Hi,

Have an issue with some scripting:
*******************************************************
#!/bin/bash
proj=w
cd /mnt/$proj/_04Images/
cat tmp00.txt
while read line
do
grep -H aaa ..etc etc.. >> tmp01.txt
grep -H bbb ..etc etc.. >> tmp01.txt
cat tmp01.txt | tr -d '\n' | awk '{printf ("%s", $0)}'>> tmp02.txt
cat tmp02.txt | awk 'BEGIN {i=(((18000*cos(-1*(atan2(8196908-8148966,585323.3-576355.2)))+576355.2))); printf (" %s\n", i)}'>> tmp02.txt

rm -f tmp01.txt
done <tmp00.txt
*******************************************************

Script runs fine & generates a multi-column file tmp02.txt:
FileName09.tif.ccc -1 -511 48799 261 0 576355 8148966 48795 585323 8196908 579665
FileName33.tif.ccc -1 -511 48799 261 0 576355 8148966 48795 585323 8196908 579665

*******************************************************
From above, I am manually generating the last column (with the atan2 function) 579665 & adding in newline (for next do-done loop).
The 579665 value is made (above cos/atan2 function) with columns of tmp02.txt (after line: cat tmp01.txt | tr -d '\n' | awk '{printf ("%s", $0)}'>> tmp02.txt)

What I'm trying to do is replace the manual numbers with that of the file to make the 579665 value. E.G. wanting to change the script line from:

cat tmp02.txt | awk 'BEGIN {i=(((18000*cos(-1*(atan2(8196908-8148966,585323.3-576355.2)))+576355.2))); printf (" %s\n", i)}'>> tmp02.txt

to

cat tmp02.txt | awk 'BEGIN {i=(((18000*cos(-1*(atan2($11-$8,$10-$7)))+$7))); printf (" %s\n", i)}'>> tmp02.txt

But the above does not input that column (e.g. $11 should be 8196908 from file tmp02.txt), but I think is either thought of as number 11 (or 0). Have tried to do other things (e.g.
using the "cut" to extract the 8196908
cat tmp02.txt | awk 'BEGIN {i=(((18000*cos(-1*(atan2($(cut tmp02.txt -f 11)-8148966,585323.3-576355.2)))+576355.2))); printf (" %s\n", i)}'>> tmp02.txt


But it keep failing.

Any straight-forward fix?

I know my script is probably not efficient etc. But am wanting to just try and get it to work

The BEGIN block in awk gets executed before the first line of the file is read. $0 and the field variables ($1, $2, etc.) simply don't exist at this point. Get rid of BEGIN
The script definitely looks over-complicated. It would be easier, if you provided some context, a sample of input data, and explained what you're trying to achieve. Then somebody here could suggest how to improve it.

Particularly,
looks redundant as tr and awk seem to do the same—remove newline characters.

I guess these two lines
could be replaced with one awk expression. And using the FILENAME variable in awk could probably eliminate the two grep lines before that as well.

1 members found this post helpful.