Re: uuencode manpage

On 9/11/21 3:18 AM, Davide Brini wrote:
> I think what the OP meant is that in the text of the man page, "-"
> could be used in place of "/dev/stdout" since it works the same and it's
> shorter.

Turns out, that's true. The name there is for the benefit of uudecode 
and uudecode does this:

>   if (  (strcmp (outname, "/dev/stdout") != 0)
>      && (strcmp (outname, "-") != 0) )
>     {
>       rval = reopen_output (outname, mode);
>       if (rval != UUDECODE_EXIT_SUCCESS)
>         goto fail_return;
>     }
meaning that the strings "/dev/stdout" and "-" are equivalent. I guess 
if you want a file named "-", you'll need to use "./-" :)

Who uses this anymore anyway? I've updated the text and someday when 
there's a new release it will say:

> @file{uuencode} will read @file{in-file} if provided and otherwise
>         read data from standard in and write the encoded form to 
> standard out.
>         The output will begin with a header line for use by 
> @file{uudecode}
>         giving it the resulting suggested file @file{output-name} and 
> access
>         mode.  If the @file{output-name} is specifically either
>         @file{/dev/stdout} or @file{-}, then @file{uudecode} will emit the
>         decoded file to standard out.
I should probably mention that this is an unauthorized extension to 
POSIX since "-" is a valid file name character and the official flavor 
of uudecode must emit the output into a file so named. deviations are 
allowed only if the output name does not come from the official file 
name character set. Honoring the "-" as stdout should probably be done 
IFF POSIXLY_CORRECT is not set. Assuming sharutils aren't de-POSIX-fied ;)