Find Information from Websites

We do 360 Spin Panoramas for Car Dealerships, we have a self contained button-iframe based on a php script on 10-15 dealership websites, containing a button to click thru to our 360 Link. We need our scripts cleaned up and standardized, and with that we need our Missing "SPINS" report functioning. I need this project completed in 10 days. No exceptions. What the missing 360/spin report needs to do: Tell me what vehicles on dealership website do not have 360 via stock# Sort by New/Used/Make/Model Erase sold vehicles as they are removed from said dealerships website. Please message me and I will show you what work has been done thus far. Also, this is what the previous programmer proposed to us. (he did not complete the job) * Standardisation of all scripts, make sure they all work and simplify generation of new scripts. * Add meaningful diagnostics to identify what else isn't working and provide ongoing monitoring. At the minute each script has 5 or more code paths that can be affected by a number of factors, it is impossible to be certain how many of these still work. (e.g. zscript is the only one that checks one of these 5 paths, so your developer must have seen a problem there but what about the others; I added a missing check to the others.) The scripts that attempt to scrape information can, and have been, broken when your partners have changed their websites and by changes in the web browsers being used by customers (particularly the increase in iPads etc.) * Make revo page test fast and low server impact Currently a page with more than one revo can cause it to make loads of slow web server to server connections. You could potentially draw complaints from your partners for hammering their servers doing this. This slows things down exponentially as each connection generates more and more work. (Under a small amount of traffic some spin buttons are taking 24 seconds, most are averaging at 6 seconds; it should really be less than 1.) * Technical Redesign All this tends to suggest combining existing monitoring, missing spins report and Perillo survey into a solid system that actually works. Cleanup of database - one database table has 482960 rows and is growing, but doesn't seem to have been checked on since August 7th 2013. That isn't going to end well if left alone. Modernisation of the revo iframe (html5) will reduce the browser support problem - iframe security features have been advanced in recent browsers. * Repair broken scripts The scraping of car details from the dealers site isn't really working very well and has the risk of breaking later if they change something, I will recommend ammended iframes as a first option and review where these are needed and how they could be implemented on a referring site by site basis. It may not be possible for the dealers to get this done easily if they have a 3rd party web designer, I can try to resurrect the scrapers that can be repaired. The scrapers should run as a regular cron job to avoid thrashing the dealers websites, currently a page with more than one spin can cause the server to repeatedly ask the dealer server for the same page (despite the fact it can't correctly get the stock number as there is more than one.)