when sending a large message after a short one, the pointer of the FrameBuffer would be stuck to the end of the short message, with all the parts of the large message overwriting each other.
Basically I was testing to send some megabytes of data. In server_socketserver.py:74 8192 bytes are read at a time. The FrameBuffer then tries to extract a frame, fails, but doesn't reset the _buffers position to the end, so that the next incoming data is appended, instead of overwriting the first 8192 bytes of the message (or parts thereof).
I just discovered coilmq today, cool project, thx for it. I am not familiar with it, so the above is just a first guess, just m2c.
Coverage decreased (-0.04%) to 91.818% when pulling b1afb8133a35e44abfb73646a5b5457dd666475d on jhb:master into 76b7fcf347144b3a5746423a228bed121dc564b5 on hozn:master.
Basically I was testing to send some megabytes of data. In server_socketserver.py:74 8192 bytes are read at a time. The FrameBuffer then tries to extract a frame, fails, but doesn't reset the _buffers position to the end, so that the next incoming data is appended, instead of overwriting the first 8192 bytes of the message (or parts thereof).
I just discovered coilmq today, cool project, thx for it. I am not familiar with it, so the above is just a first guess, just m2c.