Handling of the "silence" parameter in program_segments

[jonaskeller]

It seems like this line

pdq/pdq/host/protocol.py

Line 437 in 2880b59

silence = data.pop("silence", False)

is looking for the silence parameter one level too high and should actually be called for target_data in the loop a few lines below (if at all). Besides always setting silence to False, it also causes a KeyError exception when silence gets passed to getattr(segment, target)() twice, once by itself and once within **kwargs.

[jonaskeller]

To add more context, the current version works with the waveform defined here:

pdq/pdq/test/test_dac.py

Lines 53 to 94 in 702c069

	_test_program = [
	[
	{
	"trigger": True,
	"duration": 20,
	"channel_data": [
	{"bias": {"amplitude": [0, 0, 2e-3]}},
	{"bias": {"amplitude": [1, 0, -7.5e-3, 7.5e-4]}},
	{"dds": {
	"amplitude": [0, 0, 4e-3, 0],
	"phase": [.25, .025],
	}},
	],
	},
	{
	"duration": 40,
	"channel_data": [
	{"bias": {"amplitude": [.4, .04, -2e-3]}},
	{
	"bias": {"amplitude": [.5]},
	"silence": True,
	},
	{"dds": {
	"amplitude": [.8, .08, -4e-3, 0],
	"phase": [.25, .025, .02/40],
	"clear": True,
	}},
	],
	},
	{
	"duration": 20,
	"channel_data": [
	{"bias": {"amplitude": [.4, -.04, 2e-3]}},
	{"bias": {"amplitude": [.5, 0, -7.5e-3, 7.5e-4]}},
	{"dds": {
	"amplitude": [.8, -.08, 4e-3, 0],
	"phase": [-.25],
	}},
	],
	},
	]
	]

However, this definition contradicts the description and example at http://pdq.readthedocs.io/en/v3.0-rc2/manual.html#wavesynth-format , where silence is an entry of the spline_data dictionary rather than the spline dictionary.