Comparision of old and files for new updated records and to insert new records in old files - shell

Question

I want to compare a file which comes with new records, to the same structured file which already have records and exists in the server.
Like old file is file1:
SKU PROD NAME CODE PRICE
A001 shirt jm s02 478
B002 jean jn j02 348
C003 mwear mw m02 567
New file which comes with new record is:
SKU PROD NAME CODE PRICE
A001 shirt jm s02 680
m01 mwear mw m02 567
c02 kurta kr k04 677
d12 dr d3 d03 400
Based on the new records, either old records can be updated or new records can be appended after the old records.
I need to write a unix shell sript for the above scenerio. Please help

Answer 1

Look for the different SKUs in the new file first. If a SKU is not found there, get it from the old file.
grep SKU file1.txt > out.txt
cat file1.txt file2.txt |\
grep -v SKU | cut -d" " -f1 | sort | uniq | while read sku
do
grep $sku file2.txt >> out.txt
if [[ $? -eq 1 ]] # If SKU was not in file2.txt
then
grep $sku file1.txt >> out.txt
fi
done
Result:
SKU PROD NAME CODE PRICE
A001 shirt jm s02 680
B002 jean jn j02 348
C003 mwear mw m02 567
c02 kurta kr k04 677
d12 dr d3 d03 400
m01 mwear mw m02 567

Answer 2

Look at the problem in reverse : your new file is basically your newest data. You could simply replace the old with the new, except that there may be some old data that is not in your new file.
So, a lazy way to do this might be to read your old data, and look for stuff that isn't in your new file, and add it if you need to.
Something like this (off the top of my head, couldn't test it)
while read old_data
do
# find the key
sku=$(echo $old_data | cut -d\ -f1)
# Look for key in new data, paste the line if it's not there.
grep sku new_file > /dev/null || echo old_data >> new_file
done < old_file