How can I make it open the patch in the already running pd instance?

I use terminal a lot in my workflow, and usually launch things from terminal (I use 'lf' which is a terminal file explorer). Is there any trick to make xdg-open acknowledge an already running instance?

heres the pure data desktop file da xdg-open uses to handle .pd files:

[Desktop Entry]
Name=Pure Data (Pd)
Exec=pd-gui %F
Categories=Audio;AudioVideo;Development;Music;Graphics
Keywords=Audio;Video;Programming;Synthesis;Analysis
Comment=Visual dataflow programming platform for multimedia
Comment[ca]=Plataforma de programació visual per aplicacions multimèdia
Comment[de]=Grafische Datenflussprogrammierung für Multimedia
Comment[es]=Plataforma de programación visual para aplicaciones multimedia
Comment[fr]=Plateforme de programmation visuelle pour applications multimédia
Comment[it]=Piattaforma di programmazione visuale per applicazioni multimedia
Comment[pt]=Plataforma de programação visuais para multimedia
MimeType=text/x-puredata;
GenericName=Dataflow IDE
Type=Application
Terminal=false
Icon=puredata

I know this is unix related, I even tried asking in unix stack exchange but pure data seems to be alien to them:

https://unix.stackexchange.com/questions/752495/open-file-from-terminal-in-already-running-pure-data-instance?noredirect=1#comment1431905_752495

@zerino I have fought with this more than once without success and every *nix I have used has opened pd patches in a new instance of pd. I suspect pd does not speak with xdg and this would require a feature request on github but I may just be failing at using xdg. One of those things, I have gotten so used to always opening patches from pd itself that it is rarely an issue.

@oid said:

ve gotten so used to always opening patches from pd itself that it is rarely an issue.

I feel you my fellow. Thing is, pd open file dialog (I think its tk right?) is a pain to use. I wish i didn't have to.

@zerino As someone who has used linux since the days when TK was big and Gnome/KDE were not much better I am fairly comfortable with pd and used to its quirks. TK has some big advantages, it is very simple for the user to make changes to the UI through scripts and those changes carry over through updates without worry about things breaking. Maintaining the flexibility of TCL/TK would probably require a fairly large increase in work on the devs end. If you don't get a solution in a day or two make a feature request on the pd github.

@zerino Try using pdsend (a binary that is in your Pd distribution) from your script to send a message to Pd...
pd open patch.pd /the/path/to/patch/from/root
David.

Also you could try to
use the terminal to browse,
ctrl+shift+c copy > alt+tab switch over to Pd task > ctrl+o open > ctrl+v paste > enter.

At least this works for me with the Windows file-browser and portable Pd.

@lacuna I think the problem belongs only to Linux.... it is easy on OSX and windows.
I am sure that pdsend will work by sending that message to a [netreceive] object in a currently open patch in the currently running instance of Pd. Not sure that Pd itself will receive it though.
Still searching for an answer to that.... sending it to the Pd executable receive port maybe...... but it is probably locked.
David.

@whale-av yes, copying a selceted file in win-file-browser does full path in clipboard...

I think sending via pdsend is slick. And I believe and hope and am interested in if Pd receives it without [netreceive].

Otherwise, opening a [netreceive] patch with a startup-flag (in terminal or preferences or in a script) should do the trick:
https://forum.pdpatchrepo.info/topic/13895/opening-a-patch-inside-another-patch-by-filename/2

@lacuna So pdsend and pdreceive are simple binaries that set up network sockets.
pdreceive could maybe be bound to Pd in a plugin, but a simple patch.....
[netreceive xxxx]
|
[s pd]
would be easier...
David.

@whale-av said:

@zerino Try using pdsend (a binary that is in your Pd distribution) from your script to send a message to Pd...
pd open patch.pd /the/path/to/patch/from/root
David.

$ pd open path just opens another instance.

trying with pdsend results in:

$ pdsend open /home/zerinol/fixo/pd/patches/audio_loader.pd
usage: pdsend <portnumber> [host] [udp|tcp] [timeout(s)]
(default is localhost and tcp with 10s timeout)

@zerino You need a patch open with a [netreceive] object in it and connect the left outlet to a [s pd] like,

[netreceive 3000]
|
[s pd]

Then from the command line you send a command like echo open filename.pd $(pwd)\; | pdsend 3000 and it will open filename,pd, this assumes you are in the directory which the file is in, the $(pwd) command can be replaced by a path. You should be able to setup your file manager to do all of this automatically but I am not familiar with if, or an alias in your .bashrc or the rc of whatever shell you use, for bash alias pdopen="echo open $1 $(pwd)\; | pdsend 3000"

Edit, we can not actually use an alias like that in bash, but we can make that a function, just put this in your .bashrc instead of the alias. Need to run source ~/.bashrc in any open terminals or logout/log back in for it to be usable. Then just run pdopen <filename> and it should work.

function pdopen () {
   echo open $1 $(pwd)\; | pdsend 3000
}

Everything should work now.

Some very naive tcl coming through, but it works and hopefully someone who is not ignorant of tcl will make it sane. Seems like there should be a better way to do this than constantly checking for a file, someway to send a message but I am a complete hack when it comes to tcl and I do not really get along with it. Suppose this should be done in its own namespace as well. But it works.

Create a file called pdopen-plugin.tcl and put this in it, save it somewhere in your pd path.

file delete /tmp/pdopenfile

proc pdopen {} {
   if { [file exists /tmp/pdopenfile] == 1} { 
        set f [open /tmp/pdopenfile]
        set fp [read $f]
        file delete /tmp/pdopenfile 
        close $f
       pdsend "pd open $fp"
    }
    after 100 pdopen    
}

pdopen

And a bash script, save this as pdopen, run chmod +x pdopen and then copy this to your users path or somewhere like /usr/bin or /local/bin or where ever user installed programs go on your distro if you want it global for all users.

#!/bin/bash

fp=$(readlink -f $1)

if [[ $(pidof pd) ]]; 
    then
        f=$(basename $fp);
        p=$(dirname $fp);
        echo "$f" "$p" >> /tmp/pdopenfile
    else
        pd "$fp";
fi

Now you do not need a patch open with a [netreceive], just make pdopen the default application for .pd files and if there is a running instance of pd it will open them in that instance, if not it will start up pd.

@oid said:

echo open filename.pd $(pwd); | pdsend 3000

Yep, it really works. Thank you @oid

@oid wow, this looks solid

@zerino It is reasonably firm at best, it will fail if you try and open more than one file at a time, mostly because TCL has a tendency to leave me demoralized so I like to quit when ahead. But I did clean out my patch folder in record time last night, really made things easier.