AWK: Update data in one file while looking into secod file....

[goyalpdm]

Hi,

Please see if someone can help me on this:

I have two files:
- file_1 (col_11, col_12) and
- file_2 (col_21, col_22, col_23 ...)

Of these, file_1 is master file and is used to update data in file_2.

How to update? Read first column of file_2 (col_21) and compare it with second column of file_1 (col_12). If the two match, update first column of file_2 (col_21) with first column of file_1 (col_11).

Exception: If no matching data is found, leave the record as is AND if multiple matching records are found, use first available match.

Thanks,
Sachin

 

[PHV]

A starting point:
nawk '
NR==FNR{if(!($2 in a))a[$2]=$1;next}
{if($1 in a)$1=a[$1];print}
' file_1 file_2 > newfile_2

Hope This Helps, PH.
Want to get great answers to your Tek-Tips questions? Have a look at FAQ219-2884 or FAQ222-2244

 

[goyalpdm]

I tried this in the command prompt.

It generated newfile_2 BUT exactly same as file_2. As I said, I like to update first column of file_2 with second column of file_1.

 

[PHV]

What is the column separator ?

Hope This Helps, PH.
Want to get great answers to your Tek-Tips questions? Have a look at FAQ219-2884 or FAQ222-2244

 

[goyalpdm]

Pipe "|"... Guess, I should have mentioned it in first place.

 

[PHV]

nawk '
BEGIN{FS=OFS="|"}
NR==FNR{if(!($2 in a))a[$2]=$1;next}
{if($1 in a)$1=a[$1];print}
' file_1 file_2 > newfile_2

Hope This Helps, PH.
Want to get great answers to your Tek-Tips questions? Have a look at FAQ219-2884 or FAQ222-2244

 

[goyalpdm]

Bingo. It works great !!!

Thanks PH,
Sachin

 

[goyalpdm]

Is it possible to avoid newfile_2? Ofcourse I can mv newfile_2 file_2 but it would be much cleaner if can avoid newfile_2 and update file_2.