CODEWITHSUNDEEP
rubylinux
Charles Saag
Charles Saag

Reputation: 631

How to set a program's working directory?

I have a program that exists in the /usr/local/flower directory - flower.rb. It requires loading libraries that also exist in the same directory. I've included this flower directory in my OS environment's path. And my choice of OS is Ubuntu.

When I execute flower.rb from any other directory other than /usr/local/flower, I get error messages indicating the program can't load the libraries that also exist in the /usr/local/flower directory b/c they are being loaded as ./[library] from source code.

I realize I could change the Ruby program to hard code the /usr/local/flower/[libraries], but I'm curious if there's a way to execute this program from my home directory, e.g. /home/seattle, w/o doing this.

Also, when the program executes, it creates output via the -o switch.

My solution was to create a bash script that changed to the /usr/local/flower directory, executed the program, and then return to the PWD directory from where the flower.rb program was called. The problem is the -o switch. If I do a -o [file] the [file] gets written to the /usr/local/flower directory as opposed to where I am when the flower.rb program is run.

What is a good solution for this problem?

Upvotes: 2

Views: 370

Answers (1)

Jörg W Mittag
Jörg W Mittag

Reputation: 369438

It's not quite clear from your question, but it appears that you are using require wrongly. require is for loading scripts from the $LOAD_PATH. If you want to load a script relative to the directory of the current script, use require_relative.

You should never load scripts relative to the current working directory. The current working directory is under the control of the user, not your script. You have no idea what it is gonna be. There is a reason why the current working directory was removed from the $LOAD_PATH in 2008: because it's broken.

Upvotes: 3

Related Questions