edit a file after a newline

[zeros]

Hi.
I've got a file I'm trying to edit. The format is:

...
SOME UNIQUE EXPRESSION
ANOTHER EXPRESSION
...

I'm trying to change "ANOTHER EXPRESSION" to be that of the filename minus the suffix

Example:

bash-prompt$ ls
foo.suffix

bash-prompt$ ./some_magic_script foo.suffix > foo.suffix.changed
bash-prompt$ cat foo.suffix.changed
...
SOME UNIQUE EXPRESSION
foo
...

I know that this can be done in perl, sed, awk but am not sure how: I don't know how to use any of them (yet. I'm working on it).

I've sort of fiddled with sed so far, but haven't been able to get the N command to work so that it doesn't chop off the newline characters.

Thanks in advance.

[ghostdog74]

this assumes you have suffix of .xxx

awk '{      
      gsub(/\....$/,"",FILENAME)
      gsub("ANOTHER",FILENAME)
}1 ' file.txt

[zeros]

this assumes you have suffix of .xxx

awk '{      
      gsub(/\....$/,"",FILENAME)
      gsub("ANOTHER",FILENAME)
}1 ' file.txt

Thanks ghostdog. I was able to modify your script to fit my needs:

bash-prompt$ ls
file.suffix

bash-prompt$ awk '{
sub(/\*$/,"",file)
sub("0.1","file")
}1 ' file.suffix

0.1 was the string that I need to replace in the experimental file. Is there some way to not have this as a magic number? I don't know exactly what its going to be for some other files. I think I may not have been clear in the OP (my bad):

...
SOME_UNIQUE_STRING
SOME_UNKNOWN_STRING
...

I've actually got close to 100 files that I need to edit in the same way, and in each of them the string that I need to replace is different. It will always follow the unique string in all the files however, so I was hoping to locate the unique string and the modify the line immediatly afterwards.
I tried using some globbing characters to overwrite the line, but was either getting some missing operand failures of a replacement of more than just the desired text.

[zeros]

I decided to use perl to do this. In case anyone else runs into a similar problem, here's how I solved it:

#!/usr/bin/perl
$file=$ARGV[0];
$name="$file\n";
$name =~ s/\.[^.]*$//;
open(FILE, $file) || die("Could not open file!");
while ($line = <FILE>) {
	chomp($line);
	if ($line =~ /MAGIC_STRING/){
		print "$line\n";
		$line = <FILE>;
		chomp($line);
		$line =~ s/.*/$name/;
	}
	print "$line\n";
}
close(FILE);

It was a good opportunity to pick up some perl. Soon I'll be golfing the same

[ghostdog74]

well, I just understood your requirement after running your perl script.
Here's the Perl-less equivalent

awk '/MAGIC_STRING/{ 
    print;
    getline;
    sub(/\....$/,"",FILENAME)
    print FILENAME;
    next}1 ' file