Shell Programming and Scripting

*bump*.

After last post i decided to add some extra cleanup lines in the form of all lower cases etc. However i seemed to find some lines that dont like to work with the comparing.

In the output file i have the following lines:

Code:

4,25E+12   P00043585   242457   8,4E+10   90   19.40   ekl   JAC   alpenfohn
4,25E+12   P00052963   1577936   8,4E+10   18   12.50   ekl   JAC   alpenfohn
4,25E+12   P00052964   2001285   8,4E+10   18   12.00   ekl   JAC   alpenfohn

Dont look at the first and third column as i copied these over from excel so they look messed up. In the last field it says alpenfohn as manufacturer and according to the supplier it is ekl.

In my comparing file i have this line:

Code:

ekl,alpenfohn,alpenföhn

This should be an accepted compare and it should not have been in the compare file.

I have the idea that the script you provided only checks for the first name in the comma seperated list. Can you confirm this and possible have an idea how to get around that since this would mean i need to add double lines.

[EDIT]

For some reason i start to doubt that it also keeps tracks of the spaces in the names since i noticed that the lines below also dont get skipped (as being correct).

Code:

          
5,04E+12   P00055056   2085825   KD1250K-QS   0   169.00   black en decker   JAC   black en   decker       
5,04E+12   P00055017   2090770   KR805K   0   83.60   black en decker   JAC   black en decker

So we are initially told that we have three CSV input files with semicolons as the field delimiters, but now you are telling us that code RudiC suggested meeting those requirements does not work when one of the input files starts using spaces instead of semicolons as the field delimiters (even though some of the fields in that file contain spaces as data) and one of the input files starts using commas instead of semicolons as the field delimiter. And, instead of seven input fields we now have 10 input fields and we aren't supposed to look at two of those fields because they are messed up.

If your input data formats keep changing, it is not our job to act as your unpaid programming staff and rewrite code to meet your constantly changing input file formats.

If you would like to try modifying RudiC's suggestions to meet your new input file formats, can show us clear specifications for your new input file formats, and can explain where you are stuck trying to get your modifications to work correctly; we will be happy to help you.

If you would like us to guess at what your new input file format is for your third input file and guess at how we might modify RudiC's code to work with your new input file formats, I doubt that many of the volunteers reading your posts here will be interested in tackling that assignment.

I think we have a slight miscommunication here . All files use ; as delimiter. The linked parts above are just several lines which went through the script he wrote and that gives that output. The problem(s) i seem to have is that it ignores manufacturers which are not the first name in the list (all files are added to a previous post btw) and for some reason it does not seem to recognize several manufacters which have spaces like the ones above. The first part (the name in front) is pure to know if i need to rearrange the names but the second can be tricky since there are enough names that do have spaces. I could technicly remove all spaces but i am not sure if that would be a good solution.

The input files havent changed ever except for my miss on the column number for which i am sorry.

[Edit]
Again the code RudiC wrote is superb and the files havent changed. The problem is with the output it creates and most possible be a simple solution.

If you look at the script, you see that FS = OFS = ";", and that the respective fields are split($1, T, ",") by commas. ALL comma separated subfields are treated identical and equal, and spaces are NOT treated specially thus can be part of names.
So - I'd recommend to take a look at the input data and validate those. black en decker and black en decker are definitely NOT the same!

Thank you for that answer. Could you do comment on my question if the searched manufacturer name has to be first or does that not matter?