Add writeByte. Add support for I2C restarts

[PaulZC]

• Add writeByte for both I2C and SPI - needed by the ADS1219 on I2C
• Add support for I2C restart vs. stop
  • Many devices specify a restart between set-register-address and read-register-value
  • This is I2C-specific and may be needed on non-Arduino platforms too
  • Added setStop and getStop
• Corrected readRegisterRegion
  • The beginTransmission and endTransmission should be inside the if (bFirstInter)

[PaulZC]

TODO: readRegisterRegion should be returning sfeTkError_t to be safe and consistent with the other methods. On error, it currently returns kSTkErrFail - which makes no sense...

I propose changing:

int32_t sfeTkArdI2C::readRegisterRegion(uint8_t devReg, uint8_t *data, size_t numBytes)

to:

sfeTkError_t sfeTkArdI2C::readRegisterRegion(uint8_t devReg, uint8_t *data, size_t numBytes, size_t *readBytes)

and return the number of bytes read in readBytes.

[gigapod]

TODO: readRegisterRegion should be returning sfeTkError_t to be safe and consistent with the other methods. On error, it currently returns kSTkErrFail - which makes no sense...

I propose changing:

int32_t sfeTkArdI2C::readRegisterRegion(uint8_t devReg, uint8_t *data, size_t numBytes)

to:

sfeTkError_t sfeTkArdI2C::readRegisterRegion(uint8_t devReg, uint8_t *data, size_t numBytes, size_t *readBytes)

and return the number of bytes read in readBytes.

Yeah, I missed that in my quick revision last week - and typedef equivalence kept the compiler from complaining.

[gigapod]

I'd move the initialization of _stop{true} to the default ctor - the constructor without arguments(the method above this one).

And since _stop is a uint8_t, should the default value be 1? I would specify the number, not the value of true in the initializer).

[PaulZC]

Ah, yes. 1 is the correct choice. The Arduino source uses a mixture of uint8_t but with bool terminology. Let's stick with 1.

[PaulZC]

If you have time to jump on this, please go for it! Otherwise, I'm very happy to pick this up in the morning.

Long story short, the ToolKit is working very nicely in the ADS1219 library!

[gigapod]

Great to hear it's working !! Thanks for updates/additions.

I'll merge it and tweak the constructor ... and the Arduino version so it get's picked up by the Arduino lib bot

[SFE-Brudnerd]

TODO: readRegisterRegion should be returning sfeTkError_t to be safe and consistent with the other methods. On error, it currently returns kSTkErrFail - which makes no sense...
I propose changing:

int32_t sfeTkArdI2C::readRegisterRegion(uint8_t devReg, uint8_t *data, size_t numBytes)

to:

sfeTkError_t sfeTkArdI2C::readRegisterRegion(uint8_t devReg, uint8_t *data, size_t numBytes, size_t *readBytes)

and return the number of bytes read in readBytes.

Yeah, I missed that in my quick revision last week - and typedef equivalence kept the compiler from complaining.

I'll take a crack at this in my other branch so that I can incorporate that into my UV library too.

[gigapod]

TODO: readRegisterRegion should be returning sfeTkError_t to be safe and consistent with the other methods. On error, it currently returns kSTkErrFail - which makes no sense...
I propose changing:

int32_t sfeTkArdI2C::readRegisterRegion(uint8_t devReg, uint8_t *data, size_t numBytes)

to:

sfeTkError_t sfeTkArdI2C::readRegisterRegion(uint8_t devReg, uint8_t *data, size_t numBytes, size_t *readBytes)

and return the number of bytes read in readBytes.

Yeah, I missed that in my quick revision last week - and typedef equivalence kept the compiler from complaining.

I'll take a crack at this in my other branch so that I can incorporate that into my UV library too.

....I have this written and about to do a PR with some tweaks to the stop message API...