Some years ago I wrote a lengthy GWbasic (ie ancient MS DOS) program to manipulate a large multi-column Comma Separated Value (.csv) file that I need to use.

I have managed to get GWbasic running in Linux in DosBox. I have been unable to get a GWbasic-clone, PC-Basic, to run on my computer. Perhaps it is time to write something that runs directly in Linux.

The GWbasic program does things like remove columns, re-arrange the order of columns, sort the whole file according to the contents of one particular column, and add up the values of a particular column when the values in another column are the same.

Is there any program with scripting that can do this? Or what computer language would be easiest to write this in?

Thanks

you can use a lot of languages, like awk/perl/python. The easiest one is the one you already know, but actually it depends on the difficulty of your original program too.

The 'easiest' answer is the one pan64 gave you..it's whatever you're comfortable with. And before I could answer this, I'd ask some questions...like how big this CSV file is (rows/columns), how often you need to manipulate it, how FAST you need it manipulated, and most importantly, how it's generated. Because:

• It could very well be that whatever is generating this file can be set to output things the way you want them, and eliminate the need to manipulate the CSV at all.
• If it's HUGE (like hundreds of columns/tens-of-thousands of rows), then language choice makes a difference.
• If you only have to do this a few times a year, why bother with a program at all? Open the file in Libreoffice spreadsheet, and delete/sort/total/whatever you want, and save it back out. You could even write a macro to do this, if it's always a set number of steps/moves/etc.

Personally, I'd do it in Perl, because that's my typical go-to language for text manipulation, and as pan64 alluded to, go with what you know best.

Nice idea TBone - an ex-basic programmer should be able to handle spreadsheet macros pretty naturally.
As for language, here's a vote for awk if the OP has no learned language on Linux. I found the learning curve much easier than perl and the day-to-day usability more accessible. My usage of perl has dropped dramatically over the years in favour of awk.

Really? That's interesting, because I find perl MUCH easier to work in, mainly because of the huge library of modules you can get from CPAN that can cookie-cutter a lot of what you need. There are lots of modules already done to handle CSV files, which deal with double/single quotes, etc., without having to re-invent the wheel.

I use awk sparingly, but (like you do with awk probably), have a library of perl code I've written over the years I can re-use easily. Once you get that library built, it's hard to avoid using it.

All good points - I take it you haven't moved on to perl6 then ? ....
A new user might think "perl6 must be newer/better than perl5, I'll use that". Now that would be an interesting introduction to languages.

Well...not just yet. I'm waiting. awk is great, though, but just not something that I've spent time learning as well as I could. Like the old adage, "if all you have is a hammer, everything looks like a nail".

But as far as the Op's question goes, it would (to me), depend on how often this has to be done, and the conditions it has to run under. If this is a CSV that's shoveled over from a vendor daily, then that's something that needs to be automated with a proper program. But once a quarter? Overkill...use Excel/Libreoffice to deal with it in a few minutes and be done. Code a program to do it on a rainy day, when you're out of things to do.

How often and/or does it get over written by the creating program?

A search for:
came up with some nice awk|sed|perl options in just the first three hits.

I guess "column" isn't what is wanted. From the man page:
This did a nice job on a test file:
See the results.
Should be noted; I'm "computer" language challenged. I can tweak scripts, but have problems creating them.

My 2 cents are for perl. It's quick to make short scripts with it. The split() function would be what I'd look at to split the columns.

Edit: if the data is more complicated you might use CPAN's Parse::CSV or Text::CSV

Agreed, and the Text::CSV module is the one I'd use.