Ok so I have a few ideas for checking whether or not an instance of a process is already running but I wanna find out what other people already do/use first. I’m looking to do something like firefox does sometimes where it says firefox is already running blah blah blah, it only checks to see if there is another copy of itself running, I’m not really looking to check to see if there is just an arbitrary named process running just if there is another copy of itself.
I don’t know whether or not it would be easier to just set up a system to look for an arbitrary process and just look for itself or if it would be better to implement a system for looking for just that process.
I am trying to eventually to lead into being able to communicate with another process so that I could send messages to it.
So say to do the following just as an example: suppose you have firefox already running then you do a command of firefox URLHERE it opens up the new url in the original window that was opened.
I am also trying to figure out how to implement this so if you have any ideas on the best ways in which to do this then by all means please do let me know as well.
Thanks
Assuming Windows – other operating systems may provide similar constructs but implementation details will vary.
Look into named pipes or named mutexes.
A named mutex solution will be easier to code but it will not give you inter-process communication. The theory goes like this: your process attempts to create a named mutex. If it fails it means that another copy of the process is already running. This is guaranteed by the OS – only one named mutex can be created with a specific name. The trick is then in choosing an appropriate name for your mutex so you don’t run the risk of accidental conflict with another program wanting to create/use the same named mutex. For this you could use a GUID. Note however, that a nefarious application could create the named mutex that you’re looking for and prevent your application from ever running
The second option is to use a named pipe (same story regarding choosing the name). Your process will attempt to create a pipe with a certain name on startup. If creating the pipe fails because the pipe already exists it will go ahead and attempt to connect to the pipe and then you can have the second process exchange information with the first process (i.e. pass on its arguments so that the original process can perform an action)