How can I extract the value of an app.inventory.id command-line argument?

Question

I have a line with:

-Dapp.id=XXX -Dapp.name=YYY -Dapp.inventory.id=KKK -Dapp.inventory1.id=UNDEFINED -app.id=MMM

in an inventory.txt file. If I use the grep statement below, I get the entire line instead of only KKK:

grep -F app.inventory.id= /cme/apps/install/*/inventory/*.txt

I also tried flags like -Fx and -w, but nothing works. How can I get the just KKK using grep?

I know this is a crappy thing to say, but use `sed` or `awk` instead. It will let you control which part gets printed. `grep` will too with the `-o`, but it will include the `-D...` part too. — Mad Physicist, Mar 21 '16 at 15:25

score 0 · Answer 1 · answered Mar 21 '16 at 15:30

If you want to do this with just grep, you can use a grep with the PCRE library built in. For example:

$ pcregrep -o 'app.inventory.id=\K[^[:space:]]+' /tmp/corpus
KKK

Alternatively, if you don't have PCRE available you can use grep with extended regular expressions, and then use cut to get the correct field by using = as a field delimiter. For example:

$ grep -Eo 'app.inventory.id=\K[^[:space:]]+' /tmp/corpus | cut -d= -f2
KKK

score 0 · Answer 2 · answered Mar 21 '16 at 15:32

if all the files have same types of output wich means :

-Dapp.id=VALUE -Dapp.name=VALUE -Dapp.inventory.id=VALUE -Dapp.inventory1.id=VALUE [...]

You can then use awk like this :

grep -F app.inventory.id= /cme/apps/install/*/inventory/*.txt | awk -F ' ' '{print $3}' | awk -F '=' '{print $2}

chthonicdaemon · Accepted Answer · 2016-03-21T17:24:33.087

0

I think sed is a bit nicer here:

sed -n 's/.*-Dapp.inventory.id=\([^ ]*\).*/\1/gp' /cme/apps/install/*/inventory/*.txt

If you want the filename you can do something like

for filename in /cme/apps/install/*/inventory/*.txt; do 
    sed -n -e 's/.*-Dapp.inventory.id=\([^ ]*\).*/'$filename': \1/gp' $filename; 
done

edited Mar 21 '16 at 17:24

answered Mar 21 '16 at 15:52

chthonicdaemon

19,180
2
52
66

Thanks for your help - worked great – user3232560 Mar 21 '16 at 16:25
i am getting the value of the sed, is it possible to get the filename that is sed is operated on? I have multiple .txt files in that path. help me please – user3232560 Mar 21 '16 at 16:42
I've added a way to get the filenames as well. – chthonicdaemon Mar 21 '16 at 17:25
Thanks for your help. – user3232560 Mar 21 '16 at 20:16

pce · Answer 4 · 2016-03-21T16:04:30.627

0

grep -Eo "\s-Dapp.inventory.id=[A-Z]+\s" inventory.txt | cut -d '=' -f2

grep

-E use extended regex,
-o output only the match
[A-Z]+\s matches Uppercase-Letters to the next whitespace

To see the match: grep --color -Eo "\s-Dapp.inventory.id=[A-Z]+\s" inventory.txt

cut

-d --delimiter=DELIM delimiter of fields,
-f --fields=LIST output only the selected fields

edited Mar 21 '16 at 16:04

answered Mar 21 '16 at 15:57

pce

5,571
2
20
25

score 0 · Answer 5 · answered Mar 21 '16 at 16:10

0

awk to the rescue!

$ echo -Dapp.id=XXX -Dapp.name=YYY -Dapp.inventory.id=KKK -Dapp.inventory1.id=UNDEFINED -app.id=MMM | 
  awk -v RS=' +' '/app.inventory.id/{split($0,a,"=");print a[2]}'

KKK

answered Mar 21 '16 at 16:10

karakfa

66,216
7
41
56

score 0 · Answer 6 · answered Mar 21 '16 at 17:58

0

awk -F '[= ]' '{print $6}' file
KKK

answered Mar 21 '16 at 17:58

Claes Wikner

1,457
1
9
8

How can I extract the value of an app.inventory.id command-line argument?

6 Answers6