Welcome to factor 2 of the sockets tutorial with Python. Within the former tutorial, we found how we might ship and acquire particulars utilizing sockets, however then we illustrated the difficulty that may come up when our communication exceeds our buffer measurement. On this tutorial, we’re going to chat about overcoming this!

Useful resource code and text-primarily based mostly tutorial: https://pythonprogramming.net/buffering-streaming-data-sockets-tutorial-python-3/

Channel membership: https://www.youtube.com/channel/UCfzlCWGWYyIQ0aLC5w48gBQ/sign up for
Discord: https://discord.gg/sentdex
Help the knowledge: https://pythonprogramming.web/assistance-donate/
Twitter: https://twitter.com/sentdex
Fb: https://www.fb.com/pythonprogramming.internet/
Twitch: https://www.twitch.tv set/sentdex
G+: https://in addition.google.com/+sentdex

#socket #networking #python

supply

45 thoughts on “Sockets Tutorial with Python three half 2 – buffering and streaming info”

  2. Great video. The way how you implemented the method to get a message via header and msg length is not working however if new strings are sent every few milliseconds. On my second machine, it leads the msg to grow infinitely and never enter the == condition. This code fixed it for me if anyone has also use case with low latency applications. I simply receive only the header first because I know its length (in my case 10). Then I fetch exactly the msg length. Even if there is already a queue of strings in the socket waiting to be fetched, this method gets the messages reliably.

    while True:

    full_msg = '''
    msglen = 0
    new_msg = True
    while True:

    if new_msg:
    msg = self.clientsocket.recv(10)
    msglen = int(msg)
    new_msg = False
    else :
    msg = self.clientsocket.recv(msglen)
    #full_msg = msg.decode("utf-8")
    #print(full_msg)
    new_msg = True

    Reply

  9. Kinda bad at python, could anyone pls explain in the client code:
    msg[:HEADERSIZE]
    How does this give the message length? I tried printing a string in python from [ : i] where is just a random number and it just printed the beginning of a string up until i.
    Wouldn't msg[:HEADERSIZE] only give the first 10 characters?

    Reply

  12. can you create more vids about python sockets that is more complicated that you know hehe. u know ur expression great and u presentation too, nice video anyway continue posting python vids.

    Reply

  14. Great videos as always, many thanks! It works perfectly fine if both server and client are on my local machine, but I'm not able to establish a connection between my local machine and a cloud hosted VM. Any idea how to make that work?

    Reply

  17. How would I send data to something other than the machine that the server is running on. For example how would one send data to other computers on the same wifi or even how could I send data to my friend who lives in another city?

    Reply

  20. MY Laptop hangs/freezes after sometimes, after connecting the server to the client and then the whole PC doesn't responses, then i have to directly put it off by forcing it to shutdown by the power button. why is it so??

    Reply

  30. Awesome video! A bit of feedback, I think that it would help people if you explain more about how a socket works. It took me a while to figure out that sockets will close if there is no more data to be sent, which explains why you need buffering.

    Reply

  35. Question:
    What is the purpose of all these nested loops, junk vars, conditions and string truncating? I replaced all 21 lines of your code with:

    msg = s.recv(int(s.recv(HEADERSIZE)))
    print(msg.decode("utf-8"))

    Is there something necessary about your way? By observation, I am getting identical results, but I don't waste time looping baby chunks and juggling a bunch of extra code

    —–

    Also, what's up with:
    msg = f'{len(msg):<{HEADERSIZE}}' + msg

    Why not:
    msg = f'{len(msg):<{HEADERSIZE}}{msg}'

    and actually, why not:
    def message(msg):
    return f'{len(msg):<{HEADERSIZE}}{msg}'.encode('utf-8')

    msg = message("Welcome to the server")
    clientsocket.send(msg)

    This way you have one function that formats all of your messages perfectly without you having to write the same thing repetitively? I know you would reply something like "This is only lesson 2. We will get to that point.", but isn't that why tutorials like these drag on forever? They get built up with a bunch of junk code that you have to replace and then you spend 2 or 3 tutorials undoing all the stuff that was written poorly. It doesn't take 2 loops, 3 vars and a condition to get a number and use it. Why not just write it to the most basic PROPER way the first time and build on that? The point of this tutorial was to create a header system, but you only did that half way (maybe not even that).

    —–

    "You can just embed more brackets inside of brackets"

    EXCEPT if the opening bracket comes directly after another open bracket. "{{" is how you print 1 open bracket in an f-string.

    example:
    a, b = 10, 20
    print(f'{{a} b}')
    >>> '{a} b}'

    It will treat the double open brackets as a single open bracket CHARACTER and fail to do any of what you expect.

    Reply

  37. When getting massage from server i.e. in rcev() function you used ''int (msg[:HEADERSIZE]) " but I am getting error invalid literal for int() with base 10: how you are getting it correct , as when we use int() function on byte encoded stream it gives this error .

    Reply

  38. Is there something wrong with doing it this way? It doesn't close the connection, it also gets the whole message. Probably more effecient since it uses only one loop? I can put comments in code if someone needs it.
    1)server.py
    import socket

    import time

    buffersize = 10

    s = socket.socket( socket.AF_INET, socket.SOCK_STREAM )

    s.bind( (socket.gethostname(), 1234) )

    s.listen(5)

    while True:

    clientsocket, address = s.accept()

    print( 'Connection with: ' + str(address) )

    msg = 'Welcome to the server my friend!'

    msg = f'{len(msg):^{buffersize}}' + msg

    clientsocket.send( bytes(msg, 'utf-8') )

    while True:

    time.sleep(1)

    mes = f'The time is: {time.time()}'

    mes = f'{len(mes):^{buffersize}}' + mes

    clientsocket.send( bytes(mes, 'utf-8') )

    2) client.py
    import socket

    buffer_size = 10

    s = socket.socket(socket.AF_INET, socket.SOCK_STREAM)

    s.connect((socket.gethostname(), 1234))

    full_msg = ''

    new_msg = ''

    x = 0

    while True:

    new_msg = s.recv(buffer_size).decode('utf-8')

    full_msg += new_msg

    if x == 0:

    header = int(new_msg)

    full_msg = ''

    if len(full_msg) == header:

    print(full_msg)

    x = -1

    x += 1

    Reply

  40. For anyone who wants a life pro-tip:

    If you don't understand something done in a video, don't just copy it into the code. Actually skip it, or try to implement the code the way you think it should work. Usually you'll then see an issue, that's being resolved by the code in the video. Or be right, and your code will also work. Both are better than just copying something without understanding what is done. Happy coding!

    Reply

  44. I'm a little surprised this works consistently, with genuine Unicode data. I'd think that if, for whatever reason, a partial block of UTF-8 text were sent, that the the last character would be undecodable, and the next block would start with the remnants of an incomplete character encode. The assembly should be of bytes, and then at the very end, decode the bytes from utf-8.

    Reply

Leave a Reply

Your email address will not be published. Required fields are marked *