.NET Core SX127X library Part1

Posted on July 22, 2021 by devmobilenz

TransferFullDuplex vs. Read Write

For testing the initial versions of my .Net Core 5 dotnet/iot SX127X library I’m using a Uputronics Raspberry PiZero LoRa(TM) Expansion Board which supports both standard Chip Select(CS) pins (switch selectable which is really useful) and an M2M 1 Channel LoRaWan Gateway Shield for Raspberry PI which has a “non-standard” CS pin.

Uputronics Raspberry PIZero LoRa Expansion board on a Raspberry PI 3 device

M2M Single channel shield on Raspberry Pi 3 Device

The spiDevice.ReadByte() and spiDevice.WriteBye() version worked with a custom chip select pin(25) and CS0 or CS1 selected in the SpiConnectionSettings (but this CS line was “unusable” by other applications). This approach also worked with standard select line (CS01 or CS1) if the SpiConnectionSettings was configured to use the “other” CS line and the selected CS pin managed by the application.

namespace devMobile.IoT.SX127x.ShieldSPIWriteRead
{
	class Program
	{
		private const int SpiBusId = 0;
		private const int ChipSelectLine = 1; // 0 or 1 for Uputronics depends on the switch, for the others choose CS pin not already in use
#if ChipSelectNonStandard
		private const int ChipSelectPinNumber = 25; // 25 for M2M, Dragino etc.
#endif
		private const byte RegisterAddress = 0x6; // RegFrfMsb 0x6c
		//private const byte RegisterAddress = 0x7; // RegFrfMid 0x80
		//private const byte RegisterAddress = 0x8; // RegFrfLsb 0x00
		//private const byte RegisterAddress = 0x42; // RegVersion 0x12

		static void Main(string[] args)
		{
#if ChipSelectNonStandard
			GpioController controller = null;

			controller = new GpioController(PinNumberingScheme.Logical);

			controller.OpenPin(ChipSelectPinNumber, PinMode.Output);
			controller.Write(ChipSelectPinNumber, PinValue.High);
#endif

			var settings = new SpiConnectionSettings(SpiBusId, ChipSelectLine)
			{
				ClockFrequency = 5000000,
				Mode = SpiMode.Mode0,   // From SemTech docs pg 80 CPOL=0, CPHA=0
			};

			SpiDevice spiDevice = SpiDevice.Create(settings);

			Thread.Sleep(500);

			while (true)
			{
#if ChipSelectNonStandard
				controller.Write(ChipSelectPinNumber, PinValue.Low);
#endif

				spiDevice.WriteByte(RegisterAddress);
				byte registerValue = spiDevice.ReadByte();

#if ChipSelectNonStandard
				controller.Write(ChipSelectPinNumber, PinValue.High);
#endif

				byte registerValue = readBuffer[writeBuffer.Length - 1];

				Console.WriteLine($"Register 0x{RegisterAddress:x2} - Value 0X{registerValue:x2} - Bits {Convert.ToString(registerValue, 2).PadLeft(8, '0')}");

				Thread.Sleep(5000);
			}
		}
	}
}

The spiDevice.TransferFullDuplex worked for a standard CS line (CS0 or CS1), and for a non-standard CS line, though the CS line configured in SpiConnectionSettings was “unusable” by other applications “.

namespace devMobile.IoT.SX127x.ShieldSPITransferFullDuplex
{
	class Program
	{
		private const int SpiBusId = 0;
		private const int ChipSelectLine = 0; // 0 or 1 for Uputronics depends on the switch, for the others choose CS pin not already in use
#if ChipSelectNonStandard
		private const int ChipSelectPinNumber = 25; // 25 for M2M, Dragino etc.
#endif
		private const byte RegisterAddress = 0x6; // RegFrfMsb 0x6c
		//private const byte RegisterAddress = 0x7; // RegFrfMid 0x80
		//private const byte RegisterAddress = 0x8; // RegFrfLsb 0x00
		//private const byte RegisterAddress = 0x42; // RegVersion 0x12

		static void Main(string[] args)
		{
#if ChipSelectNonStandard
			GpioController controller = null;

			controller = new GpioController(PinNumberingScheme.Logical);

			controller.OpenPin(ChipSelectPinNumber, PinMode.Output);
			controller.Write(ChipSelectPinNumber, PinValue.High);
#endif

			var settings = new SpiConnectionSettings(SpiBusId, ChipSelectLine)
			{
				ClockFrequency = 5000000,
				Mode = SpiMode.Mode0,   // From SemTech docs pg 80 CPOL=0, CPHA=0
			};

			SpiDevice spiDevice = SpiDevice.Create(settings);

			Thread.Sleep(500);

			while (true)
			{
				byte[] writeBuffer = new byte[] { RegisterAddress, 0 };
				byte[] readBuffer = new byte[writeBuffer.Length];

#if ChipSelectNonStandard
				controller.Write(ChipSelectPinNumber, PinValue.Low);
#endif

				spiDevice.TransferFullDuplex(writeBuffer, readBuffer);

#if ChipSelectNonStandard
				controller.Write(ChipSelectPinNumber, PinValue.High);
#endif

				byte registerValue = readBuffer[writeBuffer.Length - 1];

				Console.WriteLine($"Register 0x{RegisterAddress:x2} - Value 0X{registerValue:x2} - Bits {Convert.ToString(registerValue, 2).PadLeft(8, '0')}");

				Thread.Sleep(5000);
			}
		}
	}
}

The output when the application was working as expected

Loaded '/usr/lib/dotnet/shared/Microsoft.NETCore.App/5.0.4/Microsoft.Win32.Primitives.dll'. Skipped loading symbols. Module is optimized and the debugger option 'Just My Code' is enabled.
Register 0x06 - Value 0X6c - Bits 01101100
Register 0x06 - Value 0X6c - Bits 01101100
Register 0x06 - Value 0X6c - Bits 01101100
Register 0x06 - Value 0X6c - Bits 01101100
Register 0x06 - Value 0X6c - Bits 01101100
Register 0x06 - Value 0X6c - Bits 01101100
The program 'dotnet' has exited with code 0 (0x0).

Summary

Though the spiDevice.TransferFullDuplex code was slightly more complex it worked with both standard and non-standard CS pins.

.NET Core 5 Raspberry PI SPI

Posted on July 19, 2021 by devmobilenz

I have spent a lot of time debugging Serial Peripheral Interface(SPI) device libraries and the .Net Core 5 dotnet/iot library will have its own subtleties(with SPI it’s all about timing). I have written GHI Electronics TinyCLR, Wilderness Labs Meadow, Windows 10 IoT Core, .NET MicroFramework and .NET nanoFramework libraries the SX127X family of devices so building a .Net Core 5 one seemed like a good place to start.

I’m using a Uputronics Raspberry PiZero LoRa(TM) Expansion Board which supports both standard Chip Select(CS) pins (switch selectable which is really useful) and an M2M 1 Channel LoRaWan Gateway Shield for Raspberry PI which has a “non-standard” CS pin.

Uputronics Raspberry PIZero LoRa Expansion board on a Raspberry 3 device

The Uputronics pHat has a pair of Light Emitting Diodes(LEDs) so I adapted some code from a previous post to flash these to confirm the card was working.

static void UputronicsLeds()
{
	const int RedLedPinNumber = 6;
	const int GreenLedPinNumber = 13;

	GpioController controller = new GpioController(PinNumberingScheme.Logical);

	controller.OpenPin(RedLedPinNumber, PinMode.Output);
	controller.OpenPin(GreenLedPinNumber, PinMode.Output);

	while (true)
	{
		if (controller.Read(RedLedPinNumber) == PinValue.Low)
		{
			controller.Write(RedLedPinNumber, PinValue.High);
			controller.Write(GreenLedPinNumber, PinValue.Low);
		}
		else
		{
			controller.Write(RedLedPinNumber, PinValue.Low);
			controller.Write(GreenLedPinNumber, PinValue.High);
		}

		Thread.Sleep(1000);
	}
}

The first Uputronics pHat version using spiDevice.TransferFullDuplex didn’t work. I tried allocating memory for the buffers with new and stackalloc which didn’t seem to make any difference in my trivial example. I tried different Chip Select(CS) pin options, frequencies and modes (the mode used is based on the timings specified in the SX127X datasheet).

static void TransferFullDuplex()
{
	//byte[] writeBuffer = new byte[1]; // Memory allocation didn't seem to make any difference
    //byte[] readBuffer = new byte[1];
	Span<byte> writeBuffer = stackalloc byte[1];
	Span<byte> readBuffer = stackalloc byte[1];

	//var settings = new SpiConnectionSettings(0)
	var settings = new SpiConnectionSettings(0, 0)
	//var settings = new SpiConnectionSettings(0, 1)
	{
		ClockFrequency = 5000000,
		//ClockFrequency = 500000, // Frequency didn't seem to make any difference
		Mode = SpiMode.Mode0,   // From SemTech docs pg 80 CPOL=0, CPHA=0
	};

	SpiDevice spiDevice = SpiDevice.Create(settings);

	Thread.Sleep(500);

	while (true)
	{
		try
		{
			for (byte registerIndex = 0; registerIndex <= 0x42; registerIndex++)
			{
				writeBuffer[0] = registerIndex;
				spiDevice.TransferFullDuplex(writeBuffer, readBuffer);
				//Debug.WriteLine("Register 0x{0:x2} - Value 0X{1:x2} - Bits {2}", writeBuffer[0], readBuffer[0], Convert.ToString(readBuffer[0], 2).PadLeft(8, '0')); // Debug output stopped after roughly 3 times round for loop often debugger would barf as well
				Console.WriteLine("Register 0x{0:x2} - Value 0X{1:x2} - Bits {2}", writeBuffer[0], readBuffer[0], Convert.ToString(readBuffer[0], 2).PadLeft(8, '0'));

				// Would be nice if SpiDevice has a TransferSequential
				/* 
				writeBuffer[0] = registerIndex;
				spiDevice.TransferSequential(writeBuffer, readBuffer);
				Console.WriteLine("Register 0x{0:x2} - Value 0X{1:x2} - Bits {2}", writeBuffer[0], readBuffer[0], Convert.ToString(readBuffer[0], 2).PadLeft(8, '0'));
				*/
			}

			Console.WriteLine("");
			Thread.Sleep(5000);
		}
		catch (Exception ex)
		{
			Console.WriteLine(ex.Message);
		}
	}
}

The second Uputronics pHat version using spiDevice.ReadByte() and spiDevice.WriteBye() didn’t work either.

static void ReadWriteChipSelectStandard()
{
	var settings = new SpiConnectionSettings(0) // Doesn't work
	//	var settings = new SpiConnectionSettings(0, 0) // Doesn't work
	//var settings = new SpiConnectionSettings(0, 1) // Doesn't Work
	{
		ClockFrequency = 5000000,
		ChipSelectLineActiveState = PinValue.Low,
		Mode = SpiMode.Mode0,   // From SemTech docs pg 80 CPOL=0, CPHA=0
	};

	SpiDevice spiDevice = SpiDevice.Create(settings);

	Thread.Sleep(500);

	while (true)
	{
		try
		{
			for (byte registerIndex = 0; registerIndex <= 0x42; registerIndex++)
			{
				spiDevice.WriteByte(registerIndex);
				//Thread.Sleep(5); These made no difference
				//Thread.Sleep(10);
				//Thread.Sleep(20);
				//Thread.Sleep(40);
				byte registerValue = spiDevice.ReadByte();

				Console.WriteLine("Register 0x{0:x2} - Value 0X{1:x2} - Bits {2}", registerIndex, registerValue, Convert.ToString(registerValue, 2).PadLeft(8, '0'));
			}
			Console.WriteLine("");

			Thread.Sleep(5000);
		}
		catch (Exception ex)
		{
			Console.WriteLine(ex.Message);
		}
	}
}

The third Uputronics pHat version using spiDevice.ReadByte() and spiDevice.WriteByte() with DIY Chip Select(CS) worked. In previous SPI device libraries I have found that “managing” the CS line in code can be easier to get working The MicroFramework also has more connectionSettings options for better control of CS line timings which reduces the need for DIY.

static void ReadWriteChipSelectDiy()
{
	const int CSPinNumber = 8; // CS0
	//const int CSPinNumber = 7; // CS1

	// DIY CS0 implented with GPIO pin application controls
	GpioController controller = new GpioController(PinNumberingScheme.Logical);

	controller.OpenPin(CSPinNumber, PinMode.Output);
	//controller.Write(CSPinNumber, PinValue.High);

	//var settings = new SpiConnectionSettings(0) // Doesn't work
	var settings = new SpiConnectionSettings(0, 1) // Works, have to point at unused CS1, this could be a problem is other device on CS1
	//var settings = new SpiConnectionSettings(0, 0) // Works, have to point at unused CS0, this could be a problem is other device on CS0
	{
		ClockFrequency = 5000000,
		Mode = SpiMode.Mode0,   // From SemTech docs pg 80 CPOL=0, CPHA=0
	};

	SpiDevice spiDevice = SpiDevice.Create(settings);

	Thread.Sleep(500);

	while (true)
	{
		try
		{
			for (byte registerIndex = 0; registerIndex <= 0x42; registerIndex++)
			{
				controller.Write(CSPinNumber, PinValue.Low);
				spiDevice.WriteByte(registerIndex);
				//Thread.Sleep(2); // This maybe necessary
				byte registerValue = spiDevice.ReadByte();
				controller.Write(CSPinNumber, PinValue.High);

				Console.WriteLine("Register 0x{0:x2} - Value 0X{1:x2} - Bits {2}", registerIndex, registerValue, Convert.ToString(registerValue, 2).PadLeft(8, '0'));
			}
			Console.WriteLine("");

			Thread.Sleep(5000);
		}
		catch (Exception ex)
		{
			Console.WriteLine(ex.Message);
		}
	}
}

The dotNet/IoT doesn’t support (July2021) the option to “exclusively” open a port so there could be issues with other applications assuming they control CS0/CS1.

Loaded '/usr/lib/dotnet/shared/Microsoft.NETCore.App/5.0.4/Microsoft.Win32.Primitives.dll'. Skipped loading symbols. Module is optimized and the debugger option 'Just My Code' is enabled.
Register 0x00 - Value 0X00 - Bits 00000000
Register 0x01 - Value 0X09 - Bits 00001001
Register 0x02 - Value 0X1a - Bits 00011010
Register 0x03 - Value 0X0b - Bits 00001011
Register 0x04 - Value 0X00 - Bits 00000000
Register 0x05 - Value 0X52 - Bits 01010010
Register 0x06 - Value 0X6c - Bits 01101100
Register 0x07 - Value 0X80 - Bits 10000000
Register 0x08 - Value 0X00 - Bits 00000000
Register 0x09 - Value 0X4f - Bits 01001111
Register 0x0a - Value 0X09 - Bits 00001001
Register 0x0b - Value 0X2b - Bits 00101011
Register 0x0c - Value 0X20 - Bits 00100000
Register 0x0d - Value 0X08 - Bits 00001000
Register 0x0e - Value 0X02 - Bits 00000010
Register 0x0f - Value 0X0a - Bits 00001010
Register 0x10 - Value 0Xff - Bits 11111111
Register 0x11 - Value 0X70 - Bits 01110000
Register 0x12 - Value 0X15 - Bits 00010101
Register 0x13 - Value 0X0b - Bits 00001011
Register 0x14 - Value 0X28 - Bits 00101000
Register 0x15 - Value 0X0c - Bits 00001100
Register 0x16 - Value 0X12 - Bits 00010010
Register 0x17 - Value 0X47 - Bits 01000111
Register 0x18 - Value 0X32 - Bits 00110010
Register 0x19 - Value 0X3e - Bits 00111110
Register 0x1a - Value 0X00 - Bits 00000000
Register 0x1b - Value 0X00 - Bits 00000000
Register 0x1c - Value 0X00 - Bits 00000000
Register 0x1d - Value 0X00 - Bits 00000000
Register 0x1e - Value 0X00 - Bits 00000000
Register 0x1f - Value 0X40 - Bits 01000000
Register 0x20 - Value 0X00 - Bits 00000000
Register 0x21 - Value 0X00 - Bits 00000000
Register 0x22 - Value 0X00 - Bits 00000000
Register 0x23 - Value 0X00 - Bits 00000000
Register 0x24 - Value 0X05 - Bits 00000101
Register 0x25 - Value 0X00 - Bits 00000000
Register 0x26 - Value 0X03 - Bits 00000011
Register 0x27 - Value 0X93 - Bits 10010011
Register 0x28 - Value 0X55 - Bits 01010101
Register 0x29 - Value 0X55 - Bits 01010101
Register 0x2a - Value 0X55 - Bits 01010101
Register 0x2b - Value 0X55 - Bits 01010101
Register 0x2c - Value 0X55 - Bits 01010101
Register 0x2d - Value 0X55 - Bits 01010101
Register 0x2e - Value 0X55 - Bits 01010101
Register 0x2f - Value 0X55 - Bits 01010101
Register 0x30 - Value 0X90 - Bits 10010000
Register 0x31 - Value 0X40 - Bits 01000000
Register 0x32 - Value 0X40 - Bits 01000000
Register 0x33 - Value 0X00 - Bits 00000000
Register 0x34 - Value 0X00 - Bits 00000000
Register 0x35 - Value 0X0f - Bits 00001111
Register 0x36 - Value 0X00 - Bits 00000000
Register 0x37 - Value 0X00 - Bits 00000000
Register 0x38 - Value 0X00 - Bits 00000000
Register 0x39 - Value 0Xf5 - Bits 11110101
Register 0x3a - Value 0X20 - Bits 00100000
Register 0x3b - Value 0X82 - Bits 10000010
Register 0x3c - Value 0Xf6 - Bits 11110110
Register 0x3d - Value 0X02 - Bits 00000010
Register 0x3e - Value 0X80 - Bits 10000000
Register 0x3f - Value 0X40 - Bits 01000000
Register 0x40 - Value 0X00 - Bits 00000000
Register 0x41 - Value 0X00 - Bits 00000000
Register 0x42 - Value 0X12 - Bits 00010010

The fourth Uputronics pHat version using spiDevice.TransferFullDuplex with read and write buffers two bytes long and the leading bye of the response ignored worked.

...
while (true)
{
	try
	{
		for (byte registerIndex = 0; registerIndex <= 0x42; registerIndex++)
		{
			// Doesn't work
			writeBuffer[0] = registerIndex;
			spiDevice.TransferFullDuplex(writeBuffer, readBuffer);
			Console.WriteLine("Register 0x{0:x2} - Value 0X{1:x2} - Bits {2}", registerIndex, readBuffer[0], Convert.ToString(readBuffer[0], 2).PadLeft(8, '0'));

			// Does work
			writeBuffer[0] = registerIndex;
			spiDevice.TransferFullDuplex(writeBuffer, readBuffer);
			Console.WriteLine("Register 0x{0:x2} - Value 0X{1:x2} - Bits {2}", registerIndex, readBuffer[1], Convert.ToString(readBuffer[1], 2).PadLeft(8, '0'));

			// Does work
			writeBuffer[1] = registerIndex;
			spiDevice.TransferFullDuplex(writeBuffer, readBuffer);
			Console.WriteLine("Register 0x{0:x2} - Value 0X{1:x2} - Bits {2}", registerIndex, readBuffer[1], Convert.ToString(readBuffer[1], 2).PadLeft(8, '0'));

			Console.WriteLine("");
		}

		Console.WriteLine("");
		Thread.Sleep(5000);
	}
	catch (Exception ex)
	{
		Console.WriteLine(ex.Message);
	}
}

Register 0x00 - Value 0X00 - Bits 00000000
Register 0x00 - Value 0X00 - Bits 00000000
Register 0x00 - Value 0X00 - Bits 00000000

...

Register 0x42 - Value 0X00 - Bits 00000000
Register 0x42 - Value 0X12 - Bits 00010010
Register 0x42 - Value 0X12 - Bits 00010010

The first M2M pHat version using SpiDevice.Read and SpiDevice.Write with a “custom” CS pin worked.

...
// Chip select with pin which isn't CS0 or CS1 needs M2M shield
static void ReadWriteDiyChipSelectNonStandard()
{
	const int CSPinNumber = 25;

	// DIY CS0 implented with GPIO pin application controls
	GpioController controller = new GpioController(PinNumberingScheme.Logical);

	controller.OpenPin(CSPinNumber, PinMode.Output);
	//controller.Write(CSPinNumber, PinValue.High);

	// Work, this could be a problem is other device on CS0/CS1
	var settings = new SpiConnectionSettings(0)
	//var settings = new SpiConnectionSettings(0, 0) 
	//var settings = new SpiConnectionSettings(0, 1) 
	{
		ClockFrequency = 5000000,
		Mode = SpiMode.Mode0,   // From SemTech docs pg 80 CPOL=0, CPHA=0
	};

	SpiDevice spiDevice = SpiDevice.Create(settings);

	Thread.Sleep(500);

	while (true)
	{
		try
		{
			for (byte registerIndex = 0; registerIndex <= 0x42; registerIndex++)
			{
				controller.Write(CSPinNumber, PinValue.Low);
				spiDevice.WriteByte(registerIndex);
				//Thread.Sleep(2); // This maybe necessary
				byte registerValue = spiDevice.ReadByte();
				controller.Write(CSPinNumber, PinValue.High);

				Console.WriteLine("Register 0x{0:x2} - Value 0X{1:x2} - Bits {2}", registerIndex, registerValue, Convert.ToString(registerValue, 2).PadLeft(8, '0'));
			}
			Console.WriteLine("");

			Thread.Sleep(5000);
		}
		catch (Exception ex)
		{
			Console.WriteLine(ex.Message);
		}
	}
}

The second M2M pHat version using SpiDevice.TransferFullDuplex with a “custom” CS pin also worked.

while (true)
{
	try
	{
		for (byte registerIndex = 0; registerIndex <= 0x42; registerIndex++)
		{
			writeBuffer[0] = registerIndex;
			//writeBuffer[1] = registerIndex;

			controller.Write(CSPinNumber, PinValue.Low);
			spiDevice.TransferFullDuplex(writeBuffer, readBuffer);
			controller.Write(CSPinNumber, PinValue.High);

			Console.WriteLine("Register 0x{0:x2} - Value 0X{1:x2} - Bits {2}", registerIndex, readBuffer[1], Convert.ToString(readBuffer[1], 2).PadLeft(8, '0'));
		}
		Console.WriteLine("");

		Thread.Sleep(5000);
	}
	catch (Exception ex)
	{
		Console.WriteLine(ex.Message);
	}
}

The next step was to read an array of bytes, using spiDevice .TransferFullDuplex. The SX127X transmit/receive frequency is specified in registers 0x06 RegFrMSB, 0x07 RegFrMid, and 0x08 RegFrLsb. The default frequency is 868MHz which is 0xE4, 0xC0, 0x00

static void TransferFullDuplexBufferBytesRead()
{ 
	const byte length = 3;
	byte[] writeBuffer = new byte[length + 1];
	byte[] readBuffer = new byte[length + 1];

	// Read the frequency which is 3 bytes RegFrMsb 0x6c, RegFrMid 0x80, RegFrLsb 0x00
	writeBuffer[0] = 0x06; //

	// Works, have to point at unused CS0/CS1, others could be a problem is another another SPI device is on on CS0/CS1
	//var settings = new SpiConnectionSettings(0)
	var settings = new SpiConnectionSettings(0, 0) 
	//var settings = new SpiConnectionSettings(0, 1) 
	{
		ClockFrequency = 5000000,
		Mode = SpiMode.Mode0,   // From SemTech docs pg 80 CPOL=0, CPHA=0
	};

	SpiDevice spiDevice = SpiDevice.Create(settings);

	spiDevice.TransferFullDuplex(writeBuffer, readBuffer);

	Console.WriteLine($"Register 0x06-0x{readBuffer[1]:x2} 0x07-0x{readBuffer[2]:x2} 0x08-0x{readBuffer[3]:x2}");
}

-------------------------------------------------------------------
You may only use the Microsoft .NET Core Debugger (vsdbg) with
Visual Studio Code, Visual Studio or Visual Studio for Mac software
to help you develop and test your applications.
-------------------------------------------------------------------
Loaded '/usr/lib/dotnet/shared/Microsoft.NETCore.App/5.0.4/System.Private.CoreLib.dll'. Skipped loading symbols. Module is optimized and the debugger option 'Just My Code' is enabled.
...
Loaded '/usr/lib/dotnet/shared/Microsoft.NETCore.App/5.0.4/Microsoft.Win32.Primitives.dll'. Skipped loading symbols. Module is optimized and the debugger option 'Just My Code' is enabled.
Register 0x06-0xe4 0x07-0xc0 0x08-0x00

The final step was write an array of bytes, using spiDevice.TransferFullDuplex to change the transmit/receive frequency to 915MHz. To write a value the first bit of the address byte must be set to 1 hence the 0x86 RegFrMsb address.

static void TransferFullDuplexBufferBytesWrite()
{
	const byte length = 3;
	byte[] writeBuffer = new byte[length + 1];
	byte[] readBuffer = new byte[length + 1];

	// Write the frequency which is 3 bytes RegFrMsb 0x6c, RegFrMid 0x80, RegFrLsb or with 0x00 the write mask
	writeBuffer[0] = 0x86 ;

	// Works, have to point at unused CS0/CS1, others could be a problem is another another SPI device is on on CS0/CS1
	//var settings = new SpiConnectionSettings(0)
	var settings = new SpiConnectionSettings(0, 0)
	//var settings = new SpiConnectionSettings(0, 1) 
	{
		ClockFrequency = 5000000,
		Mode = SpiMode.Mode0,   // From SemTech docs pg 80 CPOL=0, CPHA=0
	};

	SpiDevice spiDevice = SpiDevice.Create(settings);

	// Set the frequency to 915MHz
	writeBuffer[1] = 0xE4;
	writeBuffer[2] = 0xC0;
	writeBuffer[3] = 0x00;

	spiDevice.TransferFullDuplex(writeBuffer, readBuffer);
}

-------------------------------------------------------------------
You may only use the Microsoft .NET Core Debugger (vsdbg) with
Visual Studio Code, Visual Studio or Visual Studio for Mac software
to help you develop and test your applications.
-------------------------------------------------------------------
Loaded '/usr/lib/dotnet/shared/Microsoft.NETCore.App/5.0.4/System.Private.CoreLib.dll'. Skipped loading symbols. Module is optimized and the debugger option 'Just My Code' is enabled.
...
Loaded '/usr/lib/dotnet/shared/Microsoft.NETCore.App/5.0.4/Microsoft.Win32.Primitives.dll'. Skipped loading symbols. Module is optimized and the debugger option 'Just My Code' is enabled.
Register 0x06-0x6c 0x07-0x80 0x08-0x00
Register 0x06-0xe4 0x07-0xc0 0x08-0x00
The program 'dotnet' has exited with code 0 (0x0).

Summary

This exceptionally long post was to highlight that with SPI it’s all about timing, first read the datasheet, then build code to validate your understanding.

Some platforms have native TransferSequential implementations but the dotNet/IoT library only has TransferFullDuplex. SPI hardware is always full duplex, if “sequential” is available the implementation will write the provided bytes and then follow them with zeros to read the requested bytes.

.NET Core 5 Raspberry PI GPIO Interrupts

Posted on July 18, 2021 by devmobilenz

To port my Windows 10 IoT Core nRF24L01, SX123X. and SX127X LoRa libraries to .Net Core 5 I wanted to see if there were any differences in the way interrupts were handled by the dotnet/iot libraries. The initial versions of the code will being running on a Raspberry PI but I will also look at other supported Single Board Computers(SBCs).

My test-rig was a RaspberryPI 3B with a Grove Base Hat for Raspberry PI (left over from a proof of concept project), a couple of Grove Universal 4 pin 5CM cables, a Grove LED pack, and a Grove Button.

Raspberry PI test rig with Grove Base pHat, button & LED

There were some syntax differences but nothing to major

using System;
using System.Device.Gpio;
using System.Diagnostics;
using System.Threading;

namespace devMobile.NetCore.GPIOInterrupts
{
	class Program
	{
		private const int ButtonPinNumber = 5;
		private const int LedPinNumber = 16;
		private static GpioController gpiocontroller;

		static void Main(string[] args)
		{
			try
			{
				gpiocontroller = new GpioController(PinNumberingScheme.Logical);

				gpiocontroller.OpenPin(ButtonPinNumber, PinMode.InputPullDown);
				gpiocontroller.OpenPin(LedPinNumber, PinMode.Output);

				gpiocontroller.RegisterCallbackForPinValueChangedEvent(ButtonPinNumber, PinEventTypes.Rising, PinChangeEventHandler);

				Console.WriteLine($"Main thread:{Thread.CurrentThread.ManagedThreadId}");

				while (true)
				{
					Console.WriteLine($"Doing stuff");
					Thread.Sleep(1000);
				}
			}
			catch (Exception ex)
			{
				Console.WriteLine(ex.Message);
			}
		}

		private static void PinChangeEventHandler(object sender, PinValueChangedEventArgs pinValueChangedEventArgs)
		{
			Debug.Write($"Interrupt Thread:{Thread.CurrentThread.ManagedThreadId}");

			if (pinValueChangedEventArgs.ChangeType == PinEventTypes.Rising)
			{
				if (gpiocontroller.Read(LedPinNumber) == PinValue.Low)
				{
					gpiocontroller.Write(LedPinNumber, PinValue.High);
				}
				else
				{
					gpiocontroller.Write(LedPinNumber, PinValue.Low);
				}
			}
		}
	}
}

I included code to display the Thread.CurrentThread.ManagedThreadId to see if the callback was running on a different thread.

-------------------------------------------------------------------
You may only use the Microsoft .NET Core Debugger (vsdbg) with
Visual Studio Code, Visual Studio or Visual Studio for Mac software
to help you develop and test your applications.
-------------------------------------------------------------------
...
Loaded '/usr/lib/dotnet/shared/Microsoft.NETCore.App/5.0.4/Microsoft.Win32.Primitives.dll'. Skipped loading symbols. Module is optimized and the debugger option 'Just My Code' is enabled.
Main thread:1
Doing stuff
Doing stuff
Doing stuff
Doing stuff
Doing stuff
Interrupt Thread:6Doing stuff
Doing stuff
Doing stuff
Interrupt Thread:6Doing stuff
Doing stuff
Interrupt Thread:6Doing stuff
Doing stuff
Doing stuff
Doing stuff
Doing stuff
Doing stuff
The program 'dotnet' has exited with code 0 (0x0).

The ManagedThreadId for the main loop(1) was different to the callback(6) which needs some further investigation.

.NET Core 5 Raspberry PI GPIO

Posted on July 17, 2021 by devmobilenz

Next to my desk I have a stack of Raspberry PI’s and with the release of .Net Core 5 for Windows, Macintosh and Linux I decided to have another look at porting some of my nRF24L01, LoRa, and LoRaWAN libraries to .Net Core.

There are blog posts (like Deploying and Debugging Raspberry Pi .NET Applications using VS Code) about installing .Net core on a Raspberry PI, using Visual Studio Code to write an application, then deploying and debugging it over SSH which were interesting but there were a lot of steps so the likelihood me screwing up was high.

I have been using Visual Studio for C# and VB.Net code since .Net was first released (I wrote my first C# applications with Visual Studio 6) so when I stumbled across RaspberryDebugger it was time to unbox a Raspberry PI 3B and see what happened.

All coding demos start with Hello world…

using System;
using System.Diagnostics;
using System.Threading;

namespace devMobile.NetCore.ConsoleApp
{
	class Program
	{
		static void Main(string[] args)
		{
			while (true)
			{
				Console.WriteLine($"{DateTime.UtcNow:HH:mm:ss} Hello World!");

				Thread.Sleep(1000);
			}
		}
	}
}

The RaspberryDebugger is really simple to install, and “frictionless” to use. The developers have put a lot of effort into making it easy to deploy and debug a .Net Core application running on a Raspberry PI with Visual Studio. All I had to do was search for, then download and install their Visual Studio Extension(VSIX).

Then configure the connection information for the devices I wanted to use.

Visual Studio Options menu for RaspberryDebugger

On my main development system I was using multiple Raspberry PI devices so it was great to be able to pre-configure several devices.

RaspberryDebugger device(s) configuration)

I had connected to each device with PuTTY to check that connectivity was sorted.

RaspberryDebugger devices configuration device configuration

After typing in my “Hello world” application I had to select the device I wanted to use

Then I pressed F5 and it worked! It’s very unusual for things to work first time so I was stunned. The application was “automagically” downloaded and run in the debugger on the device.

-------------------------------------------------------------------
You may only use the Microsoft .NET Core Debugger (vsdbg) with
Visual Studio Code, Visual Studio or Visual Studio for Mac software
to help you develop and test your applications.
-------------------------------------------------------------------
Loaded '/usr/lib/dotnet/shared/Microsoft.NETCore.App/5.0.4/System.Private.CoreLib.dll'. Skipped loading symbols. Module is optimized and the debugger option 'Just My Code' is enabled.
Loaded '/home/pi/vsdbg/ConsoleApp/ConsoleApp.dll'. Symbols loaded.
Loaded '/usr/lib/dotnet/shared/Microsoft.NETCore.App/5.0.4/System.Runtime.dll'. Skipped loading symbols. Module is optimized and the debugger option 'Just My Code' is enabled.
Loaded '/usr/lib/dotnet/shared/Microsoft.NETCore.App/5.0.4/System.Console.dll'. Skipped loading symbols. Module is optimized and the debugger option 'Just My Code' is enabled.
Loaded '/usr/lib/dotnet/shared/Microsoft.NETCore.App/5.0.4/System.Threading.Thread.dll'. Skipped loading symbols. Module is optimized and the debugger option 'Just My Code' is enabled.
Loaded '/usr/lib/dotnet/shared/Microsoft.NETCore.App/5.0.4/System.Threading.dll'. Skipped loading symbols. Module is optimized and the debugger option 'Just My Code' is enabled.
Loaded '/usr/lib/dotnet/shared/Microsoft.NETCore.App/5.0.4/System.Text.Encoding.Extensions.dll'. Skipped loading symbols. Module is optimized and the debugger option 'Just My Code' is enabled.
Loaded '/usr/lib/dotnet/shared/Microsoft.NETCore.App/5.0.4/Microsoft.Win32.Primitives.dll'. Skipped loading symbols. Module is optimized and the debugger option 'Just My Code' is enabled.
05:50:37 Hello World!
05:50:39 Hello World!
05:50:40 Hello World!
05:50:41 Hello World!
05:50:42 Hello World!
05:50:43 Hello World!
...

Once the basics were sorted I wanted to check out the General Purpose Input & Output(GPIO) support implemented in the dotnet/iot libraries. My test-rig was a RaspberryPI 3B with a Grove Base Hat for Raspberry PI (left over from a Windows 10 IoT Core proof of concept project), a couple of Grove Universal 4 pin 5CM cables, a Grove LED pack, and a Grove Button.

using System;
using System.Device.Gpio;
using System.Diagnostics;
using System.Threading;

namespace devMobile.NetCore.ConsoleGPIO1
{
	class Program
	{
		const int ButtonPinNumber = 5;
		const int LedPinNumber = 16;

		static void Main(string[] args)
		{
			try
			{
				GpioController controller = new GpioController(PinNumberingScheme.Logical);

				controller.OpenPin(ButtonPinNumber, PinMode.InputPullUp);
				controller.OpenPin(LedPinNumber, PinMode.Output);

				while (true)
				{
					if (controller.Read(ButtonPinNumber) == PinValue.High)
					{
						if (controller.Read(LedPinNumber) == PinValue.Low)
						{
							controller.Write(LedPinNumber, PinValue.High);
						}
						else
						{
							controller.Write(LedPinNumber, PinValue.Low);
						}
					}
					Thread.Sleep(100);
				}
			}
			catch (Exception ex)
			{
				Console.WriteLine(ex.Message);
			}
		}
	}
}

After starting the application, when I pressed the button the Grove LED flashed with a 100mSec duty cycle.

The RaspberryDebugger extension is a joy to use and I’m going to figure out how I can donate some money to the developers.

Seeed LoRa-E5 LowPower problem fix

Posted on July 10, 2021 by devmobilenz

I had been soak testing Seeed LoRa-E5 equipped TinyCLR and netNF devices for the last couple of weeks and after approximately two days they would stop sending data.

After a pointer to the LowPower section of the Seeed LoRa-E5 manual I realised my code could send the next command within 5ms.

I added a 5msec Sleep after the wakeup command had been sent

public Result Wakeup()
{
   // Wakeup the E5 Module
#if DIAGNOSTICS
   Debug.WriteLine($" {DateTime.UtcNow:hh:mm:ss} AT+LOWPOWER: WAKEUP");
#endif
   Result result = SendCommand("+LOWPOWER: WAKEUP", $"A", CommandTimeoutDefault);
   if (result != Result.Success)
   {
#if DIAGNOSTICS
      Debug.WriteLine($" {DateTime.UtcNow:hh:mm:ss} AT+LOWPOWER: WAKEUP failed {result}");
#endif
      return result;
   }

   // Thanks AndrewL for pointing out delay required in section 4.30 LOWPOWER
   Thread.Sleep(5);

   return Result.Success;
}

The updated code is has been reliably running on TinyCLR and netNF devices connected to The Things Industries for the last 4 days.

Seeed LoRa-E5 LowPower problem

Posted on July 7, 2021 by devmobilenz

I have been soak testing Seeed LoRa-E5 equipped TinyCLR and netNF devices for the last couple of weeks and after approximately two days they would stop sending data. The code is still running but The Things Industries device live data tab is empty.

I have reviewed the code line by line and it looks okay. When I run the application on the device with the debugger attached the device does not stop transmitting (a heisenbug) which is a problem.

First step was to disable the sleep/wakeup power conservation calls and see what happens

#if PAYLOAD_COUNTER
   Debug.WriteLine($"{DateTime.UtcNow:hh:mm:ss} Send Timeout:{SendTimeout.TotalSeconds} Seconds payload Counter:{PayloadCounter}");
#if CONFIRMED
   result = device.Send(BitConverter.GetBytes(PayloadCounter), true, SendTimeout);
#else
   result = device.Send(BitConverter.GetBytes(PayloadCounter), false, SendTimeout);
#endif
   PayloadCounter += 1;
#endif

   if (result != Result.Success)
   {
      Debug.WriteLine($"Send failed {result}");
   }

   /*
   Debug.WriteLine($"{DateTime.UtcNow:hh:mm:ss} Sleep");
   result = device.Sleep();
   if (result != Result.Success)
   {
      Debug.WriteLine($"Sleep failed {result}");
      return;
   }
   */
   Thread.Sleep(60000);
   /*
   Debug.WriteLine($"{DateTime.UtcNow:hh:mm:ss} Wakeup");
   result = device.Wakeup();
   if (result != Result.Success)
   {
      Debug.WriteLine($"Wakeup failed {result}");
      return;
   }
   */

The devices have now been running fine for 4.5 days so it looks like it might be something todo with entering and/or exiting low power mode.

.NET Core web API + Dapper – Image Upload

Posted on July 1, 2021 by devmobilenz

Stream of Bytes or Base64 Encoded

To test my Dapper based functionality to upload images to my World Wide Importers database I used Telerik Fiddler.

Fiddler Composer with the image field name and upload file button highlighted

The currentimplementation only supports the uploading of one image at a time in a field called “image”.

This implementation supports a “Content-Type” of “application/octet-stream” or “image/jpeg”.

[HttpPost("{id}/image")]
public async Task<ActionResult> Upload([FromRoute(Name = "id")][Range(1, int.MaxValue, ErrorMessage = "StockItem id must greater than 0")] int id, [FromForm] IFormFile image)
{
	if (image == null) 
	{
		return this.BadRequest("Image image file missing");
	}

	if (image.Length == 0)
	{
		return this.BadRequest("Image image file is empty");
	}

	if ((string.Compare(image.ContentType, "application/octet-stream",true) != 0) && (string.Compare(image.ContentType, "image/jpeg", true) != 0))
	{
		return this.BadRequest("Image image file content-type is not application/octet-stream or image/jpeg");
	}

	try
	{
		using (MemoryStream ms = new MemoryStream())
		{
			await image.CopyToAsync(ms);

			ms.Seek(0, SeekOrigin.Begin);

			using (SqlConnection db = new SqlConnection(this.connectionString))
			{
				DynamicParameters parameters = new DynamicParameters();

				parameters.Add("StockItemId", id);
				parameters.Add("photo", ms, DbType.Binary, ParameterDirection.Input);

				await db.ExecuteAsync(sql: @"UPDATE [WareHouse].[StockItems] SET [Photo]=@Photo WHERE StockItemID=@StockItemId", param: parameters, commandType: CommandType.Text);
			}
		}
	}
	catch (SqlException ex)
	{
		logger.LogError(ex, "Updating photo of StockItem with ID:{0}", id);

		return this.StatusCode(StatusCodes.Status500InternalServerError);
	}

	return this.Ok();
}

After uploading the image I could download it as either a stream of bytes(displayed in Fiddler) or Base64 encoded (this had to be converted to an image)

Fiddler displaying downloaded jpeg image

This implementation doesn’t support the uploading of multiple images or the streaming of larger images but would be sufficient for uploading thumbnails etc.

NOTE : Error Handling approach has been updated

.NET Core web API + Dapper – Image Download

Posted on June 30, 2021 by devmobilenz

Stream of Bytes and Base64 Encoded

I needed to add some code using Dapper to retrieve images stored in a database for a webby client. The stockItems table has a column for a photo but they were all null…

CREATE TABLE [Warehouse].[StockItems](
	[StockItemID] [int] NOT NULL,
	[StockItemName] [nvarchar](100) NOT NULL,
	[SupplierID] [int] NOT NULL,
	[ColorID] [int] NULL,
	[UnitPackageID] [int] NOT NULL,
	[OuterPackageID] [int] NOT NULL,
	[Brand] [nvarchar](50) NULL,
	[Size] [nvarchar](20) NULL,
	[LeadTimeDays] [int] NOT NULL,
	[QuantityPerOuter] [int] NOT NULL,
	[IsChillerStock] [bit] NOT NULL,
	[Barcode] [nvarchar](50) NULL,
	[TaxRate] [decimal](18, 3) NOT NULL,
	[UnitPrice] [decimal](18, 2) NOT NULL,
	[RecommendedRetailPrice] [decimal](18, 2) NULL,
	[TypicalWeightPerUnit] [decimal](18, 3) NOT NULL,
	[MarketingComments] [nvarchar](max) NULL,
	[InternalComments] [nvarchar](max) NULL,
	[Photo] [varbinary](max) NULL,
	[CustomFields] [nvarchar](max) NULL,
	[Tags]  AS (json_query([CustomFields],N'$.Tags')),
	[SearchDetails]  AS (concat([StockItemName],N' ',[MarketingComments])),
	[LastEditedBy] [int] NOT NULL,
	[ValidFrom] [datetime2](7) GENERATED ALWAYS AS ROW START NOT NULL,
	[ValidTo] [datetime2](7) GENERATED ALWAYS AS ROW END NOT NULL,
 CONSTRAINT [PK_Warehouse_StockItems] PRIMARY KEY CLUSTERED 
(
	[StockItemID] ASC
)

I uploaded images of three different colours of sellotape dispensers with the following SQL

UPDATE Warehouse.StockItems 
SET [Photo] =(SELECT * FROM Openrowset( Bulk 'C:\Users\BrynLewis\Pictures\TapeDispenserBlue.jpg', Single_Blob) as  MyImage) where StockItemID = 

-- 203	Tape dispenser (Black)
-- 204	Tape dispenser (Red)
-- 205	Tape dispenser (Blue)

There are two options for downloading the image. The first is as a stream of bytes

[HttpGet("{id}/image")]
public async Task<ActionResult> GetImage([Range(1, int.MaxValue, ErrorMessage = "StockItem id must greater than 0")] int id)
{
	Byte[] response;

	try
	{
		using (SqlConnection db = new SqlConnection(this.connectionString))
		{
			response = await db.ExecuteScalarAsync<byte[]>(sql: @"SELECT [Photo] as ""photo"" FROM [Warehouse].[StockItems] WHERE StockItemID=@StockItemId", param: new { StockItemId = id }, commandType: CommandType.Text);
		}

		if (response == default)
		{
			logger.LogInformation("StockItem:{0} image not found", id);

			return this.NotFound($"StockItem:{id} image not found");
		}
	}
	catch (SqlException ex)
	{
		logger.LogError(ex, "Looking up a StockItem:{0} image", id);

		return this.StatusCode(StatusCodes.Status500InternalServerError);
	}

	return File(response, "image/jpeg");
}

The second is a Base64 encoded stream of bytes

[HttpGet("{id}/base64")]
public async Task<ActionResult> GetBase64([Range(1, int.MaxValue, ErrorMessage = "Stock item id must greater than 0")] int id)
{
	Byte[] response;

	try
	{
		using (SqlConnection db = new SqlConnection(this.connectionString))
		{
			response = await db.ExecuteScalarAsync<byte[]>(sql: @"SELECT [Photo] as ""photo"" FROM [Warehouse].[StockItems] WHERE StockItemID=@StockItemId", param: new { StockItemId = id }, commandType: CommandType.Text);
		}

		if (response == default)
		{
			logger.LogInformation("StockItem:{0} Base64 not found", id);

			return this.NotFound($"StockItem:{id} image not found");
		}
	}
	catch (SqlException ex)
	{
		logger.LogError(ex, "Looking up a StockItem withID:{0} base64", id);

		return this.StatusCode(StatusCodes.Status500InternalServerError);
	}

	return Ok("data:image/jpeg;base64," + Convert.ToBase64String(response));
}

I lost an hour from my life I will never get back figuring out that a correctly formatted/spelt content types “image/jpeg” and “data:image/jpeg;base64” were key to getting the webby client to render image.

NOTE : Error Handling approach has been updated

.NET Core web API + Dapper – Polly Retries

Posted on June 29, 2021 by devmobilenz

Recovering from transient failures with Polly

It’s not uncommon for SQL Azure servers and databases to suffer from “transient failures”. In application logs I have seen these occur during scale up/down events, periods where my application’s performance has been temporarily impacted (but its throughput has not changed), which I assume has been some load balancing going on in the background and when network connectivity has been a bit flakey.

Microsoft has published guidance for building Microservices applications, troubleshooting common AzureSQL errors and improving the resilience of ADO.Net connections which cover different approaches in depth.

For many years I used the Microsoft Enterprise Library Transient Fault Handling Application Block (TOPAZ), then upgraded to the .Net Core Version built by Mo Chavoshi both of which have been retired.

Now I’m using The Polly Project which builds on the concepts of TOPAZ but has been thoroughly re-engineered with lots of extensibility, an active community and modern codebase. Inspired by Ben Hyrman and several other developers I have built a minimalist wrapper for the Dapper Async methods which detects transient errors using the same approach as the Entity Framework Core library.

public static Task<int> ExecuteWithRetryAsync(
			  this IDbConnection connection,
			  string sql,
			  object param = null,
			  IDbTransaction transaction = null,
			  int? commandTimeout = null,
			  CommandType? commandType = null) => RetryPolicy.ExecuteAsync(() => connection.ExecuteAsync(sql, param, transaction, commandTimeout, commandType));

I did think about retry functionality for async methods which returned object/dynamic but have only implemented strongly typed ones for the initial version.

[HttpGet]
public async Task<ActionResult<IEnumerable<Model.StockItemListDtoV1>>> Get()
{
	IEnumerable<Model.StockItemListDtoV1> response = null;

	try
	{
		using (SqlConnection db = new SqlConnection(this.connectionString))
		{
			response = await db.QueryWithRetryAsync<Model.StockItemListDtoV1>(sql: @"SELECT [StockItemID] as ""ID"", [StockItemName] as ""Name"", [RecommendedRetailPrice], [TaxRate] FROM [Warehouse].[StockItems]", commandType: CommandType.Text);
		}
	}
	catch (SqlException ex)
	{
		logger.LogError(ex, "Retrieving list of StockItems");

		return this.StatusCode(StatusCodes.Status500InternalServerError);
	}

	return this.Ok(response);
}

I have struggled to get reproduceable transient failures without pausing execution in the Visual Studio debugger and tinkering with variables or scaling up/down my databases (limit to how often this can be done) or unplugging the network cable at the wrong time.

NOTE : Error Handling approach has been updated

.NET Core web API + Dapper – Web Caching

Posted on June 27, 2021 by devmobilenz

Web cache validation with eTags

On a couple of the systems I work on there are a number of queries (often complex spatial searches) which are very resource intensive but are quite readily cached. In these systems we have used HTTP GET and HEAD Request methods together so that the client only re-GETs the query results after a HEAD method indicates there have been updates.

I have been trying to keep the number of changes to my Microsoft SQL Azure World Wide Importers database to a minimum but for this post I have added a rowversion column to the StockGroups table. The rowversion data type is an automatically generated, unique 8 byte binary(12 bytes Base64 encoded) number within a database.

Adding a rowversion table to an existing System Versioned table in the SQL Server Management Studio Designer is painful so I used…

ALTER TABLE [Warehouse].[StockGroups] ADD [Version] [timestamp] NULL

To reduce complexity the embedded SQL is contains two commands (normally I wouldn’t do this) one for retrieving the list StockGroups the other for retrieving the maximum StockGroup rowversion. If a StockGroup is changed the rowversion will be “automagically” updated and the maximum value will change.

[HttpGet]
public async Task<ActionResult<IEnumerable<Model.StockGroupListDtoV1>>> Get()
{
	IEnumerable<Model.StockGroupListDtoV1> response = null;

	try
	{
		using (SqlConnection db = new SqlConnection(this.connectionString))
		{
			var parameters = new DynamicParameters();

			parameters.Add("@RowVersion", dbType: DbType.Binary, direction: ParameterDirection.Output, size: ETagBytesLength);

			response = await db.QueryAsync<Model.StockGroupListDtoV1>(sql: @"SELECT [StockGroupID] as ""ID"", [StockGroupName] as ""Name""FROM [Warehouse].[StockGroups] ORDER BY Name; SELECT @RowVersion=MAX(Version) FROM [Warehouse].[StockGroups]", param: parameters, commandType: CommandType.Text);

			if (response.Any())
			{
				byte[] rowVersion = parameters.Get<byte[]>("RowVersion");

				this.HttpContext.Response.Headers.Add("ETag", Convert.ToBase64String(rowVersion));
			}
		}
	}
	catch (SqlException ex)
	{
		logger.LogError(ex, "Retrieving list of StockGroups");

		return this.StatusCode(StatusCodes.Status500InternalServerError);
	}

	return this.Ok(response);
}

I used Telerik Fiddler to to capture the GET response payload.

HTTP/1.1 200 OK
Transfer-Encoding: chunked
Content-Type: application/json; charset=utf-8
ETag: AAAAAAABrdE=
Server: Microsoft-IIS/10.0
X-Powered-By: ASP.NET
Date: Sat, 26 Jun 2021 06:12:16 GMT

136
[
   {"id":5,"name":"Airline Novelties"},
   {"id":2,"name":"Clothing"},
   {"id":6,"name":"Computing Novelties"},
   {"id":8,"name":"Furry Footwear"},
   {"id":3,"name":"Mugs"},
   {"id":1,"name":"Novelty Items"},
   {"id":10,"name":"Packaging Material"},
   {"id":9,"name":"Toys"},
   {"id":4,"name":"T-Shirts"},
   {"id":7,"name":"USB Novelties"}
]
0

The HEAD method requests the maximum rwoversion value from the StockGroups table and compares it to the eTag. In a more complex scenario this could be a call to a local cache to see if a query result has bee refreshed.

[HttpHead]
public async Task<ActionResult> Head([Required][FromHeader(Name = "ETag")][MinLength(ETagBase64Length, ErrorMessage = "eTag length invalid too short")][MaxLength(ETagBase64Length, ErrorMessage = "eTag length {0} invalid too long")] string eTag)
{
	byte[] headerVersion = new byte[ETagBytesLength];

	if (!Convert.TryFromBase64String(eTag, headerVersion, out _))
	{
		logger.LogInformation("eTag invalid format");

		return this.BadRequest("eTag invalid format");
	}

	try
	{
		using (SqlConnection db = new SqlConnection(this.connectionString))
		{
			byte[] databaseVersion = await db.ExecuteScalarAsync<byte[]>(sql: "SELECT MAX(Version) FROM [Warehouse].[StockGroups]", commandType: CommandType.Text);

			if (headerVersion.SequenceEqual(databaseVersion))
			{
				return this.StatusCode(StatusCodes.Status304NotModified);
			}
		}
	}
	catch (SqlException ex)
	{
		logger.LogError(ex, "Retrieving StockItem list");

		return this.StatusCode(StatusCodes.Status500InternalServerError);
	}

	return this.Ok();
}

I used Fiddler to to capture a HEAD response payload a 304 Not modified.

HTTP/1.1 304 Not Modified
Server: Microsoft-IIS/10.0
X-Powered-By: ASP.NET
Date: Sat, 26 Jun 2021 22:09:02 GMT

I then modified the database and the response changed to 200 OK indicating the local cache should be updated with a GET.

HTTP/1.1 200 OK
Transfer-Encoding: chunked
Server: Microsoft-IIS/10.0
X-Powered-By: ASP.NET
Date: Sat, 26 Jun 2021 22:09:59 GMT

This approach combined with the use of the If-Match, If-Modified-Since, If-None-Match and If-Unmodified-since allows web and client side caches to use previously requested results when there have been no changes. This can significantly reduce the amount of network traffic and server requests.

As part of my testing I modified the eTag so it was invalid (to check the Convert.ToBase64String and Convert.TryFromBase64String error handling) and the response was much smaller than I expected.

HTTP/1.1 400 Bad Request
Content-Length: 240
Content-Type: application/problem+json; charset=utf-8
Server: Microsoft-IIS/10.0
X-Powered-By: ASP.NET
Date: Sat, 26 Jun 2021 06:28:11 GMT

This was unlike the helpful validation messages returned by the GET method of the StockItems pagination example code

{
   "type":"https://tools.ietf.org/html/rfc7231#section-6.5.1",
   "title":"One or more validation errors occurred.",
   "status":400,
   "traceId":"00-bd68c94bf05f5c4ca8752011d6a60533-48e966211dec4847-00",
   "errors": 
   {
      "PageSize":["PageSize must be present and greater than 0"],
      "PageNumber":["PageNumber must be present and greater than 0"]
   }
}

The lack of diagnostic information was not helpful and I’ll explore this further in a future post. I often work on Fintech applications which are “insert only”, or nothing is deleted just marked as inactive/readonly so this approach is viable.

NOTE : Error Handling approach has been updated

devMobile's blog

Random wanderings through Microsoft Azure esp. PaaS plumbing, the IoT bits, AI on Micro controllers, AI on Edge Devices, .NET nanoFramework, .NET Core on *nix and ML.NET+ONNX

Tag Archives: C#

.NET Core SX127X library Part1

TransferFullDuplex vs. Read Write

Summary

.NET Core 5 Raspberry PI SPI

Summary

.NET Core 5 Raspberry PI GPIO Interrupts

.NET Core 5 Raspberry PI GPIO

Seeed LoRa-E5 LowPower problem fix

Seeed LoRa-E5 LowPower problem

.NET Core web API + Dapper – Image Upload

Stream of Bytes or Base64 Encoded

.NET Core web API + Dapper – Image Download

Stream of Bytes and Base64 Encoded

.NET Core web API + Dapper – Polly Retries

Recovering from transient failures with Polly

.NET Core web API + Dapper – Web Caching

Web cache validation with eTags