views:

104

answers:

1

I am trying to create thumb images for every pdf page in PDF document and place it in a UISCrollVIew. I have succeeded in this, but scrolling is not so smooth as I want when it's too fast. And I want to optimize thumb images creating for Pdf page. I want to create one CGContextRef and reset its content after CGContextDrawPDFPage, as a consequence I wouldn't have to create a context each time and perform some other calculation, which takes a lot of resources.

Is it possible to reset CGContextRef content after CGContextDrawPDFPage? CGContextRestoreGState and CGContextSaveGState seems to doesn't help in this situation.

NSAutoreleasePool *pool = [[NSAutoreleasePool alloc] init];
GCPdfSource *pdfSource = [GCPdfSource sharedInstance];
for (int i = pageRange.location; i <= pageRange.length; i++) {

    UIView *thumbPdfView = [scrollView viewWithTag:i+1];
    if (thumbPdfView == nil) {

        CGPDFPageRef pdfPage = [pdfSource pageAt:i + 1];
        float xPosition = THUMB_H_PADDING + THUMB_H_PADDING * i + THUMB_WIDTH * i;

        CGRect frame = CGRectMake(xPosition, THUMB_H_PADDING,  THUMB_WIDTH, THUMB_HEIGHT);
        thumbPdfView = [[UIView alloc] initWithFrame:frame];
        thumbPdfView.opaque = YES;
        thumbPdfView.backgroundColor = [UIColor whiteColor];
        [thumbPdfView setTag:i+1];

        CGColorSpaceRef colorSpace = CGColorSpaceCreateDeviceRGB();
        CGContextRef context = CGBitmapContextCreate(NULL, 
                                        frame.size.width, 
                                        frame.size.height, 
                                        8,                      /* bits per component*/
                                        frame.size.width * 4,   /* bytes per row */
                                        colorSpace, 
                                        kCGImageAlphaPremultipliedLast | kCGBitmapByteOrder32Big);
        CGColorSpaceRelease(colorSpace);
        CGContextClipToRect(context, CGRectMake(0, 0, frame.size.width,frame.size.height));


        CGRect pdfPageRect = CGPDFPageGetBoxRect(pdfPage, kCGPDFMediaBox);
        CGRect contextRect = CGContextGetClipBoundingBox(context);
        CGAffineTransform transform = aspectFit(pdfPageRect, contextRect);

        CGContextConcatCTM(context, transform);
        CGContextDrawPDFPage(context, pdfPage);


        CGImageRef image = CGBitmapContextCreateImage(context);
        CGContextRelease(context);
        UIImage *uiImage = [[UIImage alloc]initWithCGImage:image];
        CGImageRelease(image);

        UIImageView *imageVIew = [[UIImageView alloc]initWithImage:uiImage];
        [uiImage release];

        [thumbPdfView addSubview:imageVIew];
        [imageVIew release];

        [scrollView addSubview:thumbPdfView];
        [thumbPdfView release];

    }
}

[pool release];//release

and aspectFit function...

CGAffineTransform aspectFit(CGRect innerRect, CGRect outerRect) {
CGFloat scaleFactor = MIN(outerRect.size.width/innerRect.size.width, outerRect.size.height/innerRect.size.height);
CGAffineTransform scale = CGAffineTransformMakeScale(scaleFactor, scaleFactor);
CGRect scaledInnerRect = CGRectApplyAffineTransform(innerRect, scale);
CGAffineTransform translation = 
CGAffineTransformMakeTranslation((outerRect.size.width - scaledInnerRect.size.width) / 2 - scaledInnerRect.origin.x, 
                                 (outerRect.size.height - scaledInnerRect.size.height) / 2 - scaledInnerRect.origin.y);
return CGAffineTransformConcat(scale, translation);

}

A: 

Try CGContextClearRect(context, contextBounds).

CGContextSaveGState and CGContextRestoreGState do not have any effect on the content of a context. They push and pop changes made to state aspects of the context like the current fill color.

Ken
Thanks! It works, now I don't have to create a context every time I draw pdf page :)CGContextClearRect(context, CGContextGetClipBoundingBox(context));It didn't help as I expected (it's not so smooth on too fast scrolling) but anyway it's better than before.
Koteg