UNIX for Dummies Questions & Answers

Hi all,

I'm a new programmer to shell script... and I have no idea how to use substring.

I want to extract the numbers from the following string and place it into a variable:
"170 unique conformations found"

The numbers can be more than three digits depending on the case. I just want to ensure that I can isolate ALL the numbers in this string for every case (no matter how many digits there are). In other words, the programming should not depend on the number of digits of the number.

Thanks! Much appreciated by me and our laboratory at University of Richmond!!

assuming no other types of charachters then in your example

Code:

tr -d [:alpha:] < filename

Hi I'm new to unix scripting myself. I learned how to do this the other day, here goes:

Assuming you have assigned the string to a variable:

Code:

x="170 unique conformations found"

You can get the numerical part and assign it to the variable y with:

Code:

y=${x%%unique conformations found}

Test:

Code:

echo $y

I hope this helps

This is somewhat like a Perl equivalent to denn's approach:

Code:

echo "170 unique conformations found" | perl -ane 'foreach my $rec (@F) {print "$rec\n" if "$rec" !~ m/[A-Za-z]/;}'

To put the output in a variable:

Code:

var=$(echo "170 unique conformations found" | perl -ane 'foreach my $rec (@F) {print "$rec\n" if "$rec" !~ m/[A-Za-z]/;}')

Code:

var=$( echo "170 unique conformations found" sed "s/[^0-9]//g")

I think you're missing a '|'.

Thanks everyone! I ended up using the perl solution!

However I have one more question! Sorry...

I have one string that has two numbers i want to isolate SEPARATELY:

"Global minimum E = 391.86 found 10 times."

I want to isolate the "391.86" number preceding the "found" word, and store that into a variable.
I also want to isolate the "10" number preceding the "times" word, and store that into a separate variable.

If you guys could help me out again, it would help a great deal!!!

Thank so much!

-Alex