UNIX for Dummies Questions & Answers

Do you mean to use this as the shell interpreter?

#!usr/xpg4/bin/awk

I did this and it nows interprets the script name as a command (which it cna't find)

No, I mean to use /usr/xpg4/bin/awk wherever you would use awk if you are using Solaris.

It doesn't seem to like that. When I execute the script "CHANGE_TEST_NUMS" I get the following error:

/home/test/bin/CHANGE_TEST_NUMS: usr/xpg4/bin/awk: No such file or directory

If you use awk in the shebang you need to use awk -f, but as I indicated in red, you left out the leading slash in the path to awk, so it cannot be found...

Ok the error is gone now but the script does nothing. I run my script on a text file called a.txt:

bash-2.03$ CHANGE_TEST_NUMS a.txt

the script syntax is:

Code:

#!/bin/bash -f
#
echo $1
/usr/xpg4/bin/awk -v start=5 -v end=9 '!($1 >= start && $1 <= end){
if($1 ~ /[0-9]{1,}/ && $1>end)
 sub(/[0-9]{1,}/,($1-(end-start)-1))
print
}' inputfile

However the script does nothing to the text file a.txt

Thanks, I have the data manipulation part working. However I need to save this to a new file. I'm using the sed program to created a new file called a_mod.txt (the original is called a.txt). Also I'm piping the output to this new file. Here's my code:

Code:

#!/bin/bash
#
OUTPUT_FILENAME = `echo $1 | sed 's/.txt/_mod.txt/'`

/usr/xpg4/bin/awk -v start=4 -v end=5 '!($1 >= start && $1 <= end){
if($1 ~ /[0-9]{1,}/ && $1>end)
 sub(/[0-9]{1,}/,($1-(end-start)-1))
print
}' $1 > $OUTPUT_FILENAME

I'm getting the following error.

/home/test/bin/CHANGE_TEST_NUMS: OUTPUT_FILENAME: command not found
/home/test/bin/CHANGE_TEST_NUMS: $OUTPUT_FILENAME: ambiguous redirect