UNIX for Dummies Questions & Answers

Trying to use SED to replace numbers that fall into a range but can't seem to get the logic to work and am wondering if SED will do this. I have a file with the following numbers

Code:

3 
26 
20 
5

. For the numbers that are greater than zero and less than 25, SED would add the word range after the number. Output would look something like this:

Code:

 3 range 
26 
20 range
5 range

. Have tried variations of this:

Code:

sed 's/[0-25]/range/g' file

. Can't get this logic to work. Any help would be appreciated.

Hi jimmyf,

Following awk may help you in same.

Code:

awk '($1+0 <= 25) {print $1 OFS "Range";next} 1'  Input_file

Output will be as follows.

Code:

3 Range
26
20 Range
5 Range

EDIT: Above code will take 0 too, to avoid taking 0 use following.

Code:

awk '($1+0 >0 && $1+0 <= 25) {print $1 OFS "Range";next} 1' Input_file

Thanks,
R. Singh

/[0-25]/ means 0-2 or 5, so it matches any line with one of the numbers: 0,1,2 or 5 (BTW In your requirements you say greater than 0, but your example would also match 0)

To match the numbers 1-25, you could use: /[1-9][0-5]\{0,1\}/

But to avoid matching "125" of example you need to anchor it front and back, for example:
/^ *[1-9][0-5]\{0,1\} /
Notice the trailing space ...

Another approach is not to use regex, but a numerical comparison. But for this you cannot use sed, which is a streamline editor. Instead you could use awk for example or another scripting language...

Code:

awk '$1>0 && $1 <=25' file

That's not the desired output.

Hi jimmyf,

I just now edited my code in post#2, kindly refer the same.

Thanks,
R. Singh

Scrutinizer, your logic works nicely but is there a way to also echo the original numeric value?

---------- Post updated at 01:34 AM ---------- Previous update was at 01:27 AM ----------

Thank you very much guys.

You lost me. No numbers are being changed. Why do you want to echo the original number and echo the original number plus the word range?

If you want to use sed instead of awk for this, try:

Code:

sed -e 's/^ *[1-9] *$/& range/' -e 's/^ *1[0-9] *$/& range/' -e 's/^ *2[0-5] *$/& range/' file

Note that using:

Code:

sed 's/^ *[1-9][0-5]\{0,1\} /&range/' file

will not match any line that does not have a trailing space, and, for lines that do have a trailing space: