AP_Scripting: add bindings for servo telemetry

[IamPete1]

Adds bindings for servo telem, this allows servo checks to be done in lua.

[IamPete1]

For refence attempt 1 cost 1500 bytes or so.

[IamPete1]

This now gets the structure and adds auto generation of a bitmask check and set. This means the script writer does not need to know there is a validity bitmask in the structure. This is currently read only, but it also adds auto setting of the mask when a value is set. The only slight problem is that once set there is no way to un-set the mask from lua.

One disadvantage to this method is that were stuck with the types in the structure, previously I had returned the temp in degrees by applying the scale in the getter, now were stuck with centi degrees.

[IamPete1]

Example of the new docs markup:

userdata AP_Servo_Telem::TelemetryData field command_position float'skip_check read valid_mask valid_types AP_Servo_Telem::TelemetryData::Types::COMMANDED_POSITION

First bit is the same, the stuff added is:

valid_mask valid_types AP_Servo_Telem::TelemetryData::Types::COMMANDED_POSITION

"valid_mask" is the keyword to enable the bitmask validity check.
"valid_types" is the variable name on the "AP_Servo_Telem::TelemetryData" object.
"AP_Servo_Telem::TelemetryData::Types::COMMANDED_POSITION" is the bitmask value that the mask should be checked against.

[IamPete1]

Now only costs about 1K, so saves 400 bytes over the individual getters.

[tpwrules]

I haven't looked in detail, but it seems this can't absolutely fix the race condition because there isn't a lock protecting the telemetry data. The race window is much shorter now however, so maybe it's fixed enough? It was said in the call that there's servos which output different data at different rates anyway, so maybe comparing commanded vs. achieved just isn't possible in general.

Also still curious what the size impact would be if you had one mega function which returned eleven values.

[rmackay9]

Looks OK to me but @tpwrules will review more thoroughly

[tpwrules]

A couple little things in the generator but should be ready to go after!

I tested in SITL using quadplane-can and the position from four servos comes through but it's the same for all four of them. I checked that the data I could get from Lua matched the data logged in CSRV, and it all did so this PR is good, but there's still probably something funky here.

fixed

race condition (mostly) fixed

[tpwrules]

Tested and still works, thanks for the fixes.

I cleaned up the history slightly and folded in another really tiny fix to the generator. Figured it was easier to do it myself then to go through review again.

[tpwrules]

Concerned whether servo telemetry needs a fix similar to #28999 but that doesn't impact this PR at all.