Extracting relevent data from a CSV file?

[shabbysquire]

A bit rusty with table processing, so looking for some advice.

I have a CSV file with approx 10 columns, but I only need to extract the data from 4 columns, like this:




And in table processing, I thought I could match up the CSV columns (A, B, F, I), like:


But when I get a new line, all the info is put into variable A, instead of the 4 x variables.

Could you tell me where I'm going wrong?

Thanks.

 

[rostonix]

Standard delimeter should be ;
Try with it instead of ,

 

[shabbysquire]

Yeah, eBay's File Exchange CSV is only comma separated ",". I tried semi-colon ";", but it drags all the line info into a single variable.

Still can't get this to grab the relevant columns.

 

[shabbysquire]

Ok, I thought I could match the CSV column header letters (i.e. A, B, F, I), but you can't.

Problem with the CSV file is that there are loads of unwanted columns, so I don't want to create a variable for each one.

Seems there's a bug - sometimes a comma separated csv is processed properly by ZP, other times it grabs every thing (the whole rows). Open Office is able to separate via commas, but ZP can't, so the file isn't corrupt, etc.

Update: Think I've found the issue. When I enable save changes to file in ZP, it screws up my CSV file somehow. And when I open it up with Open Office, all the rows are lumped all together. ZP is screwing up the file somehow.

 

[Nick]

(i'm not a support team member but still...)
That's strange because it normally works... it's ok in your case, just cut out non-required columns in OpenOffice Calc in advance, or save file into .xls, that might help. Otherwise, I would bind all columns and only use needed variables. Never tried binding only selected columns, that might be not that thouroughly tested and thus buggy.

 

[shabbysquire]

Thanks Nick, I will try some of those suggestions.