AWK(ward) pattern matching problem?

fanoflq · 03-03-2017, 12:44 PM

Code:

#works as expected
 ~/dir $ awk '{ print  $0}' inventory-shipped                                                                    
Jan 13 25 15 115
Feb 15 32 24 226
Mar 15 24 34 228
Apr 31 52 63 420
May 16 34 29 208
Jun 31 42 75 492 lastCol
Jul 24 34 67 436
Aug 15 34 47 316
Sep 13 55 37 277
Oct 29 54 68 525
Nov 20 87 82 577
Dec 17 35 61 401

# Match any word that has a "J". Works as expected
 ~/dir $ awk '/J/ { print  $0}' inventory-shipped 
Jan 13 25 15 115
Jun 31 42 75 492 lastCol
Jul 24 34 67 436

 ~/dir $ awk '/J.*/ { print  $0}' inventory-shipped 
Jan 13 25 15 115
Jun 31 42 75 492 lastCol
Jul 24 34 67 436

# /J*/ matches any word/line containing J, JJ, JJJ, JJJJ, ...  
#Does NOT work as expected.
#Incorrect output: 
#There should be ONLY outputs for lines with J, JJ, JJJ, JJJJ, ... 
 ~/dir $ awk '/J*/ { print  $0}' inventory-shipped 
Jan 13 25 15 115
Feb 15 32 24 226
Mar 15 24 34 228
Apr 31 52 63 420
May 16 34 29 208
Jun 31 42 75 492 lastCol
Jul 24 34 67 436
Aug 15 34 47 316
Sep 13 55 37 277
Oct 29 54 68 525
Nov 20 87 82 577
Dec 17 35 61 401

What did I missed?

kakistocrat · 03-03-2017, 01:01 PM

J* means any line with zero or more J's. See

Code:

man 7 regex

and then scroll down to "An atom followed by '*' matches a sequence of 0 or more matches of the atom."

Maybe you mean a plus + instead.

Code:

awk '/J+/' inventory-shipped

By the way, you can leave off print $0 since it goes without saying. There are a lot of shortcuts like that in awk

fanoflq · 03-03-2017, 01:02 PM

Thank you.

kakistocrat · 03-03-2017, 01:06 PM

No worries. Also, you can anchor the search to the beginning of the line with a caret ^ at the start of the pattern:

Code:

awk '/^J+/' inventory-shipped

That will match

Code:

Jan 13 25 15 115
Jun 31 42 75 492 lastCol
Jul 24 34 67 436

but not

Code:

aJn 13 25 15 115
uJn 31 42 75 492 lastCo
Aug 24 34 67 436 J

szboardstretcher · 03-03-2017, 01:11 PM

Quote:

An atom followed by '*' matches a sequence of 0 or more matches of the atom.

I must be reading this wrong.

If I have a string equal to: 1111111111111111111111111

And I search for '/A*/' it will not 'match' and will not exit as successful. Even though there are 0 occurrences. So which is it? Does '/A*/' successfully match when there are no matches (0) or not?

kakistocrat · 03-03-2017, 01:17 PM

Quote:

Originally Posted by szboardstretcher

I must be reading this wrong.

If I have a string equal to: 1111111111111111111111111

And I search for '/A*/' it will not 'match' and will not exit as successful. Even though there are 0 occurrences. So which is it? Does '/A*/' successfully match when there are no matches (0) or not?

Which versions of awk are you using? It should match all the lines. It works for me on GNU Awk, OpenBSD's Awk, and Mawk, just to check three.

szboardstretcher · 03-03-2017, 01:26 PM

My bad: My eyes didn't detect that i had '/AA*/' and not '/A*/'

I blame the writers of awk of course.