I tackled problem 4 here:

https://chromium-review.googlesource.com/583781

=== Problem 5 ===

configure-service and set-service-property can handle string properties, but they do not handle the full range of data types.  In particular, the new StaticIPConfig dictionary property cannot be set (I wasn't even able to make this work by manually hacking the Python code).

Problem 1: https://chromium-review.googlesource.com/c/583891/

On CL:583781, Ben said: "As we can obtain the DBus signature of a property, we could probably approximate that by taking a JSON-string from command line and convert that into an equivalent dbus type."

I have not had any luck setting StaticIPConfig as a dict from Python.

I tried this first:

    foo = dbus.Dictionary({'Address': dbus.String("1.2.3.4")})
    service.SetProperty("StaticIPConfig", foo)

This resulted in: "dbus.exceptions.DBusException: org.chromium.flimflam.Error.InvalidArguments: Property StaticIPConfig is not a string map."  I think this is because PropertyStore::SetAnyProperty recognizes that the argument can be interpreted as a Stringmap, so it doesn't even try to interpret it as a VariantDictionary.

Then I tried:

    foo = {'Address': "1.2.3.4", 'Prefixlen': 32}
    service.SetProperty("StaticIPConfig", foo)

but this failed with: "ERROR:dbus.connection:Unable to set arguments ('StaticIPConfig', {'Prefixlen': 32, 'Address': '1.2.3.4'}) according to signature None: <type 'exceptions.TypeError'>: an integer is required"

Same with this:

    foo = dbus.Dictionary({'Address': dbus.String("1.2.3.4"),
                           'Prefixlen': dbus.Int32(32)})

I think how shill_proxy.py:service_properties_to_dbus_types uses variant_level=1 may do the trick.

The following revision refers to this bug:
  https://chromium.googlesource.com/aosp/platform/system/connectivity/shill/+/69d2de6afc47c76b8cba7d609512d1136829ac51

commit 69d2de6afc47c76b8cba7d609512d1136829ac51
Author: Brian Norris <briannorris@chromium.org>
Date: Tue Jul 25 21:04:06 2017

shill: test-scripts: don't lookup Service names by substring

It's an extremely common pattern to have Wifi SSIDs in a given area
where one network is named $foo and another is named $foo-$bar (for some
values of $foo and $bar; e.g., bar == 5G). The test scripts currently
can't tell the difference between those when you try, e.g.,
"connect-service $foo". It might choose $foo or $foo-bar depending on
Python implementation details.

Let's disambiguate this for scripts that accept a Service arg by
stopping to use substrings.

BUG=chromium:735524
TEST=connect-service, disconnect-service, set-service-property for
     SSID $foo when $foo-bar exists in the same area

Change-Id: Ic424d43ea7e226536480194d981ad79599281b53
Signed-off-by: Brian Norris <briannorris@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/583781
Reviewed-by: Ben Chan <benchan@chromium.org>

[modify] https://crrev.com/69d2de6afc47c76b8cba7d609512d1136829ac51/test-scripts/disconnect-service
[modify] https://crrev.com/69d2de6afc47c76b8cba7d609512d1136829ac51/test-scripts/set-service-property
[modify] https://crrev.com/69d2de6afc47c76b8cba7d609512d1136829ac51/test-scripts/connect-service
[modify] https://crrev.com/69d2de6afc47c76b8cba7d609512d1136829ac51/test-scripts/flimflam.py

The following revision refers to this bug:
  https://chromium.googlesource.com/aosp/platform/system/connectivity/shill/+/71821bda595d5a106b5e0b59a9924b3ad8288ca7

commit 71821bda595d5a106b5e0b59a9924b3ad8288ca7
Author: Brian Norris <briannorris@chromium.org>
Date: Thu Aug 10 00:19:42 2017

shill: test-scripts: stop looking for 'Favorites'

These died with flimflam a long time ago, in a galaxy far away.

For printing services, it's probably useful to use the asterisk for
notating IsActive. Do that.

For removing services, I don't know what we want to do. But we surely
don't want the "Favorite" property, so kill it. (Note that "remove" is
still broken.)

BUG=chromium:735524
TEST=`./test-flimflam services` gives a reasonable dump

Change-Id: Ieb82028635d82572364dddeb4c56d61823c10cdc
Signed-off-by: Brian Norris <briannorris@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/583891
Reviewed-by: Ben Chan <benchan@chromium.org>

[modify] https://crrev.com/71821bda595d5a106b5e0b59a9924b3ad8288ca7/test-scripts/test-flimflam

drive-by:
I see changelists on this bug.  What's the reason it's left open and untriaged?

Several of the original problems reported are not fixed, and no one has signed up to try to fix them. These are also only debug scripts, so their importance is basically determined only by whether the people who might use them care to fix them up.

I guess "Available" is accurate.

This issue has been Available for over a year. If it's no longer important or seems unlikely to be fixed, please consider closing it out. If it is important, please re-triage the issue.

Sorry for the inconvenience if the bug really should have been left as Available.

For more details visit https://www.chromium.org/issue-tracking/autotriage - Your friendly Sheriffbot