> > On Sun, 31 Jul 2011, Graham Lawrence wrote: > > > I'm writing a bash script to automate my sign-ins at various web > > sites. Is it possible for the script to access the content of the > > blue highlighted line near the bottom of the lynx screen, so that it > > can determine if it is currently at a (NORMAL LINK) or at a (Text > > Entry Field), etc? > > Not easily, I think. The usual response to questions about > screen-scraping would be to use "expect", but I believe that's > too hard (because of all of the screen optimization made by curses). > > -- > Thomas E. Dickey
There are a handful of tools you can do this sort of thing with. Look for something that mentions pseudoterminals. I've had good experience using C-Kermit to drive Lynx. I may not understand your strategy, but spending a lot of time extracting the blue highlighted line seems too passive. Decide where you're going to, and go there. Aggressive use of URL #fragments, link numbering and and finding GET tackons and directory paths of URLs in ADDRLIST/LIST views of the page can simplify navigating there. At least that's my experience. :-) This is exactly the situation where including the form html elements in ADDRLIST views of the page would simplify things. (Hint, hint, nudge, nudge, I submitted a wish list request about this some time back. ;-) ) Regards, Dallas E. Legan II / [email protected] / [email protected] http://isthereanotherquestion.blogspot.com ************************************************************************ "Tekeli-li! Tekeli-li!" - "At the Mountains of Madness", H.P. Lovecraft --------------------------------------------- This message was sent using Endymion MailMan. http://www.endymion.com/products/mailman/ _______________________________________________ Lynx-dev mailing list [email protected] https://lists.nongnu.org/mailman/listinfo/lynx-dev
