On Aug 19, 6:38 pm, aspineux <aspin...@gmail.com> wrote: > On Aug 19, 5:00 pm, lzlu123 <lzlu...@gmail.com> wrote: > > > > > > > > > > > I have an instrument that has a RS232 type serial comm port and I need > > to connect to and control. I use Python 3.2 in Windows XP, plus > > pySerial module. I have a problem when I execute a script consisting > > of statements that open the comm port, configure it, write strings to > > and receive strings from it. Thoese strings aer either commands > > pertinent to the instrument (control) or responses from the instrument > > (response). > > > When those statements are executed in a python interpreter > > interactively (at >>>), I get what I expect and the results are good > > and correct. However, when I execute the script, either being invoked > > within the interpreter or run file, I don’t get what I want. The > > statements in the script is the same as what I use in the interactive > > interpreter. > > > Why do I get the strange behavior and how can I change the script to > > make it to behave like in interactive interpreter? > > > ----------------------script----------------------- > > def read(comport): > > > wrt_str=b'movt 3000'+b'\r\n' > > ret_str=comport.write(wrt_str) > > > wrt_str=b'scan'+b'\r\n' > > ret_str=comport.write(wrt_str) > > > rsp_str=comport.readlines() #########1 > > You use readlines() with a s at the end ! > > * Note that when the serial port was opened with no timeout, that > readline() > * blocks until it sees a newline (or the specified size is reached) > * and that readlines() would never return and therefore refuses to > work > * (it raises an exception in this case)! > > > > > > > > > > > > > wrt_str=b'hllo'+b'\r\n' > > ret_str=comport.write(wrt_str) > > > rsp_str=comport.readlines()#########2 > > ---------------------------------------------------------- > > > The problem is with the lines above with #######. In interactive mode, > > there is about 1 second delay at #1, and 9 seonds delay at #2. I get > > correct responses there. However, if I execute the script above, there > > is no delay at all and I get incorrect results (garbage). I set the > > read timeout to 0 in comm port set up, as > > > comport.timeout=0 > > So the comport should be in blocking mode if it waits for the end of > > line or end of file. > > Wrong : > > timeout = None: wait forever > timeout = 0: non-blocking mode (return immediately on read) > timeout = x: set timeout to x seconds (float allowed) > > > > > > > > > > > > > I tried many things, like exec (execfile in 2.7), but at no avail. > > > I have an update to the original post I made a few days ago. I think I > > know what the problem is and want to know if anyone has a solution: > > > After putting "print" and "time.sleep(delay)" after every statement, I > > found when the script is running, it appears going around the pyserial > > statement, such as "comport.write(..)" or "comport.readlines(...)" > > while the pyserial command is executing (appearing as waiting and > > busying doing some thing, you know serial port is slow). So for > > example, when I exec all statements in a python interactive shell, I > > am not able to type and run a new statement if the previous one is not > > returned. Let's ay, if comport.readlines() is not returning, I can't > > type and run the next comport.write(...) statemtn. However, in a > > script that is running, if the comport.readlines() is busy reading, > > the next statement is running, if the next statement happens to be a > > comport.write() which will abort the reading. > > > Is there any way to force the python script to behave like running > > exactly sequentially? > > You have some new things to try
I tried to set the timeout to different values other than 0, but still got same result -- ruuning script directly behaves differently from executing statements in interactive shell. I even tried using read instead of readline(s). -- http://mail.python.org/mailman/listinfo/python-list
