On Mon, 2013-11-25 at 18:38 +0100, Jan-Philip Gehrcke wrote:
> PS . Jason:
> "parmed.py ../molecule.prmtop 'printDetails .7'" does not, which I did
> not expect. In the help message the last argument is described as
> "[<script>]". I read from this that <script> should be string containing
> parmed commands. However, the latter is interpreted as path to a file.
> But there is an additional option (-i FILE) for specifying a path to a
> script file. So, are there currently two ways to specify a path to a
> script file and no way to directly provide a command string?
This might make more sense in the context of ParmEd development. I took
most of the design principles from cpptraj and ptraj (in fact, I think
only readline support showed up in ParmEd first). Like ptraj/cpptraj,
there was only one way to provide input in the beginning:
parmed.py <prmtop> <script>
and it behaved like ptraj. Then like cpptraj, it started to support
multiple topology files, commands that did not require a topology file
at all, and an option to source command files and read multiple scripts.
The only way to handle this generally was to provide a flag for prmtop
and script files that could be used multiple times, so the usage became
parmed.py -p <parm1> -p <parm2> ... -i <script1> -i <script2> ...
To avoid breaking backwards compatibility (akin to what cpptraj does), I
support both options. I never saw the need to add support for
command-line flags to give a single ParmEd command since parmed reads
stdin. I always do something like:
echo "command" | parmed.py [options]
or
parmed.py [options] << EOF
command1
command2
EOF
I will clarify the help statement to indicate that the scripts are all
files. (The -e option is improperly documented as well)
Thanks,
Jason
--
Jason M. Swails
BioMaPS,
Rutgers University
Postdoctoral Researcher
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Mon Nov 25 2013 - 11:00:02 PST
