grep not matching EXACTLY what I tell it to 1

[630111]

I have a text file...

usdalnpd01 PROD_WINDOWS
usdalphm01 PROD_WINDOWS
usdalqdb01 NON_PROD_WINDOWS
usdalsvc01 NON_PROD_WINDOWS
usdaluat01 NON_PROD_WINDOWS

When I do this...

cat text.file | grep PROD_WINDOWS

I get both the PROD_WINDOWS and NON_PROD_WINDOWS entries.
I also tried

grep 'PROD_WINDOWS' text.file

and the same thing happens. I only want the lines that are PROD_WINDOWS. Is grep not the right tool, or is there some switch I'm missing?

Thanks, in advance!
630111

 

[PHV]

Have you tried this ?
awk '$2=="PROD_WINDOWS"' text.file
Or this ? ([<Space><Tab>])
grep '[ ]PROD_WINDOWS' text.file

Hope This Helps, PH.
Want to get great answers to your Tek-Tips questions? Have a look at FAQ219-2884 or FAQ222-2244

 

[duncdude]

I'm only really competent in Perl (just) but this is how i'd do it - and it does work on the command line:-

cat text.file | grep "[^\_]PROD_WINDOWS"

returns:-
[red]usdalnpd01 PROD_WINDOWS
usdalphm01 PROD_WINDOWS[/red]

grep is doing what it is told as all the lines contain that string. you need to tell grep NOT - as in this example - to be preceeded by an underscore




Kind Regards
Duncan

 

[vgersh99]

grep '[^A-Za-z_]PROD_WINDOWS' file

vlad
+----------------------------+
| #include<disclaimer.h> |
+----------------------------+

 

[duncdude]

PHV - quick as lightning!

630111 - I'd listen to PHV - he DOES know what he/she is talking about!!!


Kind Regards
Duncan

 

[guggach]

my solaris grep knows a 'w' option, and sed always works
/\<abc\>/p


don't forget, RTFMP guggach