Shell Programming and Scripting

LOL, no worries.

The command that I first posted will apply to all uncommented lines. So, suppose you have a file like this:

Code:

OTHER_STUFF_1=foo
MIN_PASSWORD_LENGTH=7
#MIN_PASSWORD_LENGTH=2
OTHER_STUFF_2=bar
MIN_PASSWORD_LENGTH=3
OTHER_STUFF_3=baz
OTHER_STUFF_4=qux
#MIN_PASSWORD_LENGTH=6
MIN_PASSWORD_LENGTH=8
MIN_PASSWORD_LENGTH=11

The grep portion of the command given earlier will filter out everything but:

Code:

MIN_PASSWORD_LENGTH=7
MIN_PASSWORD_LENGTH=3
MIN_PASSWORD_LENGTH=8
MIN_PASSWORD_LENGTH=11

So, those 4 lines will be passed on to the awk statement. So, the output of the original command (adding in the -F= part...) would be:

Code:

Value is good
Danger, Will Robinson!
Value is good
Value is good

So, that's what it would do with multiple matching lines. The real question, though, is what do you want it to do?

well, in that case I only want the last
MIN_PASSWORD_LENGTH=11
how can I tell grep to give me the last value only?
Thanks Andrew!
K.

Ah, good question. There are a couple of ways. On some systems, you can use either the 'tac' (that is, 'cat' backwards) or the tail -r commands. That takes the input and, in a manner of speaking, simply flips it upside down. Then, the command 'head -n 1' keeps only the first line. I mention that only because it's nice to have in the command line toolbox. In your situation, though, I'd probably use this approach:

Code:

grep ^MIN_PASSWORD_LENGTH security | awk -F= 'END { if ($2 > 6) print "Value is good"; else print "Danger, Will Robinson!" }'

For testing, though, you might want to have it include the actual line in the output so you can be sure that it's giving you what you want; then just use the command above once that's proven:

Code:

grep ^MIN_PASSWORD_LENGTH security | awk -F= 'END { if ($2 > 6) print $0 " Value is good"; else print $0 " Danger, Will Robinson!" }'

Hope that helps.

cool, I got the tail and head commands ( and piping it as output from the grep command ) but what does END in your example do? That's the first time I used it
Thanks Andrew!
K.

hm, one more thing, what about if the MIN_PASSWORD_LENGTH is not at the beginning of the line ( if there are white spaces at the beginning ... ) ? then your grep won't return anything, how could I fix this?

Thank you
K.

I'd bet that there are loads of ways to do this, probably searchable with the keywords "remove leading spaces" or "remove leading whitespace". My usual way to do it is:

Code:

sed 's/^\ *//'

So, this full expression should give you what you need:

Code:

sed 's/^\ *//' security | \
grep ^MIN_PASSWORD_LENGTH security | \
awk -F= 'END { if ($2 > 6) print "Value is good"; else print "Danger, Will Robinson!" }'

You might want to go googling for a while to see if there are expressions to remove leading whitespace that better meet your needs. One should start to get a little nervous as more and more external commands are piled on. On the other hand, if the performance is satisfactory, go for it.

Regarding the END expression in the awk statement... There are both BEGIN and END in awk. BEGIN essentially provides to awk a set of instructions that it is to run before it looks at whatever it will be operating on. END provides instructions that are evaluated after the arguments are evaluated. The practical effect of the END statement in this case, though, is that only the last relevant line ("relevant" meaning those lines that got through the sed and grep filtering) is processed to stdout. My understanding is that the last line, and only the last line, is kept by awk and remains available even to instructions with the END section. An awk wizard might be able to shed more light on the particulars.

BEGIN is illustrated here:

Code:

sed 's/^\ *//' security | \ 
grep ^MIN_PASSWORD_LENGTH security | \
awk -F= 'BEGIN {print "Are you ready?"} ; END { if ($2 > 6) print "Value is good"; else print "Danger, Will Robinson!" }'

Awk is being told to execute the command {print "Are you ready?"} before it even looks at the lines being passed by the sed and grep combo.

Great! Thank you so much Andrew, I learned so much!
K.