The testserver is single threaded (assuming you're using django-
admin.py runserver), so it would appear to hang until the os.popen
call completes. If it's really crashing, then something horrible must
be going wrong with the child process.
In a production setting, where you're using a multiprocess or
multithreaded server like lighttpd+fastcgi or apache+mod_python/
mod_wsgi you'd almost get this code to work as is, but it's not really
a best practice. Chances are the server would decide your script is
taking too long and give up, presenting the user an error page after a
long wait.
The simplest option is to save the unprocessed file to disk, then have
a background job process the file and stick it in the database.
Basically just don't do everything in the web process.
You could use cron to schedule the background process or you could get
fancy and implement a daemon that listens to work requests via a queue
(perhaps using the multiprocessing module)
There might be a way to use filesystem notification events to autorun
a script if a file is added to a directory (MacOS definitely has
something like this via launchd)
The catch with all this multiprocess jazz is that you might still need
a way to tell your user what you've done. Since HTTP works on a pull
models, so you can't exactly send them a message once processing is
done (could use email, tweets, rss, or tricky comet stuff). A simple
workaround is to either metarefresh the page or use ajax to poll for
processing completion. You might also just neglect to tell the user
about the encoding step and only let them know when you've finished
uploading the file.
The easiest way to communicate between the background process and the
web process is to use a plain old file in a known location, or to use
the database (best option IMHO) or a shared queue (why bother)
-Jacob
On Feb 13, 7:55 am, "Amirouche B." <aneg...@gmail.com> wrote:
> Hello,
>
> I 'd like to process some file before saving it, let's say it's an
> audio file that I want to be converted to ogg before it lands in the
> database forever.
>
> I tried to run ffmpeg2theora in the save method of the models but this
> doesn't work verywell, the testserver hangup after the processing is
> done.
>
> ## FILE : models.py, ##
>
> class Song:
> [...]
>
> def save(self):
> super(Song, self).save()
> path = self.file.path.split("/")
> path, file = path[1:-1], path[-1]
> joined_path = "/"
> for e in path:
> joined_path += e + "/"
>
> os.chdir(joined_path)
>
> res = os.popen("/usr/bin/ffmpeg2theora " + path)
>
> any help ?
--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups
"Django users" group.
To post to this group, send email to django-users@googlegroups.com
To unsubscribe from this group, send email to
django-users+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/django-users?hl=en
-~----------~----~----~----~------~----~------~--~---
