Hi Thomas! I have done this (extended the link dialog), and I can say that its _not_ an easy task (even having the sourcecode available). I did it by using a shadowed "Util/_dialogs/HyperlinkProperties.aspx" where I added a new input control to the dialog. I then modified the script to combine the url with my "extra input" to form one url in the form "http://<<original url>>$$CSTRACK=<<extra input>>" Then I also needed an editor plugin to run some specialized javascript (overriding the scripts/HyperlinkProperties.js) to look for url's having this "$$CSTRACK=" parameter, and if found, extract that extra input and put it into a "onclick=" attribute of the A-element. This approach would not be ideal in that it requires modifications to episerevr core files. Even though it can be done through shadowed files, there's always the potential problem with upcoming service releases. In my opinion, I think you should consider some other alternative: *) create an editorplugin to modify an existing A-element. This requires however the editor (person) to take an extra step when creating such links. *) derive a new Property from PropertyLongString and insert the appropriate onclick-javascript upon submitting by parsing the htmlcode server side. (this would be the best option if the onclick-generation could be totally automated and requires no extra user input). *) create an editorplugin where you take care of inserting the links yourself. probably the "best" solution if extra user input is needed, but requires a lot of coding... If you need more details, send me an email and I'll try to "package" the stuff together for you. Regards, Johan Olofsson EPiServer AB

I will definitely not go for the alternative with extending the link editor! Just wanted to get that confirmed as a bad idea :-) Alternative no. 2: I guess if I can hook up to the event where the editor content is saved, I can run an algorithm to find all elements and insert the onclick javascript code. The code is really quite static so I don't see the need for an extra property - the only variable part is the URL of the download, and I already have it from the href attribute. But, with this approach, I have to keep a list of what files are considered downloads, to I don't end up counting normal links to other pages (double-counting). Alternative no. 1 then on the other hand seems to leave it up to the editors to decide which links they count and which links they don't, but as you said it adds an extra step, plus I'm unsure of how much effort I have to put in to create such an editor plug-in. Alternative no. 3 seems like too much work... I guess a fourth alternative, or workaround, would be for editors to link to a proxy page that holds the file to be downloaded, which when viewed runs the javascript before redirecting to the actual file. But that could perhaps diminish the user experience plus it adds more work to the editors. Alternativ no. 2 seems my best option.

I've fiddled with both the link editor and the image editor and extended it quite a bit. It's not easy like Johan says, but definately not impossible. I recently made a change to the link-editor and added a style dropdown for selecting a link css style, like you can do with text in the editor (styles are read from the css just like "EditMenuName"). It took about 4-5 hours. If you document the changes it wouldn't take long to reproduce in case you need to upgrade and you're unlucky enough to have the link editor replaced by an updated one. It would think it takes about the same time to create a hack this way as making a routine that parse the saved text and look for links etc. Parsing text could be affected by an updated aswell if the link format is changed in some way.

If you want a fast and dirty method of just inserting that onclick in the links of the body part or so I suggest doing the algorithm for insert to the tag when you show the page. Like then you want to do your myBodyLiteral.Text = MainBody you do some search and replace and insert your onclick. That way no modding needed at all and no extra step when creating the page. Hope that helps =)

One easy way would be to add client script and attach a method to the onclick event of all anchors. I have done similar things on many different sites and for different purposes. This example might lead you in the right direction. The code below is from http://www.naturvardsverket.se that I made last winter. @ http://www.naturvardsverket.se/sv/Nedre-meny/Fragor-och-svar/Artskydd/ you can se the script in action. You need also need the 'eol.js' that can be found @ http://www.robertnyman.com or more specific @ http://www.robertnyman.com/2006/10/03/elo-encapsulated-load-object-the-ultimate-way-to-handle-window-load-events/ Regards, HAXEN // --- var arrQandAs = new Array(); function initFAQ() { var arrQuestions = getElementsByClassName(document, "span", "question"); var oQandASection; for(var i=0; i