I have a directory full of files and the format of the names vary somewhat, for example:
John Doe - Data.dat John_Doe_-_Info.dat JohnDoe_Account-Nov.dat
Nov_-_information_-_John_Doe.Dat
There are other patterns but loosely follow the ones above. What I need is help writing a shell script "bash or perl" to put all these files in one standard context. This will be my first attempt at anything other than simple one line shell scripts. How do and Where do I begin?
running F20 and KDE 4.11.3 if that makes any difference.
Thanks

You've only stated half of your requirements. What standard format to you want all of the files to eventually end up as? What have you tried so far? What you ask is a relatively simple task in bash but this sounds like homework so I'm less inclined to give you an answer outright.

The final format doesn't so much matter as long as the group is cohesive, that said however, data before user, eg. Data-Nov_John-Doe.dat would be sufficient. There is over 6000 of these files. The only thing I have tried so far is mv *.dat *.dat the star meaning source and destination names. This is the only way I know. Sorry it sounds like homework to you but to me it is a masters thesis on a topic i did not study.
Thanks

Just to be clear: this is more of a hobbyist forum where like minded individuals come together and discuss various topics. It's not really a "script farm" (best term I could come up with) for generating scripts upon request. In any case, I can point you in the right direction for researching how this could be accomplished. One thing to keep in mind is among the plethora of available tools in Linux there are hundreds of ways this can be accomplished.

First I'll reference sources recommended for general reading and then more to the point.

• Bash guide for beginners
• Advanced bash-scripting guide
• Bash manual
• coreutils manual

The feature you referred to as *.dat is known as name globbing. This will get you nowhere with just the mv command alone. What you need to do is iterate through all of the file names using a loop (learn about loops in the bash docs see compound commands). I would likely use sed to perform a substitution from the coreutils package which would involve using it likely with an echo command with command substitution (see command substitution under shell expansions in the bash manual). With sed I'd be using a regex to match common patterns. Then once I have the source file name and destination file name (likely each stored in a variable) I would then perform the mv command. This process would happen through each iteration of the loop. Be aware that this approach has a few non-obvious pitfalls such as not handling leading or trailing spaces.

This is a start for you.

Thanks, as I am new to Linux. A place to start is all I ask. I never want a hand out only a hand up and you have given it to me. Now that I know where to start I am no longer overwhelmed.

I understand which is why I worked so hard on crafting an answer for you. If I thought all you wanted was a simple script I wouldn't have bothered. I just wanted to convey the general attitude of this community because the framing of your question looked a lot like elementary "introduction to Linux" course work college students generally sludge through. Often, students come here expecting to be handed answers without effort on their own part and then label the community useless when their expectations are not met.

I'm not categorizing you with that. I'm just stating where my original mentality stemmed.

Also, the find utility is occasionally used to solve a problem like this as well. Both name globbing in a for loop and the find utility generally work.

e.g.
Definitely use the man pages. Almost every utility ever comes with a man page. There's even a man page for the man command.

Learn the navigation shortcuts because man pages (such as man bash) can be very large.

Unrelated but also a valuable resource: You should also check out IRC. That's just another place to get help on any subject involving Linux or programming. See freenode.net for that. There's a #bash channel on freenode.

Thanks. I will spend the night seeing what damage I can accomplish. Much Respect.

Once you make progress on a script feel free to post it in this thread. I can help you work out the details and improve it. You'll generally find this community is most helpful when you are able to provide a small sample of what you're trying to accomplish. Members usually jump at the opportunity to collaborate in this regard.