[SOLVED] Remove a word after a specified word

CSVvenu · 02-10-2014, 09:47 AM

Hi All,

I have a file with words like this.

M0 FFB OBRSS 7225 w=1u l=1u
M1 FFC OBRSS 9253 w=1u l=1u
M1 FFC MOSFP 92365 w=1u l=1u
...etc

I want to remove number after OBRSS. i.e

M0 FFB OBRSS w=1u l=1u
M1 FFC OBRSS w=1u l=1u
M1 FFC MOSFP 92365 w=1u l=1u
..etc

Could you please help me?

Thanks,
Venu.

schneidz · 02-10-2014, 09:47 AM

what have you tried and where are you stuck ?

awk and sed seem like likely solutions.

CSVvenu · 02-10-2014, 09:52 AM

I don't know much about linux. I just searched in web but haven't found any query related to this. So I just raised query in this forum.

schneidz · 02-10-2014, 10:00 AM

i think you would want to reserch the sed command and its substitute functionality (especially how it operates with regular expressions).

CSVvenu · 02-11-2014, 10:36 PM

I tried sed command. But haven't get it. Please let me if any one knows how to get it..

kooru · 02-12-2014, 12:13 AM

Quote:

Originally Posted by CSVvenu

I tried sed command. But haven't get it. Please let me if any one knows how to get it..

Could you post what you tried?
Anyway if you search into this forum, you will find several posts about your same issue.

pan64 · 02-12-2014, 12:43 AM

here is a tip:

Code:

sed 's         # stands for substitute
/OBRSS digits  # look for the given string and the numbers
/OBRSS         # replace it with OBRSS
/g             # flags if required
' inputfile > outfile

see man sed to complete it

danielbmartin · 02-12-2014, 07:15 AM

Quote:

Originally Posted by CSVvenu

Hi All,

I have a file with words like this.

M0 FFB OBRSS 7225 w=1u l=1u
M1 FFC OBRSS 9253 w=1u l=1u
M1 FFC MOSFP 92365 w=1u l=1u
...etc

I want to remove number after OBRSS. i.e

M0 FFB OBRSS w=1u l=1u
M1 FFC OBRSS w=1u l=1u
M1 FFC MOSFP 92365 w=1u l=1u
..etc

Could you please help me?

Thanks,
Venu.

If OBRSS does appear in a line, will it always be the third word?

Daniel B. Martin

danielbmartin · 02-12-2014, 11:23 PM

With this InFile ...

Code:

M0 FFB OBRSS 7225 w=1u l=1u
M1 FFC OBRSS 9253 w=1u l=1u
M1 FFC MOSFP 92365 w=1u l=1u

... this awk ...

Code:

awk '{for (j=1;j<=NF-1;j++) {if ($j=="OBRSS") $(j+1)=""}; print}' $InFile >$OutFile

... produced this OutFile ...

Code:

M0 FFB OBRSS  w=1u l=1u
M1 FFC OBRSS  w=1u l=1u
M1 FFC MOSFP 92365 w=1u l=1u

Note the (possibly insignificant) flaw: a double blank after OBRSS.

Daniel B. Martin

danielbmartin · 02-12-2014, 11:28 PM

With this InFile ...

Code:

M0 FFB OBRSS 7225 w=1u l=1u
M1 FFC OBRSS 9253 w=1u l=1u
M1 FFC MOSFP 92365 w=1u l=1u

... this awk ...

Code:

awk '{p1=index($0,"OBRSS ")+5;
      if (p1>5) 
        {str1=substr($0,1,p1-1)
         str3=substr($0,p1+index(substr($0,p1+1)," "))
         $0=str1 str3}
      print}' $InFile >$OutFile

... produced this OutFile ...

Code:

M0 FFB OBRSS w=1u l=1u
M1 FFC OBRSS w=1u l=1u
M1 FFC MOSFP 92365 w=1u l=1u

Daniel B. Martin

CSVvenu · 02-14-2014, 06:55 AM

Thanks very much Daniel B. Martin, Your code is working. And thanks kooru,pan and schneidz for your useful suggitions.

schneidz · 02-14-2014, 08:33 AM

Quote:

Originally Posted by schneidz

what have you tried and where are you stuck ?

awk and sed seem like likely solutions.

i was waiting for a little effort from the op but this is what i came up with:

Code:

[schneidz@hyper ~]$ cat csvvenu.txt 
M0 FFB OBRSS 7225 w=1u l=1u
M1 FFC OBRSS 9253 w=1u l=1u
M1 FFC MOSFP 92365 w=1u l=1u
[schneidz@hyper ~]$ sed s/"OBRSS [0-9]* "/"OBRSS "/g csvvenu.txt 
M0 FFB OBRSS w=1u l=1u
M1 FFC OBRSS w=1u l=1u
M1 FFC MOSFP 92365 w=1u l=1u

CSVvenu · 02-14-2014, 09:26 AM

Thanks schneidz, I followed your hint and then I was able to replace OBRSS but not numbers. Now I understand how to replace it. Previously if I spent much time on this then I can able to get this. But i don't have much time because I am working on a project which is tight scheduled. Sorry for this and thanks very much for your hint and answer.

danielbmartin · 02-15-2014, 01:20 PM

Quote:

Originally Posted by schneidz

Code:

 sed s/"OBRSS [0-9]* "/"OBRSS "/g csvvenu.txt

Now, just for fun, let's generalize this problem. Instead of saying we want to delete the word following OBRSS, let's say we want to delete the word following any word contained in a second input file.

With this InFile1 ...

Code:

M0 FFB OBRSS 7225 w=1u l=1u
M1 FFC OBRSS 9253 w=1u l=1u
M1 FFC MOSFP 92365 w=1u l=1u
M2 FFB OBRSS 8336 w=1u l=1u
M3 FFC OBRXX 5454 w=1u l=1u
M4 FFC MOSFP 42424 w=1u l=1u

... and this InFile2 ...

Code:

OBRSS
OBRXX
OBRYY

... this code ...

Code:

sed 's|\(.*\)|s/\1 [0-9]* /\1 /g|' $InFile2  \
|sed -f - $InFile1 >$OutFile

... produced this OutFile ...

Code:

M0 FFB OBRSS w=1u l=1u
M1 FFC OBRSS w=1u l=1u
M1 FFC MOSFP 92365 w=1u l=1u
M2 FFB OBRSS w=1u l=1u
M3 FFC OBRXX w=1u l=1u
M4 FFC MOSFP 42424 w=1u l=1u

Daniel B. Martin