UNIX for Dummies Questions & Answers

I have grep MHz psrinfo-v.out

it gives

The i386 processor operates at 3000 MHz,
The i386 processor operates at 3000 MHz,
The i386 processor operates at 3000 MHz,
The i386 processor operates at 3000 MHz,



how to get instead of these 4 lines:
CPU speed: 3000 MHz

i.e. CPU speed: $6<space>$7<trailing comma removed>

Many thanks

Here's an ugly Perl one-liner:

cat temp.txt | perl -pe 's/^.*\s+(\d+\s+mhz).*$/CPU speed: $1/i' | sort -u

Many thanks

You're very welcome. I realize that I should probably explain it a bit.

Code:

cat temp.txt | perl -pe 's/^.*\s+(\d+\s+mhz).*$/CPU speed: $1/i' | sort -u

This uses a regular expression to find everything from the beginning of the line (^.*) up to one or more blank spaces (\s+), followed by one or more digits followed by a space and the letters "mhz" (\d+\s+mhz). It then captures everything else up to the end of the line (.*$). Surrounding the part I want in parenthesis lets me refer to it later with $1 (additional sections would be $2, etc.). It's case-insensitive because of the "i" after the final forward-slash.

Then, "sort -u" makes the output unique. So if all the lines are the same, there will only be one line of final output. If three lines were the same and one was different, there would be two lines of output -- one of each. Making the "sort -u" into "sort | uniq -c" would print out how many of each value there were.

ShawnMilo