heads-project / arduino_msgpack Goto Github PK

since the Print class has a virtual method write(data, len), and some sub class like HardwareSerial has optimized block write method, we should use it instead of writing bytes one by one
same for write_string

See https://github.com/HEADS-project/arduino_msgpack/blob/master/examples/test_uno_writer/test_uno_writer.ino#L37 it gives compile error.
This line should be:
msgpck_write_float(&Serial, 0.5f);

Hi and thanx for your work

I am looking to encode/decode an array that would look like this
PHP arrays
$data = [ '1234567',
['9999999', '7654321', '64'],
['62', ['64513451334' => '' ],
[
'pin1'=>['5,2,4'],
'ipin2'=>[5,2,'readIC'=>3],
4=>[2=>80]
]
]
];

I looked at the examples but I can not understand how can I use it for such arrays which are more real life and would like to pack everything and send it in 1 package rather than splitting it in multiple parts like the examples show.

Thank you.

https://github.com/HEADS-project/arduino_msgpack/blob/master/msgpck.cpp#L327~L331 currently you just treat 8-bytes float as error and discard it.
can you read double the same way you read float, by defining a 8-bytes union?

union double_to_byte {
      double;
      byte b[8];
    } b2d;

also, these 4 lines looks like typo https://github.com/HEADS-project/arduino_msgpack/blob/master/msgpck.cpp#L321~L324, should they be

b &= ...

Hi,

I believe there is an error in the implementation of msgpck_write_integer(Stream * s, int16_t i).

If i > 127 and i <= 254, the else-part is executed. But then the number is written as an int8_t, so it will be interpreted as a negative number on the other side.

It seems that there is a case-distiction missing in the else-part: If i > 127, it should be written as uint8_t, otherwise as int8_t.

Thanks