Here I come up with the second post of the series and my recent free lance programming experience.
This project was also from Eric Willis who happily chose me for another of his projects. This time I developed http://crowdminz.com . It’a a live web crawler that crawls the www in real time.
As you will notice, the first page has the “web” category pre-selected. And there are three columns showing results from — twitter, digg and friendfeed. You might be surprised to know that search box uses no forms. It just happened that I coded the php portion while the site layout was coded by a third party coder. When I tried to put a form into the right place, the page broke. I then put a single input box and used javascript to make it behave like a html form. When you click the search button, the value of the search box is collected by javascript and the window.location.href is assigned a new value that necessarily contains a new URI with query string. To make it behave like a real form, I had to filter key presses and make sure that the search is performed even when the user pressed the Enter button after typing something in the box.
I used the twitter, digg and friendfeed RSS search results to gather data from them. I used SimpleXML with php to parse the data and form the results. For video, I had to use Youtube and Flickr for image results.
It was a $50 project and I completed it in almost 6 working hours. I took time to dig into the HTML layouts where I lost track every now and then. Finally it was done and I did enjoy coding the javascript portion.