BItBlt issue when using enhanced meta file DC as source DC - mfc

I am working on an application which is using emf for buffer drawing. I am trying to save this emf to a bitmap image file using BitBlt. But no drawings are saved to bitmap. I know I can use PlayEnhMetaFile() but I have to use BitBlt or GDI/GDI+ calls for this. As there will be some other drawing calls to emf after saving it to bitmap.
Sample code.
void CTestGUIApplicationView::OnDraw(CDC* pDC)
{
CRect oRect(0, 0, 640, 434);
HDC hdc = pDC->GetSafeHdc();
//Meta file creation
int iWidthMM = GetDeviceCaps(hdc, HORZSIZE);
int iHeightMM = GetDeviceCaps(hdc, VERTSIZE);
int iWidthPels = GetDeviceCaps(hdc, HORZRES);
int iHeightPels = GetDeviceCaps(hdc, VERTRES);
CRect rect;
rect.left = (oRect.left * iWidthMM * 100) / iWidthPels;
rect.top = (oRect.top * iHeightMM * 100) / iHeightPels;
rect.right = (oRect.right * iWidthMM * 100) / iWidthPels;
rect.bottom = (oRect.bottom * iHeightMM * 100) / iHeightPels;
HDC hMetaDC = CreateEnhMetaFile(hdc, NULL, &rect, NULL);
CDC *pMetaDC = CDC::FromHandle(hMetaDC);
//Drawing on meta file DC.
RECT drawingRect;
CPen penBlue, pen2, pen3, pen4;
penBlue.CreatePen(PS_SOLID | PS_COSMETIC, 1, RGB(0, 0, 255));
pen2.CreatePen(PS_SOLID | PS_COSMETIC, 1, RGB(0, 255, 255));
pen3.CreatePen(PS_SOLID | PS_COSMETIC, 1, RGB(255, 0, 255));
pen4.CreatePen(PS_SOLID | PS_COSMETIC, 1, RGB(128, 0, 56));
auto pOldPen = pMetaDC->SelectObject(&penBlue);
pMetaDC->Arc(oRect,
CPoint(oRect.right, oRect.CenterPoint().y),
CPoint(oRect.CenterPoint().x, oRect.right));
pMetaDC->SelectObject(&pen2);
pMetaDC->Ellipse(oRect.left + 50, oRect.top + 25, oRect.Width(), oRect.Height());
//copy meta file to DC.
CopyToBitMap(_T("StateImage_EMFDC.bmp"), pMetaDC, oRect);
//some other drawing calls on meta file DC.
pMetaDC->SelectObject(&pen3);
pMetaDC->Ellipse(oRect.left + oRect.Width() / 2, oRect.top + oRect.Height() / 2, oRect.Width(), oRect.Height());
pMetaDC->SelectObject(pOldPen); pOldPen = NULL;
//Copy meta file to window DC.
HENHMETAFILE hMeta = CloseEnhMetaFile(pMetaDC->GetSafeHdc());
if (hMeta != NULL)
{
//save meta file to disk to view its contents.
HENHMETAFILE hMeta2 = CopyEnhMetaFile(hMeta, _T("test.emf"));
DeleteEnhMetaFile(hMeta2);
PlayEnhMetaFile(hdc, hMeta, oRect);
DeleteEnhMetaFile(hMeta);
}
//some other drawing on Window DC.
pOldPen = pDC->SelectObject(&pen4);
pDC->Ellipse(oRect.left + 25, oRect.top + 50, oRect.Width() - 30, oRect.Height() - 60);
pDC->SelectObject(pOldPen);
CopyToBitMap(_T("StateImage_WindowDC.bmp"), pDC, oRect);
}
//Routine to copy data from DC to bitmap.
void CTestGUIApplicationView::CopyToBitMap(CString filePath, CDC* pDC, CRect & windRect)
{
//creating bitmap
CBitmap bitmap;
bitmap.CreateCompatibleBitmap(pDC, windRect.Width(), windRect.Height());
CDC memDC;
memDC.CreateCompatibleDC(pDC);
CBitmap *oldBMP = memDC.SelectObject(&bitmap);
memDC.FillSolidRect(windRect, RGB(0, 255, 0));
BOOL result = BitBlt(memDC.GetSafeHdc(), 0, 0, windRect.Width() , windRect.Height(), pDC->GetSafeHdc(), 0, 0, SRCCOPY);
memDC.SelectObject(oldBMP);
//Saving bitmap to disk.
CImage image;
image.Attach(bitmap);
image.Save(filePath, Gdiplus::ImageFormatBMP);
image.Detach();
image.Destroy();
bitmap.DeleteObject();
}
"StateImage_WindowDC.bmp" have complete drawing. while "StateImage_EMFDC.bmp" is complete green image which should have drawing on it.

A metafile DC isn't like a regular DC, it doesn't draw on a bitmap; it's only used for recording drawing instructions. From Microsoft's "Enhanced Metafile Creation" page:
When CreateEnhMetaFile succeeds, it returns a handle that identifies a special metafile device context. A metafile device context is unique in that it is associated with a file rather than with an output device. When the system processes a GDI function that received a handle to a metafile device context, it converts the GDI function into an enhanced-metafile record and appends the record to the end of the enhanced metafile.
You say that you can't call PlayEnhMetaFile because you have additional drawing commands you want to append to the metafile after you capture it. In that case I can see two options:
Draw to a regular DC in parallel with the metafile DC.
Use GetEnhMetaFileBits and SetEnhMetaFileBits to make a copy of the metafile that you can use with PlayEnhMetaFile.

Related

Anti aliasing in MFC

I'm trying to implement anti-aliasing in my MFC app, I'm using the technique described in this tutorial.
Create a bitmap (2x, 4x, 8x) the size of the original bitmap.
Draw on the resized bitmap (I'm only using simple figures (lines, circles and etc)).
Set StretchBlt Mode to HalfTone.
And Resize with StretchBlt to the original size.
Using this way, drawing in the resized bitmap it works, but I want to create a more generic function that receives a bitmap with the drawing already made and return with the anti-aliasing, I tried this:
static HBITMAP AntiAliasing(HBITMAP hBitmap)
{
int escala = 4;
HBITMAP bmp = __copia(hBitmap); // Copy the bitmap.
HDC hMemDC = CreateCompatibleDC(NULL);
HBITMAP bmpAntigo1 = (HBITMAP)::SelectObject(hMemDC, bmp);
BITMAP bitmap;
::GetObject(hBitmap, sizeof(BITMAP), &bitmap);
// Create a bitmap (2x, 4x, 8x) the size of the original bitmap.
HDC hDCDimensionado = ::CreateCompatibleDC(hMemDC);
HBITMAP bmpDimensionado = ::CreateCompatibleBitmap(hDCDimensionado,
bitmap.bmWidth * escala,
bitmap.bmHeight * escala);
HBITMAP hBmpVelho = (HBITMAP)::SelectObject(hDCDimensionado, bmpDimensionado);
// I also tried with {BLACKONWHITE, HALFTONE, WHITEONBLACK}
int oldStretchBltMode2 = ::SetStretchBltMode(hDCDimensionado, COLORONCOLOR);
// Resize the bitmap to the new size.
::StretchBlt(hDCDimensionado,
0, 0, bitmap.bmWidth * escala, bitmap.bmHeight * escala,
hMemDC,
0, 0, bitmap.bmWidth, bitmap.bmHeight,
SRCCOPY);
/*
* Here the bitmap has lost his colors and became black and white.
*/
::SetStretchBltMode(hDCDimensionado, oldStretchBltMode2);
// Set StretchBltMode to halfTone so can mimic the anti aliasing effect.
int oldStretchBltMode = ::SetStretchBltMode(hMemDC, HALFTONE);
// resize to the original size.
::StretchBlt(hMemDC,
0, 0, bitmap.bmWidth, bitmap.bmHeight,
hDCDimensionado,
0, 0, escala * bitmap.bmWidth, escala * bitmap.bmHeight,
SRCCOPY);
::SetStretchBltMode(hMemDC, oldStretchBltMode);
::SelectObject(hMemDC, bmpAntigo1);
::DeleteDC(hMemDC);
::SelectObject(hDCDimensionado, hBmpVelho);
DeleteDC(hDCDimensionado);
return bmp;
}
But this function doesn't work, the result loses its colors (all drawings became black) and there isn't anti aliasing.
Any help will be appreciated!
From documentation for CreateCompatibleBitmap:
Note: When a memory device context is created, it initially has a
1-by-1 monochrome bitmap selected into it. If this memory device
context is used in CreateCompatibleBitmap, the bitmap that is created
is a monochrome bitmap. To create a color bitmap, use the HDC that was
used to create the memory device context, as shown in the following
code:
Change the code and supply hdc for the desktop as show below:
HDC hdc = ::GetDC(0);
HBITMAP bmpDimensionado = ::CreateCompatibleBitmap(hdc, ...)
::ReleaseDC(0, hdc);
This will show the image, however this method will not produce the desired effect because it simply magnifies each pixel to larger size and reduces it back to the original pixel. There is no blending with neighboring pixels.
Use other methods such Direct2D with Gaussian blur effect, or use GDI+ instead with interpolation mode:
Gdiplus::GdiplusStartup...
void foo(HDC hdc)
{
Gdiplus::Bitmap bitmap(L"file.bmp");
if(bitmap.GetLastStatus() != 0)
return 0;
auto w = bitmap.GetWidth();
auto h = bitmap.GetHeight();
auto maxw = w * 2;
auto maxh = h * 2;
Gdiplus::Bitmap membmp(maxw, maxh);
Gdiplus::Graphics memgr(&membmp);
memgr.SetInterpolationMode(Gdiplus::InterpolationModeHighQualityBilinear);
memgr.DrawImage(&bitmap, 0, 0, maxw, maxh);
Gdiplus::Graphics gr(hdc);
gr.SetInterpolationMode(Gdiplus::InterpolationModeHighQualityBilinear);
gr.DrawImage(&membmp, 0, 0, w, h);
}
If target window is at least Vista, use GDI+ version 1.1 with blur effect. See also How to turn on GDI+ 1.1 in MFC project
#define GDIPVER 0x0110 //add this to precompiled header file
void blur(HDC hdc)
{
Gdiplus::Graphics graphics(hdc);
Gdiplus::Bitmap bitmap(L"file.bmp");
if(bitmap.GetLastStatus() != 0)
return;
Gdiplus::Blur blur;
Gdiplus::BlurParams blur_param;
blur_param.radius = 3; //change the radius for different result
blur_param.expandEdge = TRUE;
blur.SetParameters(&blur_param);
bitmap.ApplyEffect(&blur, NULL);
graphics.DrawImage(&bitmap, 0, 0);
}

How to maintain transparency when writing text on a Windows System Tray icon from unmanaged C program

I have a MFC C++ (unmanaged) Windows application that uses a “standard” icon in the System Tray. This icon was created & edited using Visual Studio and is 32x32 pixels with only 4bit colour (according to VS's Resource Editor).
With Visual Studio, I also set a transparent background (shown as white in the “before” image).
I wish to dynamically change the icon by writing 2 digits (1-99) on top of it.
Using the code below (based on that in this question: How to draw text with transparency using GDI?) to superimpose “55” in yellow on the icon, it works except that the transparency disappears (it appears black in the “after” image and on the System Tray). My actual code differs very very slightly in that the font size (20), font name (Courier New), text colour (yellow - RGB(255, 255, 0)) and the numeric value (55) are run-time variables rather than fixed values.
Any suggestions on how to make the background remain transparent as far as the System Tray is concerned gratefully received.
These images have been captured using MS’s Snipping tool with the image open in MS Paint as a 32x32 icon wouldn't be very visible as-is.
Before Image:
After image:
Code:
void CreateNewIcon(HICON &hNewIcon, HICON hBackgroundIcon)
{
::DestroyIcon(hNewIcon);
// First create font
LOGFONT lf = { 0 };
lf.lfHeight = -20;
lf.lfWeight = FW_BOLD;
lf.lfOutPrecision = OUT_TT_PRECIS;
lf.lfQuality = CLEARTYPE_QUALITY;
wmemset(lf.lfFaceName, 0, LF_FACESIZE);
lstrcpy(lf.lfFaceName, L"Courier New");
HFONT hFont = ::CreateFontIndirect(&lf);
ICONINFO ii = { 0 };
::GetIconInfo(hBackgroundIcon, &ii);
BITMAP bm = { 0 };
::GetObject(ii.hbmColor, sizeof(bm), &bm);
SIZE szBmp = { bm.bmWidth, bm.bmHeight };
HDC hDc = ::GetDC(NULL);
HDC hMemDC = ::CreateCompatibleDC(hDc);
HGDIOBJ hOldBmp = ::SelectObject(hMemDC, ii.hbmColor);
HGDIOBJ hOldFont = ::SelectObject(hMemDC, hFont);
::SetBkMode(hMemDC, TRANSPARENT);
::SetTextColor(hMemDC, RGB(255, 255, 0));
::TextOut(hMemDC, 0, 8, L"55", 2);
::SelectObject(hMemDC, hOldFont);
::SelectObject(hMemDC, hOldBmp);
// We need a simple mask bitmap for the icon
HBITMAP hBmpMsk = ::CreateBitmap(szBmp.cx, szBmp.cy, 1, 1, NULL);
ICONINFO ii2 = { 0 };
ii2.fIcon = TRUE;
ii2.hbmColor = ii.hbmColor;
ii2.hbmMask = hBmpMsk;
// Create updated icon
hNewIcon = ::CreateIconIndirect(&ii2);
// Cleanup
::DeleteObject(hBmpMsk);
::DeleteDC(hMemDC);
::ReleaseDC(NULL, hDc);
::DeleteObject(ii.hbmColor);
::DeleteObject(ii.hbmMask);
::DeleteObject(hFont);
}
There are multiple issues with our code:
You are trying to draw cleartype-quality text over transparent icon part. But cleartype font rendering must be performed over opaque background because it needs to inspect background color. So you should switch to Anitialiased quality (or to not antialiased quality) or provide opaque background for your text.
When you create hBmpMsk you skip it's content initialization by supplying NULL for bits pointer, so resulting icon will actually have completely random transparency. You need to fill this mask bitmat appropriately.
Also you probably need to switch to higher bit depth because monochrome bitmap mask can't handle semitransparent parts of antialiased text.
update
I think you should draw cleartype text but with opaque background, then get text rectangle using something like GetTextExtentPoint32, then copy data from the original bitmap mask into hBmpMsk and then finally fill white (text) rectangle on it so the new icon will preserve transparency from original and has opaque text block.
Thanks for all the help from VTT without which I wouldn't have been able to get this far. This appears to work for me.
void CreateNewIcon(HICON &hNewIcon, HICON hBackgroundIcon)
{
::DestroyIcon(hNewIcon);
HDC hDc = ::GetDC(NULL);
HDC hMemDC = ::CreateCompatibleDC(hDc);
// Load up background icon
ICONINFO ii = { 0 };
::GetIconInfo(hBackgroundIcon, &ii);
HGDIOBJ hOldBmp = ::SelectObject(hMemDC, ii.hbmColor);
// Create font
LOGFONT lf = { 0 };
lf.lfHeight = -20;
lf.lfWeight = FW_BOLD;
lf.lfOutPrecision = OUT_TT_PRECIS;
lf.lfQuality = ANTIALIASED_QUALITY;
wmemset(lf.lfFaceName, 0, LF_FACESIZE);
lstrcpy(lf.lfFaceName, L"Courier New");
HFONT hFont = ::CreateFontIndirect(&lf);
HGDIOBJ hOldFont = ::SelectObject(hMemDC, hFont);
// Write text
::SetBkMode(hMemDC, TRANSPARENT);
::SetTextColor(hMemDC, RGB(255, 255, 0));
::TextOut(hMemDC, 0, 8, L"55", 2);
// Set up mask
HDC hMaskDC = ::CreateCompatibleDC(hDc);
HGDIOBJ hOldMaskBmp = ::SelectObject(hMaskDC, ii.hbmMask);
// Also write text on here
HGDIOBJ hOldMaskFont = ::SelectObject(hMaskDC, hFont);
::SetBkMode(hMaskDC, TRANSPARENT);
::SetTextColor(hMaskDC, RGB(255, 255, 0));
::TextOut(hMaskDC, 0, 8, L"55", 2);
// Get handle to create mask bitmap
HBITMAP hMaskBmp = (HBITMAP)::SelectObject(hMaskDC, hOldMaskBmp);
// Use new icon bitmap with text and new mask bitmap with text
ICONINFO ii2 = { 0 };
ii2.fIcon = TRUE;
ii2.hbmMask = hMaskBmp;
ii2.hbmColor = ii.hbmColor;
// Create updated icon
hNewIcon = ::CreateIconIndirect(&ii2);
// Cleanup bitmap mask
::DeleteObject(hMaskBmp);
::DeleteDC(hMaskDC);
// Cleanup font
::SelectObject(hMaskDC, hOldMaskFont);
::SelectObject(hMemDC, hOldFont);
::DeleteObject(hFont);
// Release background bitmap
::SelectObject(hMemDC, hOldBmp);
// Delete background icon bitmap info
::DeleteObject(ii.hbmColor);
::DeleteObject(ii.hbmMask);
::DeleteDC(hMemDC);
::ReleaseDC(NULL, hDc);
}

How do I save GDI+ Graphics / HDC to a file?

I am unable to save a GDI+ Graphics object, which is derived from a Device Context HDC, to a file.
What works: I am able to save a GDI+ Graphics derived from a Bitmap. Sample code (Win32):
Color color(255, 0, 0);
Pen pen(color, 2.0f);
CLSID pngClsid;
if(GetEncoderClsid(L"image/bmp", &pngClsid) < 0) // calls GetImageEncoders()
return;
// Graphics from Bitmap - works OK
Bitmap bitmap(300, 300, PixelFormat24bppRGB); // create Bitmap first
Graphics *graphics = new Graphics(&bitmap); // create Graphics second
graphics->Clear(Color(255, 255, 255, 255));
Status stat = graphics->DrawEllipse(&pen, 50, 50, 100, 100);
assert(stat == Ok);
stat = bitmap.Save(L"C:\\temp\\test1.bmp", &pngClsid, NULL);
assert(stat == Ok);
delete graphics;
Result:
What fails: If the Graphics object is derived from an HDC, I get a black rectangle. Regardless of whether I create the Bitmap at point [1], [2] or [3], I always get a black rectangle. Code:
CLSID pngClsid;
if(GetEncoderClsid(L"image/bmp", &pngClsid) < 0) // calls GetImageEncoders()
return;
// Graphics from HDC - fails
HDC hdc = GetDC(NULL);
Graphics *graphicsDC = new Graphics(hdc); // create Graphics first
graphicsDC->SetPageUnit(UnitPixel);
//Bitmap bitmapDC(300, 300, graphicsDC); // [1] create Bitmap second. Black rectangle if called here
graphicsDC->Clear(Color(255, 255, 255, 255));
//Bitmap bitmapDC(300, 300, graphicsDC); // [2] black rectangle if called here
HPEN penGDI = CreatePen(PS_SOLID, 3, RGB(0, 255, 0)); // old school GDI
HPEN oldPen = (HPEN)SelectObject(hdc, penGDI);
Ellipse(hdc, 50, 50, 150, 150);
DeleteObject(SelectObject(hdc, oldPen));
Bitmap bitmapDC(300, 300, graphicsDC); // [3] black rectangle if called here
Status stat = bitmapDC.Save(L"C:\\temp\\test2.bmp", &pngClsid, NULL);
assert(stat == Ok);
delete graphicsDC;
ReleaseDC(NULL, hdc);
Result:
Why I need this: I am converting code that contains thousands of calls to the GDI API. I want to start using GDI+ gradually without converting all GDI calls to GDI+ at once. I have mixed GDI/GDI+ successfully in other cases, for instance when creating gradients. The only difference is that in the other cases, I wasn't trying to save to a file.
I can't use CImage because it doesn't have groovy anti-aliasing.
So, how can I save a Graphic as image when starting with an HDC?

Using TransparentBlt for drawing in MFC

I want do draw a vector of CRect into a device context. The CRects that overlap should add up in a way that the intersection of all turn to a brighter green. Therefore I came up with the following code:
void Grid::tag(CDC* pDC){
CBrush brushGreen;
brushGreen.CreateSolidBrush(RGB(0, 100, 0));
CDC dcMemory;
dcMemory.SelectObject(&brushGreen);
dcMemory.CreateCompatibleDC(pDC);
for (size_t i = 0; i < taglist.size(); i++){
dcMemory.FillRect(taglist[i], &brushGreen);
pDC->TransparentBlt(frame.left, frame.top, frame.Width(), frame.Height(), &dcMemory, taglist[i].left, taglist[i].top, taglist[i].Width(), taglist[i].Height(),RGB(0,100,0));
}
DeleteObject(brushGreen);
}
Unfortunately, it turns out black. It seems like nothing is drawn into pDC. What am I doing wrong? Is this a valid approach to begin with?
In your example you have to fill memory dc with a transparent color. This will initialize the background color, so to speak. Then draw on memory dc and use TransparentBlt with that transparent color.
void CMyWnd::OnPaint()
{
CWnd::OnPaint();
CClientDC dc(this);
CRect rc;
GetClientRect(&rc);
//paint any custom background
dc.FillSolidRect(&rc, RGB(200,200,255));
//choose a color which you don't otherwise need, use it for transparency
COLORREF transparent_color = RGB(1, 1, 1);
//create memory dc and initialize with transparent_color:
CDC memdc;
memdc.CreateCompatibleDC(&dc);
CBitmap bitmap;
bitmap.CreateCompatibleBitmap(&dc, rc.right, rc.bottom);
memdc.SelectObject(bitmap);
memdc.FillSolidRect(&rc, transparent_color);
//start custom drawing on memeory dc:
CBrush brushGreen;
brushGreen.CreateSolidBrush(RGB(0, 100, 0));
CRect small_rc(10, 10, rc.right - 10, 20);
memdc.FillRect(small_rc, &brushGreen);
//end custom drawing
//Finish by copying memeory dc to destination dc:
dc.TransparentBlt(0, 0, rc.Width(), rc.Height(),
&memdc, 0, 0, rc.Width(), rc.Height(), transparent_color);
}
Instead of TransparentBlt - which does color keying during blt, you can use AlphaBlend, you have also other issues in your code. Below are some fixes, and ideas how to correct your code (I have not tested if this compiles).
CBrush brushGreen;
brushGreen.CreateSolidBrush(RGB(0, 100, 0));
CDC dcMemory;
//SO: what is the use of this? Also before creating DC
//dcMemory.SelectObject(&brushGreen);
dcMemory.CreateCompatibleDC(pDC);
//SO: for memory DC you need also a bitmap to be selected (dont forget to release it):
HBITMAP hbmp = CreateCompatibleBitmap((HDC)dc, 500, 500);
auto oldDcMemoryBmp = dcMemory.SelectObject(hbmp);
for (size_t i = 0; i < taglist.size(); i++){
dcMemory.FillRect(taglist[i], &brushGreen);
// SO: this is not needed
//pDC->TransparentBlt(frame.left, frame.top, frame.Width(), frame.Height(), &dcMemory, taglist[i].left, taglist[i].top, taglist[i].Width(), taglist[i].Height(),RGB(0,100,0));
// SO: Instead use albhaBlt
BLENDFUNCTION BlendFunction;
BlendFunction.AlphaFormat = AC_SRC_ALPHA;
BlendFunction.BlendFlags = 0;
BlendFunction.BlendOp = AC_SRC_OVER;
BlendFunction.SourceConstantAlpha = 15; // value 0 (transparent) to 255 (opaque)
dc.AlphaBlend(taglist[i].left, taglist[i].top, taglist[i].Width(), taglist[i].Height(), &dcMemory, 0, 0, taglist[i].Width(), taglist[i].Height(), BlendFunction);
}
//DeleteObject(brushGreen);

saving to bitmap with win32 API or GDI+

Question is rather simple but I couldn't find nice & clean solution to my problem on the Internet.
What I got are some drawings on my window. Now, I can save those using BitBlt function from window device context to image device context, and also from there to bitmap handle:
HDC bitmapDC = CreateCompatibleDC(dc);
HBITMAP bitmap = CreateCompatibleBitmap(bitmapDC, 200, 200);
SelectObject(bitmapDC,bitmap);
BitBlt(bitmapDC, 0, 0, 200, 200, dc, 200, 200, SRCCOPY);
But from there I'm lost. I had a look at GDI+ Bitmap class which got save function, and I found how to implement code for retrieving CLSID of picture encoding. However I don't know if I use loading to that class correctly. There's overloaded constructor for HBITMAP, but it's also asking for some palette, which I set to NULL:
Bitmap image(bitmap,NULL);
I tried to save png file but it resulted in black quare without those drawings I was expecting. If you'd like, full code for my painting procedure:
void GetCLSID(const WCHAR* format, CLSID* pClsid){
UINT num = 0; // number of image encoders
UINT size = 0; // size of the image encoder array in bytes
ImageCodecInfo* pImageCodecInfo = NULL;
GetImageEncodersSize(&num, &size);
pImageCodecInfo = (ImageCodecInfo*)(malloc(size));
GetImageEncoders(num, size, pImageCodecInfo);
for(UINT j = 0; j < num; ++j)
{
if( wcscmp(pImageCodecInfo[j].MimeType, format) == 0 )
{
*pClsid = pImageCodecInfo[j].Clsid;
free(pImageCodecInfo);
}
}
}
void OnPaint(HDC dc){
RECT rect; rect.bottom = 0; rect.top = 20; rect.left = 0; rect.right = 100;
HBRUSH blueBrush = CreateSolidBrush(RGB(0,0,200));
FillRect(dc, &rect, blueBrush);
Graphics graphics(dc);
Pen pen(Color(255, 0, 0, 255));
graphics.DrawLine(&pen, 0, 0, 200, 100);
SolidBrush greenBrush(Color(0,200,0));
Rect ellipseRect(20,20,20,20);
graphics.FillEllipse(&greenBrush, ellipseRect);
SolidBrush redBrush(Color(200,0,0));
Rect boxRectangle(0,40,20,100);
graphics.FillRectangle(&redBrush, boxRectangle);
pen.SetColor(Color(200,0,200));
pen.SetWidth(20);
graphics.DrawBezier(&pen, 100, 20, 130, 40, 200, 10, 230, 20);
HDC bitmapDC = CreateCompatibleDC(dc);
HBITMAP bitmap = CreateCompatibleBitmap(bitmapDC, 200, 200);
SelectObject(bitmapDC,bitmap);
BitBlt(bitmapDC, 0, 0, 500, 500, dc, 500, 500, SRCCOPY);
Bitmap image(bitmap,NULL);
CLSID clsID;
GetCLSID(L"image/png", &clsID);
image.Save(L"pic.png", &clsID);
}
I couldn't even imagine that simple saving will be such problem, so I'll be glad for any help, thanks!
I gave code here which does pretty much what you want:
How to save the client area of a child Window to a Bitmap file?
It is very verbose in C. It's a lot better in C++ because of CImage