I have found some libraries or web services in PHP that does the job. The problem is that the conversion is done when the page is fully loaded, I would like to convert the page to PDF after some content dynamically added via AJAX in onload event.
Thank you very much, Omar
When you add content with AJAX, that's only happening on the client's machine, so the server-side PHP conversion code will not work with it. You'll need to come up with a server-side method of including the dynamic content if you want to do the PDF creation on the server.
This is a client-side requirement due to the Ajax situation. If they're using a Mac they can simply save as PDF but otherwise you have little control over what they can do.
What you can do however for a server-side implementation is keep a record of the page and all their AJAX requests since loading that page, and construct the HTML and PDF on the server from that record if it is requested. Of course this is not a simple task and quite overkill. You'd probably be better off having a different mechanism to create a server-side PDF report of a page other than turning HTML into PDF.
You could possibly implement an AJAX call that sent the page content/state back to serverside, after the dynamic contens is added, where it could be rendered as a pdf. You might not need all the page, depending on where this "dynamic content" is going to go.
Seems a bit messy, but without knowing the project it's hard to say whether there could be a cleaner method to do what you're trying to do.
You could post back document.getElementsByTagName('html')[0].innerHTML to the server (possibly using AJAX) and generate a PDF from that.
If your page can be updated solely on the client with no trip to the server, you'll have to post back up to your app. That way you'll have all the content and will not have rebuild, which would not be possible if there are client side only interactions.
Wow, thank you everyone I didn't know that this is comunity is so active. To answer Beepcake about the project:
When the page loads the app retrieves, from more than 40 servers, biological information via AJAX request, then a unique view is displayed where you can manipulate the graphic with many options.
So, the cool thing will be to print when the user makes his own version of the graphic. I think that the best solution is to POST the entire HTML with document.getElementsByTagName('html')[0].innerHTML as RoBorg has said and then generate the PDF with a library such as dompdf
But document.getElementsByTagName('html')[0].innerHTML give incorrect html with missing double-quotes & end tags in IE & Firefox. Hence exception is thrown by iText .