[Zel2008]

Hi everybody,
I'm trying to split a file based on context of certain lines, to transform it from this:

<random stuff>KEYWORD<random stuff>
more lines section 1
<random stuff>KEYWORD<random stuff>
more lines section 2
<random stuff>KEYWORD<random stuff>
more lines section 3

to this:

FILE 1
<random stuff>KEYWORD<random stuff>
more lines section 1

FILE 2
<random stuff>KEYWORD<random stuff>
more lines section 2

FILE 3
<random stuff>KEYWORD<random stuff>
more lines section 3

Basically, split a big file so any line that has the KEYWORD in it (anywhere in the line) is the first line of a new file. I want to number those files 1.txt, 2.txt, 3.txt, etc.

Since I'm going to be working with potentially 500+ files after the split, I can't use csplit, and that's the only unix utility I know that can do anything like this. Does anyone know of a possible 1 or 2 liner (or failing that, a short script) that can also do this? I'd appreciate any help.

Thanks,
Zel2008

[mufasa]

hopefully not grave digging, but have you looked at using a small bash script and just checking a line for your keyword/phrase? it may not be super efficient but it would work. you could check if the keyword existed by doing something like:

if [[ $line == *KEYWORD* ]]; then
     # do some fun stuff
fi