Beefy Boxes and Bandwidth Generously Provided by pair Networks
The stupid question is the question not asked

Re: Trying to read from serial port - not getting a response

by graff (Chancellor)
on Dec 05, 2012 at 02:20 UTC ( #1007180=note: print w/replies, xml ) Need Help??

in reply to Trying to read from serial port - not getting a response

Do you have any method at all (apart from a perl script) that can let you know that stuff really does come in from the serial port? It would be a comfort, at least, to rule out hardware problems (e.g. having a badly wired RS-232 connector).

Apart from that, you might need to be more explicit about the file handle discipline being used for reading from the port. You seem to be using a default read protocol, whereby a line-termination pattern (probably "\r\n" in your case) needs to be detected by perl in order to return a value from the file handle via the <PORT> read operation.

How about trying a read or sysread instead? That is, something that will read some quantity of bytes, regardless of whether there's any specific "record delimiter" character to be looked for. update:It wouldn't be a bad idea to use one of these functions to read just one byte at a time; that way, it's unlikely that anything will be missed by virtue of being "cached" by the file-handle object.

Replies are listed 'Best First'.
Re^2: Trying to read from serial port - not getting a response
by adiuva (Sexton) on Dec 05, 2012 at 10:23 UTC
    Well, if i connect to the modem directly with Putty for example and execute the same command I'm getting the expected feedback. I also tried different modems with the same result. I think for some reason the response is not ending up the file handle. If you look at the first code I posted you see that the loop to output the response should only be entered if there is anything in the filehandle. But it never entered it. In the second code it's just getting stuck as there is nothing in the filehandle as well. Could you maybe explain more detailed (or provide links) what you mean with "more explicit file handle discipline"? I'm not a very experienced programmer so I would need more input to understand :) I tried both now - read and sysread. But both loops are not beeing entered at all:
    my $read; my $sysread; #Reading the response print "reading...\n"; while ( read(PORT,$read,2048) ) { print "using read:\n"; print $read; } while (sysread(PORT, $sysread, 10000)) { print "using sysread:\n"; print $sysread; }
    Here's what's happening:
    c:\Dropbox\Scripts>perl Enter AT command:ATI0 Sending: 'ATI0' reading... done! c:\Dropbox\Scripts>
      When you say "both loops are not being entered at all", I would guess that it's because you have asked the "read" or "sysread" to return at the point where the serial-port device has sent 2048 or 10000 bytes worth of content, and maybe the device isn't sending that much data. (How much data are you expecting to get back after you send a given "AT" command to the device?)

      Have you tried it this way?

      my $buffer; while (1) { read( PORT, $byte, 1 ); $buffer .= $byte; if ( length( $buffer )) { printf( "Buffer now has %d bytes; last byte read was 0x%02x\n" +, length( $buffer ), $byte ); } }
      You can kill that process (i.e. ^C or some other method) when you feel that it has run long enough so you know what's going on. Then you can add logic inside the while loop so that when the input buffer attains some particular state (or the last byte was some particular value), you break out of the loop and do something else.

      The term "discipline" applied to reading from a file handle covers things like: (a) how to delimit input records (i.e. either by number of bytes, or by watching for a particular byte value or byte sequence; and (b) whether or not your process should wait for data when it tries to read (i.e. using a "blocking" vs. "non-blocking" read - "blocking" is the default).

      It's been a while since I've tried to play with non-blocking reads, but you should be able to look that up. The basic idea is that when "read" is called, it always returns pretty much right away; if the device has data to be read, your input buffer gets the data, otherwise your input buffer remains empty and your process continues to execute. This is how people handle devices where input data arrives at unpredictable intervals, but other things have to be done while waiting for input.

      The example above assumes blocking input (read doesn't return until the device has delivered a byte to be read). If you were to convert it to use a non-blocking input, you'd want to set $byte to empty string or undef before the read call, then check if it contains a value after the read; if it does, then append to the buffer and print the report. (When using non-blocking input, it can also a good idea in many cases to include a "sleep()" call, to keep the loop from consuming too many cpu cycles.)

        I just tried the code you posted. But I'm getting the same result like with my code. The loop is somehow not entered as well:
        c:\Dropbox\Scripts>perl Enter AT command:ATI Sending: 'ATI' reading...
        From my understanding it should at least print the buffer size one time if there is 1 byte of data. The AT-command (ATI) I'm trying should give me something like this:
        ATI Manufacturer: Company Name Model: 1234 Revision: AABBCCDDEE-123456 1 [Dec 01 2012 14:00:00] IMEI: 123456789012345 +GCAP: +CGSM,+DS,+ES OK
      I found something else that might be useful regarding read() and sysread():

Log In?

What's my password?
Create A New User
Node Status?
node history
Node Type: note [id://1007180]
and all is quiet...

How do I use this? | Other CB clients
Other Users?
Others chilling in the Monastery: (6)
As of 2018-01-16 10:30 GMT
Find Nodes?
    Voting Booth?
    How did you see in the new year?

    Results (177 votes). Check out past polls.