Web scraping / store data

Hello, I would like to scrape a website, repeating on a variable time frequency. I'd like the data to be stored in a database that I can access. There are three very similar sources. They are at: [login to view URL] [login to view URL] [login to view URL] I haven't revisited the site all that many times, so there's a chance that there are some elements that change that will make this tricky. But I've visited enough that it seems to be rather constant. Here's the data model that I propose: plays ------- id (PK autoincrementing int) created_at (timestamp) play_name (varchar 255) is_non_musical (tinyint) is_on_broadway (tinyint) locations ---------- id (PK autoincrementing int) created_at (timestamp) location_name (varchar 255) location_url (varchar 255) crawls -------- id (PK autoincrementing int) created_at (timestamp) location_id (int) http_return_code (int) listing_updated_at (timestamp, nullable) -- this is from the crawled inner html listings -------- id (PK autoincrementing int) created_at (timestamp) play_id (int) play_time is_tomorrow (tinyint) percent_off (int - nullable) configuration_parameters --------------------------------- id (PK autoincrementing int) polling_frequency_in_minutes (int) Basically, I want to get the listings from the grid and populate in the database. There should be three calls - once for each of the three above URLs (which will be stored in locations.location_URL). It should crawl the site every n minutes, where n is defined in configuration_parameters.polling_frequency_in_minutes. To start, let's set the polling frequency to 60. Capture the "crawl" top-level info in the crawls table. Then break the listings data into the listings. Check first to see if a listing's play's name exists already in the table; if so, refer to the existing plays record. If not, insert a new listing into the plays table and refer to that play_id in the listings data. Make sure to capture the http return code (should be 200 if successful) so that we can see if the URL changes (they don't seem to have 404s for bad URLs but a 301 to redirect to the homepage) and/or if they start blacklisting our IP address and give a non-200 answer. If you can set this up with [login to view URL], ideally we have zero maintenance costs since this becomes effectively free. If the data model needs to change slightly to accommodate [login to view URL], that's fine. Please let me know what ends up changing from the data model proposed above. Alternatively, if [login to view URL] is not practical, I'd like you to set this up on a server with a mysql database (or other simple db) that I can access with a local client. Obviously we need a very simple server setup. I'd like to use an inexpensive commercial server -- I do not want to rely on a machine that you maintain. Please suggest an inexpensive setup for ongoing access, and I will pay for it directly. We just need *very* basic stuff here, obviously. Alternatively, if the server maintenance is unreasonably high and if [login to view URL] is not practical, I'd like you to set this up so that it can work locally on my machine. By accepting this job, you agree that you claim no rights to the work or to derivatives of the work and that all rights belong to me. You cannot show off this project to others, including in your portfolio here on Freelancer.com or elsewhere.