[vox-tech] Changing data with awk

[Dylan Beaudette]

> I have a large flat file generated by SQL Loader that I'd like to mess
> around with; specifically, I'd like to replace all of the carriage returns
> in one field with some other character, since they're messing up my data
> load.
>
> I figured I'd use awk, since it's a pretty powerful little tool for
> getting right to the data.  If I use:
>
> $ awk -F^ {print $6} test.dat
>
> I get the field that I want.  But how do I change the characters in that
> field and replace them in test.dat?


i recently had a similar problem: trying to convert
this:
1<CR>
2<CR>
3<CR>
...

into this: 1, 2, 3...

here is how i did it:

append a comma+space to the end of each line with sed
then remove each CR using tr:

sed -e 's/$/, /g' input_file | tr -d "\n" > output_file

so something like this might do the trick:

awk -F^ {print $6} test.dat | sed -e 's/$/, /g' | tr -d "\n" > output_file


.. you would be left with one column of data that would have to be
re-instered into the DB, or added back to the original file.

the command 'paste' might be helpful for appending the data to the
original...


good luck!

Dylan