[realized I sent my prior response to Neil alone. Summary was: Thank you!]
I see that I can get the event telling me that something changed. As far as I can tell the event contains no information about *what* changed, it simply alerts that *something* changed. Likewise, the monitor is only for the directory that I point it at, not for the subdirectories. That leaves me two questions: - I could keep a copy of the directory state in (RAM | the DB) and then, when I get the filesystem event, I could walk down through the filesystem and compare against that copy to know what happened. Is there a better way to do this? - Is there a way to monitor a directory and all its subdirs without creating separate event monitors for each one? On Thu, Jan 19, 2017 at 3:58 PM, Neil Van Dyke <[email protected]> wrote: > David Storrs wrote on 01/19/2017 03:08 PM: >> >> of events but I still know nothing about detecting filesystem change >> events. I've looked through PLaneT and found nothing that seems like >> an FS-monitoring package. Can anyone suggest how to do this? > > > Here's a simple example, using `sync`: > > #lang racket > (define my-fs-change-evt (filesystem-change-evt "/home/user")) > (printf "sync...~n") > (sync my-fs-change-evt) > (printf "synced!~n") > > I ran this example on GNU/Linux, and then used the command `touch > /home/user/x`, which caused the `sync` procedure to return. > > See the documentation in the vicinity of that for `sync`, for other other > ways to use synchronizable objects. They're essential for some kinds of I/O > programming, not just GUI. > > Don't be scared away by the events (and ports documentation. I've done a > bunch of low-level programming in C, of I/O and such, and I still find some > of the Racket documentation for related concepts to be intimidating. (Well, > there is one big thing in there that's scared me off the few times I've > wanted to use it, but each time, I was on the clock for a consulting client, > so I did things in the known-quantity way instead. Much of the primitives > are simple, however, and the work is in building correct and efficient > programs around the primitives.) > > BTW, You don't want "PLaneT", which was the old (and wise) package system. > Almost all current development has moved from PLaneT to > "pkgs.racket-lang.org". > -- You received this message because you are subscribed to the Google Groups "Racket Users" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. For more options, visit https://groups.google.com/d/optout.
