On Wed, 8 Jun 2016, at 17:16, Bob wrote: > Well Tomas, I like to share with you that your guess is again correct. > This link quickly redirects to http://<linkk> (seems the session checked > here) and then to home page. I'm tying to figure out the cookie part now.
It's important that at the start of this process you have an empty cookie file. Then you give its name to curl on every call; curl will store any cookies that the site sends to you in that file (you shouldn't have to pay any attention to what they actually are). In every subsequent curl command, you again specify the same file and the cookies whose values are stored in it are sent back to the server by curl. Each time the server might change cookies but the whole thing should just work. At the very end of the process, throw the cookie file away. Also, in the curl command you can specify names of files into which curl will place the HTML of the page you just fetched, and you can sepcify a second file to have the headers (which accompany every page sent to a browser, but users don't ever see them) stored. It's good to do this because those headers will include any server error codes - useful for diagnosing problems in the whole process - and they also contain eg redirection instructions, if the server wants you to make another request for a different page. if you don't have the headers file for your script to read and check it can be really difficult to make this sort of thing work! -- Jeremy Nicoll - my opinions are my own.