bad interpreter: No such file or directory

[IrateRover]

Hi, I am a linux newbee, using fedora core 5.

I get the following message when I try to run a simple test script. The same script runs on another machine very similar to my own.

./mydate: /bin/bash^M: bad interpreter: No such file or directory

The file does has execute rights.

[Cabhan]

I take it that you wrote the script on a Windows machine? And ran it under Cygwin, I suppose?

In Windows, the end of a line is CRLF (Carriage-Return, Line-Feed). In Unix-like systems, it is just LF. So when you try to run that script, every line has an extra CR on the end.

The way to fix this is to use some sort of utility to fix that. You might use the utility dos2unix if you have it installed, or just go download it.

Alternatively, the following script should work:

#!/usr/bin/perl

die "Usage: $0 < files >\n" unless @ARGV;

for $file (@ARGV)
{
    open IN, $file or die "$0: Cannot open $file for input!\n";

    my @lines = <IN>;

    close IN;
    open OUT, "> $file" or die "$0: Cannot open $file for output!\n";

    s/\r$// for @lines;
    print OUT for @lines;
}

[IrateRover]

Hi Cabhan,

Thanks for replying but I am running pure Fedora core 5, the script was written in Fedora so there is no CRLF incompatability here.

[drl]

Hi, IrateRover.

My first thought was the same as Cabhan's, so I tried adding a ^M after the bash in the shebang (first line: #!/bin/bash^M ) and did get an error message, but it was Command not found.

Googling for "bad interpreter linux" produced a number of hits and our friends in NZ provided a number of possible causes on their wiki. Best wishes ... cheers, drl

http://www.wlug.org.nz/CommonErrors

( edit 1, 2: typos )

[IrateRover]

Hi Cabhan, I tried out your script against my test script and now it works.

many thanks

IrateRover

[saxykat]

worked for me too! thanks!