DFReader.py: add bitmask data to verbose print output

[peterbarker]

.. taking the conservative approach to this new feature failing....

@shancock884

[peterbarker]

2024-03-28 15:13:19.699: POWR
    TimeUS: 211395439 µs
    Vcc: qnan V
    VServo: qnan V
    Flags: 0
        !BRICK_VALID (main brick power supply valid)
        !SERVO_VALID (main servo power supply valid for FMU)
        !USB_CONNECTED (USB power is connected)
        !PERIPH_OVERCURRENT (peripheral supply is in over-current state)
        !PERIPH_HIPOWER_OVERCURRENT (hi-power peripheral supply is in over-current state)
    AccFlags: 0
        !BRICK_VALID (main brick power supply valid)
        !SERVO_VALID (main servo power supply valid for FMU)
        !USB_CONNECTED (USB power is connected)
        !PERIPH_OVERCURRENT (peripheral supply is in over-current state)
        !PERIPH_HIPOWER_OVERCURRENT (hi-power peripheral supply is in over-current state)
    Safety: 6

2024-03-28 15:13:19.700: RCI2
    TimeUS: 211396039 µs
    C15: 1501 us
    C16: 1501 us
    OMask: 0
    Flags: 1
        HAS_VALID_INPUT (true if the system is receiving good RC values)

[shancock884]

Looks nice! A couple of thoughts you are free to consider or ignore :-) ....

I know it is the same as dump_message_verbose in mavutil, but I don't really like the use of "!" for False values.... when I first came across it my eyes read it as a check-mark meaning true (i.e. opposite of what it means)! But maybe its best to remain consistent until someone has a strong enough argument/opinion to change both.

I was thinking it could be good to split the dump_verbose_bitmask function into two: get_field_metadata and dump_verbose_bitmask, as it should make it easier to implement showing enums too in a later PR without looking up metadata twice (and probably needed to do following suggestion...).

It would be nice to show the hex value for bitmasks, alongside the decimal (as dump_message_verbose does). If you do create get_field_metadata, then you can get the field metadata first and do something like if hasattr(field_metadata, 'bitmask'): f.write(" 0x%X (%d)\n" % (val, val)).

[peterbarker]

I know it is the same as dump_message_verbose in mavutil, but I don't really like the use of "!" for False values.... when I first came across it my eyes read it as a check-mark meaning true (i.e. opposite of what it means)! But maybe its best to remain consistent until someone has a strong enough argument/opinion to change both.

You wouldn't be surprised to hear I find the ! thing readable, but definitely open to alternatives.

I was thinking it could be good to split the dump_verbose_bitmask function into two: get_field_metadata and dump_verbose_bitmask, as it should make it easier to implement showing enums too in a later PR without looking up metadata twice (and probably needed to do following suggestion...).

It would be nice to show the hex value for bitmasks, alongside the decimal (as dump_message_verbose does). If you do create get_field_metadata, then you can get the field metadata first and do something like if hasattr(field_metadata, 'bitmask'): f.write(" 0x%X (%d)\n" % (val, val)).

Seems reasonable. If you can put those suggestions in patch form I can apply them... otherwise I'll come back to this in a while.

Or I could merge and accept patches later?

[peterbarker]

@shancock884 I had another look at your suggestions. I've re-arranged things a little to be somewhat like what you're suggesting, but haven't modified the printing of the value. Too many cases to test for this PR ;-)

Further patches would probably be something like field_metadata = field_metadata_by_name.get(c, None), test for None instead of presence-in-field-metadata-by-name in the call to dump_verbose_bitmask, and then some weird stuff when printing the values. It's complicated by the py2 compat code.