First line and replace

fs11 · 04-24-2008, 03:40 PM

Hello All,

I have a small problem, i would like to solve with bash scripting.(sed grep anything..).

I have a file and i want to search for the start of the file..if the pattern is not present and then to delete those lines..

forexample:

---------
>SEQ216
------------------------------------------------------------
------------------------------------------------------------
--------------------GLIHGGT--KAAPLNEGA----------------------
------------------------------------------------------------
---------------------------------------------------YELNQCASI
---------
>SEQ250
------------------------------------------------------------
------------------------------------------------------------
--------------------GTIYGFN--VMGVVCLYG----------------------
------------------------------------------------------------
---------------------------------------------------FEHDLCDAV
---------
>SEQ251
------------------------------------------------------------
------------------------------------------------------------
--------------------GVMKGAN--TMGVVCLYG----------------------
------------------------------------------------------------
---------------------------------------------------FKHDLCDAI
---------
>SEQ252
------------------------------------------------------------
------------------------------------------------------------
--------------------GVVSGGH--TMGVVCLYG----------------------
------------------------------------------------------------
---------------------------------------------------YEIERCDEI
---------

$$$$$$$$$$$$$$$$$$$$$

I would like that the first element in the file is a '>'..If this is not the case , then delete this line....

Any help would be highly appreciable.

Thanks in advance..

jschiwal · 04-24-2008, 05:46 PM

Code:

sed '1{/^>/!d}' file >newfile
or
sed -i '1{/^>/!d}' file

The "1" selects the first line. The "/^>/!" test passes if the line doesn't start with "<". If that is the case the line is deleted.

fs11 · 04-24-2008, 06:04 PM

The code doesn't seem to work.

I am still getting all the lines in the file.
Please help !

jschiwal · 04-25-2008, 12:17 AM

Maybe I misunderstood what you want to do. From your description, you wanted to delete line 1 if it doesn't begin with at "<" character.

Code:

sed '1{/^>/!d}' testfile >newfile
jschiwal@hpmedia:~> head newfile
>SEQ216
------------------------------------------------------------
------------------------------------------------------------
--------------------GLIHGGT--KAAPLNEGA----------------------
------------------------------------------------------------
---------------------------------------------------YELNQCASI
---------
>SEQ250
------------------------------------------------------------
------------------------------------------------------------

Also, you probably should have put your sample in [ code ] blocks to preserve formatting. I think that as posted, the lines are split up so I couldn't test it on a read sample. The sample you posted has 32 lines, and the first was deleted because it didn't start with a "<".

If you want to delete all lines until one starts with a "<", then:
sed '1,/^>/{/^>/!d}' testfile