Question

Using a Webview, I would like to find the links inside this page.

-(void)webView:(WebView *)sender didFinishLoadForFrame:(WebFrame *)frame {
    DOMDocument *myDOMDocument = [[self.webview mainFrame] DOMDocument];

looks like a good starting point, but I find the WebScriptObject Class Reference a little cryptic. Clearly I don’t want to evaluate some Javascript to get the links. I want to directly read the DOM.

How may I find which nodes in the DOM are links, and get the address they are pointing to ?

Answer 1

Find the DOMNodes for images

see walkNodeTree @ http://cocoadev.com/wiki/DOMCore

— complete sample to find image nodes, get their src and make nsimages

@implementation DDAppDelegate

- (void)applicationDidFinishLaunching:(NSNotification *)aNotification {
    [self.webview.mainFrame loadRequest:[NSURLRequest requestWithURL:[NSURL URLWithString:@"http://dominik.pich.info/Home.html"]]];
}

-(void)webView:(WebView *)sender didFinishLoadForFrame:(WebFrame *)frame {
    DOMDocument *myDOMDocument = [[self.webview mainFrame] DOMDocument];

    NSMutableArray *imgs = [NSMutableArray array];
    [self walkNodeTree:myDOMDocument imgsCollected:imgs];

    //bad code, demo
    NSMutableArray *nsImages = [NSMutableArray array];
    for (DOMNode *img in imgs) {
        for(int i = 0; i < img.attributes.length; i++) {
            DOMNode *attr = [img.attributes item:i];
            NSLog(@"%@", attr.nodeName);
            if([attr.nodeName.lowercaseString isEqualToString:@"src"]) {
                NSString *urlstring = [attr nodeValue];
                NSURL *url = [NSURL URLWithString:urlstring relativeToURL:[NSURL URLWithString:@"http://dominik.pich.info/"]];
                NSImage *nsimg = [[NSImage alloc] initWithContentsOfURL:url];
                if(nsimg)
                    [nsImages addObject:nsimg];
            }
        }
    }

    NSLog(@"%@", nsImages);
}

- (void)walkNodeTree:(DOMNode*)parent imgsCollected:(NSMutableArray*)imgs {
    DOMNodeList *nodeList = [parent childNodes];
    unsigned i, length = [nodeList length];
    for (i = 0; i < length; i++) {
        DOMNode *node = [nodeList item:i];

        NSLog(@"%@", node.nodeName);
        if([node.nodeName.lowercaseString isEqualToString:@"img"]) {
            [imgs addObject:node];
        }
        else {
            //recurse
            [self walkNodeTree:node imgsCollected:imgs];
        }
    }
}
@end