gattc: use GetUUID() to allow for bare metal use of short UUID.

[deadprogram]

This PR adds a GetUUID() accessor to both DeviceService and DeviceCharacteristic to simplify use on bare metal with short UUIDs.

[deadprogram]

This PR is intended to address the comments from #11

[aykevl]

GetUUID works, but Effective Go suggests leaving out the Get prefix: https://golang.org/doc/effective_go.html#Getters

To make that possible, perhaps you can unexport the UUID field (renaming it to uuid) or rename it to something else such as UnderlyingUUID?

[deadprogram]

That was what I was trying originally but I did not immediately figure out how to handle the naming conflict you mentioned, which is why I took an "easier" way out.

I realized I can do what I wanted using a type alias, which I have just now changed and pushed to this branch.

[aykevl]

This is the function to convert from C.ble_uuid_t to the UUID type: sd_ble_uuid_encode

[deadprogram]

I think I have made all of the changes requested but am unable to test them due to #15

[aykevl]

You could use any development from Nordic to test these changes. But I will also try to test it.

[aykevl]

Unfortunately this doesn't work correctly on a SoftDevice:

$ tinygo gdb -target=pca10056-s140v7 ./examples/scanner/
[...]
(gdb) bt
#0  runtime.nilPanic () at /home/ayke/src/github.com/tinygo-org/tinygo/src/runtime/panic.go:32
#1  0x0002713e in (tinygo.org/x/bluetooth.Addresser).Set ()

It appears that the MAC address is being set on a nil Address.

[deadprogram]

Unfortunately this doesn't work correctly on a SoftDevice:

$ tinygo gdb -target=pca10056-s140v7 ./examples/scanner/
[...]
(gdb) bt
#0  runtime.nilPanic () at /home/ayke/src/github.com/tinygo-org/tinygo/src/runtime/panic.go:32
#1  0x0002713e in (tinygo.org/x/bluetooth.Addresser).Set ()

It appears that the MAC address is being set on a nil Address.

Seems like this problem exists on dev branch even without this PR.

[deadprogram]

Made a great deal of progress here. Now I am testing and discovering some differences between the results being returned by Linux vs. the SoftDevice.

$ go run ./examples/discover E4:B7:F4:11:8D:33
enabling
scanning...
found device: E4:B7:F4:11:8D:33 0 TinyGo
connected to  E4:B7:F4:11:8D:33
discovering services/characteristics
- service 0000180d-0000-1000-8000-00805f9b34fb
-- characteristic 00002a37-0000-1000-8000-00805f9b34fb
- service 00001801-0000-1000-8000-00805f9b34fb
-- characteristic 00002a05-0000-1000-8000-00805f9b34fb

vs. my current code on SD:

found device: E4:B7:F4:11:8D:33 -51 Go HRS
evt: connected in central role
connected to  E4:B7:F4:11:8D:33
discovering services/characteristics
evt: discovered primary service 3
evt: discovered primary service 2
evt: discovered primary service 1
- service 00180000-0000-0000-0000-000000000000
evt: discovered characteristics 3
evt: discovered characteristics 3
evt: discovered characteristics 2
evt: discovered characteristics 1
- service 01180000-0000-0000-0000-000000000000
evt: discovered characteristics 1
- service 0d180000-0000-0000-0000-000000000000
evt: discovered characteristics 1
evt: discovered characteristics 0

First thing I notice is that the short UUIDs on SD do not look like what I was expecting. Is that an error?

In any case, also why are more services being returned? And likewise with the number of characteristics per service? I think I am getting pretty close here.

[deadprogram]

Found the error in converting short UUID to UUID, and pushed a commit to correct that.

Now, here is my debug output:

found device: E4:B7:F4:11:8D:33 -51 Go HRS
evt: connected in central role
connected to  E4:B7:F4:11:8D:33
discovering services/characteristics
evt: discovered primary service 3
evt: discovered primary service 2
evt: discovered primary service 1
- service 00001800-0000-1000-8000-00805f9b34fb
evt: discovered characteristics 3
evt: discovered characteristics 3
evt: discovered characteristics 2
evt: discovered characteristics 1
- service 00001801-0000-1000-8000-00805f9b34fb
evt: discovered characteristics 1
- service 0000180d-0000-1000-8000-00805f9b34fb
evt: discovered characteristics 1
evt: discovered characteristics 0

[deadprogram]

OK, now after testing, I am pretty confident that the SD version is working exactly as it is supposed to with my latest commits:

connected to  E4:B7:F4:11:8D:33
discovering services/characteristics
- service 00001800-0000-1000-8000-00805f9b34fb
-- characteristic 00002a00-0000-1000-8000-00805f9b34fb
-- characteristic 00002a01-0000-1000-8000-00805f9b34fb
-- characteristic 00002a04-0000-1000-8000-00805f9b34fb
-- characteristic 00002aa6-0000-1000-8000-00805f9b34fb
- service 00001801-0000-1000-8000-00805f9b34fb
-- characteristic 00002a05-0000-1000-8000-00805f9b34fb
- service 0000180d-0000-1000-8000-00805f9b34fb
-- characteristic 00002a37-0000-1000-8000-00805f9b34fb
Done.

Note this is the "heartbeat" example.

[aykevl]

Looks good, except that I discovered some unsafe code. Other than that it looks good to me.

[aykevl]

Great! Now we can discover services and characteristics from SoftDevice chips.