Data mining from web service

The source: [login to view URL] that will show you a top-of-tree of 30 different links (branches). On each of these links may be a list of one or more links. However, at the very bottom of the list on the second page, there is also a link that starts with: "V?"LJ ALLA INOM" which let's you see all links in that branch (e.g. [login to view URL]) On the left hand side of the (30) pages shown by doing the above, there are four interesting choices under the heading "Omsättning" (means Turnover), namely the ones from turnovers 1000tkr and above. Luckily, these are named by just adding "/xo/4", "/xo/5", "/xo/6" and "/xo/7" respectively (so you have e.g. [login to view URL]). On all of these, there will be different number of records, of which 10 new links are shown to each page (next page is called by just adding "/page/2" (e.g.: [login to view URL]) with 10 new records, etc. When following every one of these links (you get to a company-presentation page like: [login to view URL]), I would like to have the information that sits on these pages in columns (sql or xls) that are on that page including the data for the three years you find in the yellowish part for the last three years (fiscal years ending in the month at the top): The only information you DON'T need to collect is "Visa på karta" (show on map), PLATS F?-R BOLAGETS PRESENTATION, Kreditupplysning, Årsredovisningar, and you don't need to follow any links in the left menu, nor the link: "Fler bokslut- och nyckeltal" towards the bottom. In total I think there could be about 200 000 records like this, maybe more.