Differences

This shows you the differences between two versions of the page.

--- en:edu:zmitac:s2:aih:lab:nrf52zephyrbleperipheral [2024/01/13 14:19] – [Adding the possibility of reading the characteristic] ktokarz
+++ en:edu:zmitac:s2:aih:lab:nrf52zephyrbleperipheral [2024/01/13 15:23] (current) – [Adding the possibility of reading the characteristic] ktokarz
@@ Line 102: / Line 102: @@
   * write the callback function that calls bt_gatt_attr_read() (example function in the following code)
   * register callback function within the BT_GATT_CHARACTERISTIC macro using the "_read" parameter.
-An example of the callback function which simulates the thermometer value can look like this:
+Notice that the temperature data sent in the original example by indications is encoded in the following format, which can be automatically decoded by nRF Connect software.
-<code>
-static ssize_t read_ht(struct bt_conn *conn,
-		       const struct bt_gatt_attr *attr, void *buf,
-		       uint16_t len, uint16_t offset)
-{
-	uint8_t lvl8 = 36;
-	return bt_gatt_attr_read(conn, attr, buf, len, offset, &lvl8, sizeof(lvl8));
-}
-</code>
-Notice that the temperature data which is sent in the original example by indications is encoded in the following format:
 <code>
 bool 	        temp_in_fahr_units
@@ Line 122: / Line 112: @@
 uint8_t 	temp_type
 </code>
-All boolean flags are set to 0 (the first byte which is sent is 0x00), so the only data is the temperature expressed as a 32-bit floating point number encoded as an 8-bit exponent and 24-bit mantissa. This format is not compatible with single or double-precision formats used in microprocessors, so conversion is required.
+All boolean flags are set to 0 (the first byte which is sent is 0x00), so the only data is the temperature expressed as a 32-bit floating point number encoded as an 8-bit exponent and 24-bit mantissa (ieee_float32_t).
+<note>
+In the ieee_float32_t format, the exponent is encoded as a signed 8-bit binary 10-based number. The mantissa is encoded as a signed 24-bit binary number. This format is incompatible with single or double-precision formats used in microprocessors, so conversion is required.
+</note>
+An example of the callback function which simulates the thermometer value can look like this:
+<code>
+static ssize_t read_ht(struct bt_conn *conn,
+		       const struct bt_gatt_attr *attr, void *buf,
+		       uint16_t len, uint16_t offset)
+{
+	uint8_t temp_buffer[5];
+	double temp_read;
+	uint32_t mantissa;
+	uint8_t exponent;
+	temp_read = 36.6;	//constant temperature for reading
+	temp_buffer[0] = 0x00;	//flags for temerature data formatting
+        //conversion from double to 4 bytes
+	//mantissa is 24 bits signed binary
+	//exponent is 8 bits signed 10-base binary
+	mantissa = (uint32_t)(temp_read * 100);
+	exponent = (uint8_t)-2;
+	sys_put_le24(mantissa, (uint8_t *)&temp_buffer[1]);
+	temp_buffer[4] = exponent;
+	return bt_gatt_attr_read(conn, attr, buf, len, offset, &temp_buffer, sizeof(temp_buffer));
+}
+</code>
 ===== Tasks to do during the laboratory =====
   - Import the peripheral_ht example.