Hello Martin, Hello list,

now there is the need to implement it ;)

however, i get lots of devices ttyS[2-31] with the following

DEVNAME /dev/ttyS10
DEVPATH /devices/platform/serial8250/tty/ttyS10
ID_MM_CANDIDATE 1
MAJOR 4
MINOR 74
SUBSYSTEM tty
USEC_INITIALIZED 91478

which are actually not real devices (running on ubuntu 13.10) - i assume they get added
automatically ...

and two devices which are actually existing on my hardware

DEVNAME /dev/ttyS1
DEVPATH /devices/pnp0/00:08/tty/ttyS1
ID_MM_CANDIDATE 1
MAJOR 4
MINOR 65
SUBSYSTEM tty
USEC_INITIALIZED 82505

DEVNAME /dev/ttyS0
DEVPATH /devices/pnp0/00:09/tty/ttyS0
ID_MM_CANDIDATE 1
MAJOR 4
MINOR 64
SUBSYSTEM tty
USEC_INITIALIZED 5299

so the question arises how to filter out the non-existing ones ?

is it safe to ignore all those "DEVPATH /devices/platform/serial8250/tty*" ?

thanks for any hints,
marcus.

On 2013-09-08 11:20, Martin Pitt wrote:
Hello Marcus,

Marcus Priesch [2013-07-22 14:39 +0200]:
i have a software written in python that needs to know the list of
available serial devices on the system, including on-board serial
devices and USB serial adapters.

in the past i used hal for this and got nice results like this (lshal output):

You get pretty much the same information with libudev these days,
which is the recommended interface not just for enumerating the
current hardware, but also to get notified about any hardware changes.
For C programs you'd just use libudev (or libgudev for GObject-y
stuff), for Python you have the choice between pyudev
(http://pyudev.readthedocs.org/en/latest/) or using GUdev through
introspection (http://www.freedesktop.org/software/systemd/gudev/),
like this:

from gi.repository import GUdev
client = GUdev.Client.new(subsystems=['tty'])
devs = client.query_by_subsystem('tty')

Then you can iterate over devs and call stuff like .get_device_file()
or .get_property('ID_SERIAL') on those.

Martin

_______________________________________________
devkit-devel mailing list
devkit-devel@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/devkit-devel

Reply via email to