Too few input values to make output type

1 次查看（过去 30 天）

Jason Gauthier 2018-5-31

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/403447-too-few-input-values-to-make-output-type

评论： Walter Roberson 2018-5-31

I'm reading a binary stream from a serial port. The binary stream looks like this:

start byte (actually byte series)
length of subsequent data
important data
checksum

The stream is coming from a QUARC block, 16 bytes at a time. So, I'm put this into a MATLAB function that uses persistent variables to process the stream, detect the sync sequence, grab the length byte, and then read the next x bytes. The last 8 bytes (which is a float) is the checksum.

I want to convert that byte stream into a float, and then I can determine from the important data if it matches.

I've successfully used typecast to do this, but in this case it's simply not working.

I suspect, because the other complication, is that this is being built for code generation.

So, to provide specific details:

'bytes' is read in, and determined to be 48 bytes. the 40-48 bytes is the checksum. But I'm not always guaranteed to read in 48 bytes. I'd like to have some flexibility, so I cannot guarantee the size of the bytes that I am going to read (based on what I am sending, which is why I am sending the length)

So, 'bytes' (variable) now contains the length of the data, including the 8-byte checksum.

so, I try to type case that:

checksum = typecast(uint8(data(bytes-7:bytes)), 'double');

And it seems that it's not going to work when I build it.

    "Too few input values to make output type.
    Function 'Subsystem/MATLAB Function3' (#59.1056.1084), line 56, column 19:
    "typecast(uint8(d), 'double')"
    Launch diagnostic report."

If I hard code in some values:

checksum = typecast(uint8(data(32-7:32)), 'double');

The application builds fine.

I've been banging my head at reading this byte stream for far too long, and I finally made some progress and now hit another wall. Any thoughts are welcome.