wrong with an elite as long as I'm allowed to be part of it!! - Mike Burston Oct 23, 2001
Sonork ID 100.10002:MeanManOzI live in Bob's HungOut now
|
|
|
|
|
I'm having a hell of a time replying here, I think you must be deleting posts ?
Anyhow, you said
Tim
i'm afraid i'm correct there, i AM getting memory allocated equal to 100 * sizeof(char) bytes somewhere. your suggested statement is considered illegal by some compilers - i'm using MS VC++
and I am trying to reply with this:
Tim is right, under VC++ the statement char * pChar = new char [100] allocates an array of 100 chars, and the statement char * pChar = new (char) (100) will ignore the first set of brackets and create room for one char, initialised to 100. I don't know what strange settings could cause VC to barf when asked to allocate an array, or why you think your statement creates an array, it quite plainly does not.
I'm guessing you've realised you were wrong ? In any case, char * pChar = new char[100]; is so valid that any compiler that rejects it should be burned at the stake.
Christian
After all, there's nothing wrong with an elite as long as I'm allowed to be part of it!! - Mike Burston Oct 23, 2001
Sonork ID 100.10002:MeanManOzI live in Bob's HungOut now
|
|
|
|
|
char *text = new (char) (100);
What are you trying to do here?
1. Allocate 100 characters?
2. Allocate 1 character initialized to 100?
The code does #2. If you want to do #1:
char *text = new char [100];
Tim Smith
Descartes Systems Sciences, Inc.
|
|
|
|
|
LOL, well ok.
Don't say I didn't try to help.
You might want to actually try to debug you code and see exactly what "char *array = new (char) (count)" actually does. Which is what I did. It equates to the code:
char *array = new char [1];
array [0] = count;
new operator
Tim Smith
Descartes Systems Sciences, Inc.
|
|
|
|
|
General 
News 
Suggestion 
Question 
Bug 
Answer 
Joke 
Praise 
Rant 
Admin 
Submit an article or tip
