edit file to extract certain lines only - sed probably 1

[kasparov]

I want to automatically edit the file produced from an Oracle database by running this:

alter database backup controlfile to trace;

so that I only get the first definition of the control file - i.e. the first part that begins

STARTUP NOMOUNT
CREATE CONTROLFILE REUSE DATABASE "OUT_DATABASE" NORESETLOGS  NOARCHIVELOG
    MAXLOGFILES 5
    MAXLOGMEMBERS 5

and ends

  '/our_database/indx03.dbf',
  '/our_database/indx04.dbf',
  '/our_database/cftest_LOB_TS.dbf',
  '/our_database/cftest_tools01.dbf'
CHARACTER SET WE8ISO8859P1
;

This section appears twice in the file (with small differences) and I only want the first occurrence. Both occurrences begin with 'STARTUP NOMOUNT' and end with ';'.

I've tried this (UUOC I know )

cat control.sql | sed -n '/^STARTUP NOMOUNT/,/^;/p'

which gives me both occurrences (and correctly strips everything else out). But I don't know how to only get the first occurrence.

I can't be certain this will always be on the same lines so I can't print certain line numbers only. I could hack away at removing the 2nd occurrence of the lines but there must be a simpler way to do it.

Hope this makes sense to you all. Any advice appreciated as usual.

Thanks, Chris

 

[feherke]

Hi

sed -n '/^STARTUP NOMOUNT/,/^;/[highlight]{[/highlight]p[highlight];/^;/q}[/highlight]' /input/file

Tested with GNU [tt]sed[/tt].

Feherke.

http://free.rootshell.be/~feherke/

 

[kasparov]

Thanks for the reply feherke

Unfortunately I'm getting

sed: command garbled: /^STARTUP NOMOUNT/,/^;/{p;/q}

I'm on Solaris 10. What exactly should

{p;/^;/q}

do? I'm guessing it's:
p - print
/^;/q - don't print after the semi-colon

That's what I'm aiming for anyway.

Cheers, Chris

 

[feherke]

Hi

Try to add more semi-colons ( ; ), for example :

sed -n '/^STARTUP NOMOUNT/,/^;/{p;/^;/q[highlight];[/highlight]}' /input/file

Feherke.

http://free.rootshell.be/~feherke/

 

[kasparov]

God bless the semi-colon

It works . Thanks you feherke