Guerrilla

OK this is weird! I tried compiling it several times yesterday, each time ubot crashed. I tried compiling it again today and it compiles and works! A big relief but I hope I dont have the crashing bug again next time i compile.

The problem I get is that when I add the $application folder to the path then ubot crashes when I try to compile it.

In my bot I populate some menus from text files. The problem I have run into that I just cant see a way round is how to do change the logic for distributing the bot. I can only get the menu to populate when I use an exact path like "C:\folder\file.txt". If I use "$application folder\file.txt" then it doesn't work. This is a real problem. Right now the way my bot works is that user fills out a configuration file with the user/pass for various sites. They then run the setup script and all the text files for the menus are generated. Right now those menus are only loading with hardco

Yeah, all someone would have to do is use a network sniffer and get all the URL's for my source code library I spent the last year creating. Were talking about a lot of script pages here. It's just not an option to have all that code in a single bot, especially as I would only want to show maybe 4-5 tabs to the user. Ideally what ubot needs to do is have the ability to create a compiled include file, so even if someone gets your include files, they cant just open then up and take all your work. Is there any solution to this problem that works now?

Make some javascript code that changes the value of a hidden field on the page. You can then use this in your logic to create different actions determined by user selection. You could use a waitfor command to discover when they have made choice but this is very buggy. The way I do it is I add in a pause, and after any choice is made I make text appear in window telling them to click play to continue. Its haphazzard but works. I use this for sakiving logs after an operation completes. As often an operation will fail for some reason, i output message saying "if operation completed succe

I have run into a problem I wonder if anyone can help me with. I have my bot code organised into a library (folders containing bot source code). Then when I want to create a bot I just include the libraries I need and have instant access to all my routines. The problem is that when I compile the bot, the includes are not compiled with it. The only way of getting the compiled bot to work is to actually include the bot source code for any includes. I don't want to give away my source code with releases. Is there a way I can get around this? (I am using 3.5 still)

The is a programming library I came across once (cant remember name) that allowed you to feed in a pdf and it would export the PDF as HTML. I even remember finding a few websites that you could upload a pdf to and get a HTML version of it. This would be the route to go down to scrape the text via ubot.

I use search & replace to remove all quote marks with unique placeholder then replace them back in when i need to use data.

A member of the forum made an app that does this which you can call from ubot. He shared it here for free. Apparently next version of ubot will have this feature built in.

I wrote an adwords scraper. The annoying thing was that google change the html from search to search. There are 3 main types of pattern (well there were when i did it). Also to scrape multiple pages you will need to disable cookies otherwise when you programatically access several pages google just starts showing you the same set of ads over and over again.

I just ran into the 5000 word count issue. I am sending already spun articles to the API to add nested spins so I have to be careful as a break in wrong place would break whole article. Logic I am following is this: - Break article into a list seperating by $newline and removing blank entries. - Send each paragrpah one at a time - Re-assemble article A longer way about it but work. If I get time I will write some JS that will make this better.

Thanks Praney

How much will upgrade cost?

Does utorrent have a log file? If it does you can parse this to work out what is going on.

Another trick if you don't want to enable legacy mode is to call your bot iexplore.exe (or the same filename of any other supported browser) and it will work.

set the random text to a variable first then you can get the folder name easy.

Here you go. I tried to make it in way so you can modify it to fit your own bot. Add a keyword in like "ReD" in text box and it will return all variation regardless of case from the string in the "write browser" command. scrape non case sensitive.ubot

I think this may be a windows permissions error. Are both user accounts administrators? I would suggest clearing your event logs then trying to run the bot and see if any entries appear in the security/application log. If they do post them here. If you are on windows 7 or vista you could also try disabling UAC.

From what I understand Ubot anly supports http proxies so to hide your IP on the email you will need to setup a vps on the machine running the bot. A lot of VPS come with a web admin panel that lets you log in and change the IP of the VPS so you could automate that with ubot (although most will limit the number of IP changes you are allowed within a given period of time). Alternatively you could configure a mail relay on a dedicated server that rotated proxies. I am not sure exactly the steps to do this but if you need a new IP very frequently then this could be the only option.

I had this issue on my xp box because when I installed it it asked me to select admin account to install it with and I think I selected a different account to the one i use. I then had trouble updating so i re-installed it and found that it was working much faster.

Do you know which reg key needs to be changed to disable cookies? Or the COM command?

I have noticed some sites using cookies to prevent scraping. Is there any way I can make my bot disable cookies before scraping and then re-enable them after?

Here you go, proof of concept - http://dl.dropbox.com/u/19226543/adwords%20scraper.exe Think its working pretty well now, will tidy this up into a proper bot now.

I take it back, google rotate different class ID's making it hard to scrape. I am trying to find a solution but it is more of a time consuming job than i first thought

I have been meaning to make a bot like this so I decided to give it a go. The answer is yes it's easy (although I hit a few snags but got past them easy enough). To code in allowances for multiple datacentres and format data to be pretty will be a bit more effort but this is a pretty straightforward thing to do in ubot.