usinf sed

[Henry Chua]

Hi Guys,

I'll need to extract data from the following field "WJV82LV83A1.ICZ34K.ws". The data I need is captured between W and the . , i.e. JV82LV83A in this case. as the data may differ in lenght i tried using a combo of awk and sed to do it.. but the script got quite bulky though it did the job., Is there a easier way. I'll need to reduce the speed and alot of data to capture. any advise?

thank u

[Olivier LEGRAND]

Hi

echo "WJV82LV83A1.ICZ34K.ws"|awk -F"." '{print $1}'|cut -c2-

Best regards

[Steve Steel]

Hi


Use the SHELL variable IFS and cut

Or cut with the -d option if the
data field format is stable

echo "WJV82LV83A1.ICZ34K.ws". |cut -f1 -d"."|cut -f2 -d"\""

Steve Steel

[Muthukumar_5]

Try this as,
echo "WJV82LV83A1.ICZ34K.ws"| awk -F"." '{ print substr($1,2); }'

HTH.

[Gordon Morrison_1]

Hi Henry,
Assuming you always want to drop the first character and keep everything after it up until the first dot (.) then:

str1="WJV82LV83A1.ICZ34K.ws"
tmp=${str1#?}
echo $tmp
JV82LV83A1.ICZ34K.ws
str2=${tmp%%.*}
echo $str2
JV82LV83A1

[Muthukumar_5]

IF you are date contains W*****.I** then, it is easy to do with sed as,

echo "WJV82LV83A1.ICZ34K.ws"| sed 's/^W\(.*\).I.*/\1/'

HTH.

[Henry Chua]

Hi guys,

Thanks for the input, my case is alittle more troublesome, as the I only want the data between W and the 1, and as the number of characters for the data differs from time to time it may be alittle difficult to achieve. I have tried using a script to do it.. but its really bulky.. do u guys have any idea?


Thank u!!

[Gordon Morrison_1]

Hi Henry,
If you just want everything between (but not including) the fiorst character on the line, and the last character before the first dot, then:

str1="WJV82LV83A1.ICZ34K.ws"
tmp=${str1#?}
echo $tmp
JV82LV83A1.ICZ34K.ws
str2=${tmp%%?.*}
echo $str2
JV82LV83A

[Hein van den Heuvel]

Hmmm, maybe there is something you are stil not specifying, as at first observation all posted suggestion would work, and would allow for a variable match count.

Here is yest another awk solution:

echo WJV82LV83A1.ICZ34K.ws | awk '{print substr($0,2,index($0,".")-2)}'

Or if the W is not not really only the first character as sugggested, but the fist W then:

echo WJV82LV83A1.ICZ34K.ws | awk '{w=index($0,"W")+1;print substr($0,w,index($0,".")-w)}'

With perl, you don't get the 'newline'.
It could look like:

echo WJV82LV83A1.ICZ34K.ws | perl -ne 'print $1 if (/W([^.]+)/)'

the ERE says: When you see a W, start to remember "(" one or more "+" non-periods "^."

In this case it looks like you could also write:
echo WJV82LV83A1.ICZ34K.ws | perl -ne 'print $1 if (/W([A-Z0-9]+)/)'

Here the ERE says: one or more characters from the A-Z range or 0-9 range.

With a perl solution you cn often do much more of the final processing all rigt there in perl.

Anyway... what specification detail are we missing?

Cheers,
Hein.

[Henry Chua]

Thanks for the help Guys.. Hello Hein, actually, what I meant was to capture "JV82LV83A" out of "WJV82LV83A1.ICZ34K.ws".. I have tried the suggestion and it gives "JV82LV83A1" instead, is there anyway I can get rid of the 1? .. "JV82LV83A" is the data i wanted, and as there are no fix pattern or number to it, I could not just cut out the string... can this be done?

sorry for the trouble..

thank u!!
Henry

[Olivier LEGRAND · Accepted solution]

Hi
echo "WJV82LV83A1.ICZ34K.ws"|awk -F"." '{print $1}'|cut -c2-|sed s/.$//

Best regards
Olivier

[Henry Chua]

thanks Olivier, thats wat I am lookin for!!

thanks guys!!