R help - Feb 2012 - optparse::parse_args, using equals sign or not

Hi

We've found that when using parse_args(..., positional_arguments=FALSE), 
it is permissible to invoke our script with either "--myfoo=bar" or 
"--myfoo bar"; that is, whether or not the equals sign is present
makes
no difference, and in fact both usage forms are demonstrated in the 
optparse vignette.

However, we've found that when using parse_args(..., 
positional_arguments=TRUE), it is no longer possible to use the equals 
sign; that is, "--myfoo=bar" will show up as a positional argument,
with
"myfoo" set to its default value instead.

Wondering if this is really the intended behavior? Some of our scripts 
use positional arguments and some do not, and we'd like to arrive at a 
standard usage across all scripts.

Many thanks,
Ben

> Hi
>
> We've found that when using parse_args(...,
positional_arguments=FALSE), it
> is permissible to invoke our script with either "--myfoo=bar" or
"--myfoo
> bar"; that is, whether or not the equals sign is present makes no
> difference, and in fact both usage forms are demonstrated in the optparse
> vignette.
>
> However, we've found that when using parse_args(...,
> positional_arguments=TRUE), it is no longer possible to use the equals
sign;
> that is, "--myfoo=bar" will show up as a positional argument,
with "myfoo"
> set to its default value instead.
>
> Wondering if this is really the intended behavior? Some of our scripts use
> positional arguments and some do not, and we'd like to arrive at a
standard
> usage across all scripts.
>
> Many thanks,
> Ben
This bug when using long flag arguments with an equals sign when
positional_arguments=TRUE was fixed in the version of optparse just
uploaded to CRAN (version 0.9.5).  Windows and MAC binaries are being
compiled (and are also available on R-Forge).

Trevor