Re: Matching word and save text line

* mailbox20112002-SED <[email protected]> [2013-06-14 18:09]:
> It possible with Sed, matching one word,
> or more word in single line: YES and saving,
> in other file the X lines previous,
> and the Y following lines?
> You have any example?

generally possible in some way?  yes.

but - does it *have* to be sed?   and why?
out out funds for a good operating system? ;)

because this really looks like a good example
for the capabilities that "grep" already has.

example1:

  $ X=3 Y=2 word=Prism file=/usr/share/dict/ngerman ofile1=results1.txt
  $ grep -B $X -A $Y $word $file > $ofile1

here are the relevant options for grep:

  -A, --after-context=NUM   print NUM lines of trailing context
  -B, --before-context=NUM  print NUM lines of leading context

the rest comes from the evaluation of your shell, ie
expansion of variables and redirection of output.

you can also use grep with an extended set of regular
expressions which allows alternative search patterns:

example2:

  $ X=3 Y=2 word1=Prism word2=arnev file=/usr/share/dict/ngerman ofile2=results2.txt
  $ grep -E -B $X -A $Y "($word1|$word2)" $file > $ofile2

so.. if this is purely some academic homework..
maybe you want to share *your* answer with us?

Sven

----------------
$ cat $ofile1
Prioritätsordnung
Prioritätsrecht
Prioritätssteuerung
Prismen
Pritschenwagen
Privatadresse

$ cat $ofile2
Karlspeises
Karlsruhe
Karlsruher
Karneval
Karnevale
Karnevalen
Karnevals
Karnevalsmaske
Karnickel
Karnickeln
--
Prioritätsordnung
Prioritätsrecht
Prioritätssteuerung
Prismen
Pritschenwagen
Privatadresse