Problem with deleting the first n characters on a line

[CyCliC]

I have a file containing this:

gpu/aps/mcp/mc_server/index.html
gpu/aps/mcp/mca_rtsp/index.html
gpu/oms/subagent/testagent/index.html
sys/netbsd/src/sys/netstreams/index.html

and I use this to delete gpu on the first line:
($str contains "gpu")

 open(FOUND, "+<found");
 $first_line = <FOUND>;
    $tmp = length $str;
    $tmp = $tmp +1;
    $linetmp = substr($first_line, $tmp);
    print FOUND $linetmp;

but I get this as a result:

gpu/aps/mcp/mc_server/index.html
gpu/aps/mcp/mca_rtsp/index.html
gpu/oms/subagent/testagent/index.html
sys/netbsd/src/sys/netstreams/index.html
gpu/aps/mcp/mc_server/index.html <-- where did this line come from
aps/mcp/mc_server/index.html

please help!

[nysus]

I would use a substitution operator to get rid of the leading "gpu" like so:

$first_line =~ s/^gpu//;
print FOUND $first_line;

See perlop for more about substitution and perlre for more on regular expressions. Also, it doesn't look like you are using strict. Strict forces you to used good coding practices and helps you catch a lot of coding errors. Always, always put "use strict" shortly after the shebang line.

Update: If you just want to get rid of the first 3 arbitrary characters, use s/^.{3}//;

[Cubes]

You're not resetting your file position before writing back into FOUND (and even if you did, I don't think this code would do what you expect because the new string you write out is shorter than the original).

The quick 'n' easy way: open up a second file for output. Loop through the entire input file, changing lines as necessary before printing them to the output file. At the end, remove the original and rename your output file to the original's filename.

[CyCliC]

thanks for your help

the "use strict" tip was very needed
but I still have the same problem, the extra line in the found file