On Monday 06 November 2006 05:14, Andrew Sackville-West wrote: > On Sun, Nov 05, 2006 at 10:08:12AM -0800, Steve Lamb wrote: > > Andrew Sackville-West wrote: > > > On Mon, Nov 06, 2006 at 01:10:08AM +1100, John O'Hagan wrote: > > >> Or the whole thing could even be done with (I think!): > > >> > > >> #tr -d '\n' < IN | tr ' ' '\n' | grep -B1 Processor | grep -v > > >> 'Processor\|--'
[...] > > so that part strips all the newlines. the next tr replaces all spaces > with newlines so that each word is now on an individual line, then > grep -B1 Processor returns the line before any occurence of > Processor. not sure what the second grep does with that \-- > in it, but I assume it is supposed to eliminate any "Processor" result > when the input includes 'Processor Processor'. > The first grep -B1 returns two lines per result, "Processor" and the number we are after; also, grep -B separates each result with a "--". The second grep -v is just to remove the unwanted lines. E.g., if IN contains: junk info 18 Pro cessor ggjgh 34 Processor dgjhj 19 Processor yytyr fee 45 Processor tt 5 6 Proce ssor rgrge Then (after the newline business) grep -B1 Processor gives: 18 Processor -- 34 Processor -- 19 Processor -- 45 Processor -- 56 Processor Grep -v 'Processor\|--' gives just the numbers we want. Regards, John -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
