Re: [Ranger-users] passing / testing macros to shell

[hut]

On Thu, Jan 21, 2016 at 06:38:03AM +0100, Diederich Hessling wrote:
> Hi there. I couldn't find much about this on the web, so I thought I'd just 
> ask here.
> I am not very familiar with this stuff, so... my question might be more 
> related to bash or python(?) than being ranger-specific.
> My aim is to use ranger to create a list of files and directories for later 
> use. I found that the --choosefiles option works only inside the directory 
> one is currently browsing. As the manual says, the "%c" macro is the only one 
> that transcends the current dir.
> So I browse my file tree and mark files and dirs with the "yank/copy" 
> function.
> I mapped a key to something like 'map td shell echo "%c" >> list_of_stuff' 
> with some sed piping to give me line breaks and it works fine.
> I would like that to work also when nothing is yanked/copied, and the shell 
> expression then should use "%f" instead. 
> I tried to insert a simple shell test in my rc.conf like '[[ -z "%c" ]] &&' 
> but apart from not knowing how to correctly pass the macro it seems that 
> ranger always gives an error if "%c" is empty.
> What would be the best way to decide whether the %c or %f macro will be 
> passed to the shell?

There is the command "save_copy_buffer" which writes the copy buffer to
~/.config/ranger/copy_buffer, maybe that helps you.

To determine if nothing is yanked, you can check the variable
fm.copy_buffer for emptiness.  For example:

  :eval fm.execute_console("echo SOMETHING IS COPIED") if fm.copy_buffer else 
fm.execute_console("echo NOTHING IS COPIED")

you can bind this to a key like any other command with "map X eval ..."

regards, hut