Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > comp.lang.python > #47287 > unrolled thread

trigger at TDM/2 only

Back to article view | Back to comp.lang.python

Contents

  trigger at TDM/2 only cerr <ron.eggler@gmail.com> - 2013-06-06 17:03 -0700
    Re: trigger at TDM/2 only MRAB <python@mrabarnett.plus.com> - 2013-06-07 01:49 +0100
      Re: trigger at TDM/2 only cerr <ron.eggler@gmail.com> - 2013-06-07 09:38 -0700
    Re: trigger at TDM/2 only Dave Angel <davea@davea.name> - 2013-06-06 20:43 -0400
      Re: trigger at TDM/2 only cerr <ron.eggler@gmail.com> - 2013-06-07 09:35 -0700

#47287 — trigger at TDM/2 only

Hi,

I have a process that I can trigger only at a certain time. Assume I have a TDM period of 10min, that means, I can only fire my trigger at the 5th minute of every 10min cycle i.e. at XX:05, XX:15, XX:25... For hat I came up with following algorithm which oly leaves the waiting while loop if minute % TDM/2 is 0 but not if minute % TDM is 0:
	min = datetime.datetime.now().timetuple().tm_hour*60 + datetime.datetime.now().timetuple().tm_min
	while not (min%tdm_timeslot != 0 ^ min%(int(tdm_timeslot/2)) != 0):
		time.sleep(10)
		logger.debug("WAIT "+str(datetime.datetime.now().timetuple().tm_hour*60 + datetime.datetime.now().timetuple().tm_min))
		logger.debug(str(min%(int(tdm_timeslot/2)))+" - "+str(min%tdm_timeslot))
		min = datetime.datetime.now().timetuple().tm_hour*60 + datetime.datetime.now().timetuple().tm_min
	logger.debug("RUN UPDATE CHECK...")

But weird enough, the output I get is something like this:
I would expect my while to exit the loop as soon as the minute turns 1435... why is it staying in? What am I doing wrong here?

WAIT 1434
3 - 3
WAIT 1434
4 - 4
WAIT 1434
4 - 4
WAIT 1434
4 - 4
WAIT 1434
4 - 4
WAIT 1434
4 - 4
WAIT 1435
4 - 4
WAIT 1435
0 - 5
WAIT 1435
0 - 5
WAIT 1435
0 - 5
WAIT 1435
0 - 5
WAIT 1435
0 - 5
WAIT 1436
0 - 5
RUN UPDATE CHECK...


Thank you for any assistance!
Ron

[toc] | [next] | [standalone]

#47288

On 07/06/2013 01:03, cerr wrote:
> Hi,
>
> I have a process that I can trigger only at a certain time. Assume I have a TDM period of 10min, that means, I can only fire my trigger at the 5th minute of every 10min cycle i.e. at XX:05, XX:15, XX:25... For hat I came up with following algorithm which oly leaves the waiting while loop if minute % TDM/2 is 0 but not if minute % TDM is 0:
> 	min = datetime.datetime.now().timetuple().tm_hour*60 + datetime.datetime.now().timetuple().tm_min
> 	while not (min%tdm_timeslot != 0 ^ min%(int(tdm_timeslot/2)) != 0):
> 		time.sleep(10)
> 		logger.debug("WAIT "+str(datetime.datetime.now().timetuple().tm_hour*60 + datetime.datetime.now().timetuple().tm_min))
> 		logger.debug(str(min%(int(tdm_timeslot/2)))+" - "+str(min%tdm_timeslot))
> 		min = datetime.datetime.now().timetuple().tm_hour*60 + datetime.datetime.now().timetuple().tm_min
> 	logger.debug("RUN UPDATE CHECK...")
>
> But weird enough, the output I get is something like this:
> I would expect my while to exit the loop as soon as the minute turns 1435... why is it staying in? What am I doing wrong here?
>
> WAIT 1434
> 3 - 3
> WAIT 1434
> 4 - 4
> WAIT 1434
> 4 - 4
> WAIT 1434
> 4 - 4
> WAIT 1434
> 4 - 4
> WAIT 1434
> 4 - 4
> WAIT 1435
> 4 - 4
> WAIT 1435
> 0 - 5
> WAIT 1435
> 0 - 5
> WAIT 1435
> 0 - 5
> WAIT 1435
> 0 - 5
> WAIT 1435
> 0 - 5
> WAIT 1436
> 0 - 5
> RUN UPDATE CHECK...
>
Possibly it's due to operator precedence. The bitwise operators &, |
and ^ have a higher precedence than comparisons such as !=.

A better condition might be:

     min % tdm_timeslot != tdm_timeslot // 2

or, better yet, work out how long before the next trigger time and then
sleep until then.

[toc] | [prev] | [next] | [standalone]

#47339

MRAB,

Thanks for the hint! Yep, that's much easier!

Thanks! :)
Ron

On Thursday, June 6, 2013 5:49:55 PM UTC-7, MRAB wrote:
> On 07/06/2013 01:03, cerr wrote:
> 
> > Hi,
> 
> >
> 
> > I have a process that I can trigger only at a certain time. Assume I have a TDM period of 10min, that means, I can only fire my trigger at the 5th minute of every 10min cycle i.e. at XX:05, XX:15, XX:25... For hat I came up with following algorithm which oly leaves the waiting while loop if minute % TDM/2 is 0 but not if minute % TDM is 0:
> 
> > 	min = datetime.datetime.now().timetuple().tm_hour*60 + datetime.datetime.now().timetuple().tm_min
> 
> > 	while not (min%tdm_timeslot != 0 ^ min%(int(tdm_timeslot/2)) != 0):
> 
> > 		time.sleep(10)
> 
> > 		logger.debug("WAIT "+str(datetime.datetime.now().timetuple().tm_hour*60 + datetime.datetime.now().timetuple().tm_min))
> 
> > 		logger.debug(str(min%(int(tdm_timeslot/2)))+" - "+str(min%tdm_timeslot))
> 
> > 		min = datetime.datetime.now().timetuple().tm_hour*60 + datetime.datetime.now().timetuple().tm_min
> 
> > 	logger.debug("RUN UPDATE CHECK...")
> 
> >
> 
> > But weird enough, the output I get is something like this:
> 
> > I would expect my while to exit the loop as soon as the minute turns 1435... why is it staying in? What am I doing wrong here?
> 
> >
> 
> > WAIT 1434
> 
> > 3 - 3
> 
> > WAIT 1434
> 
> > 4 - 4
> 
> > WAIT 1434
> 
> > 4 - 4
> 
> > WAIT 1434
> 
> > 4 - 4
> 
> > WAIT 1434
> 
> > 4 - 4
> 
> > WAIT 1434
> 
> > 4 - 4
> 
> > WAIT 1435
> 
> > 4 - 4
> 
> > WAIT 1435
> 
> > 0 - 5
> 
> > WAIT 1435
> 
> > 0 - 5
> 
> > WAIT 1435
> 
> > 0 - 5
> 
> > WAIT 1435
> 
> > 0 - 5
> 
> > WAIT 1435
> 
> > 0 - 5
> 
> > WAIT 1436
> 
> > 0 - 5
> 
> > RUN UPDATE CHECK...
> 
> >
> 
> Possibly it's due to operator precedence. The bitwise operators &, |
> 
> and ^ have a higher precedence than comparisons such as !=.
> 
> 
> 
> A better condition might be:
> 
> 
> 
>      min % tdm_timeslot != tdm_timeslot // 2
> 
> 
> 
> or, better yet, work out how long before the next trigger time and then
> 
> sleep until then.

[toc] | [prev] | [next] | [standalone]

#47289

On 06/06/2013 08:03 PM, cerr wrote:
> Hi,
>
> I have a process that I can trigger only at a certain time. Assume I have a TDM period of 10min, that means, I can only fire my trigger at the 5th minute of every 10min cycle i.e. at XX:05, XX:15, XX:25... For hat I came up with following algorithm which oly leaves the waiting while loop if minute % TDM/2 is 0 but not if minute % TDM is 0:
> 	min = datetime.datetime.now().timetuple().tm_hour*60 + datetime.datetime.now().timetuple().tm_min
> 	while not (min%tdm_timeslot != 0 ^ min%(int(tdm_timeslot/2)) != 0):

You might have spent three minutes and simplified this for us.  And in 
the process discovered the problem.

(BTW, min() is a builtin function, so it's not really a good idea to be 
shadowing it.)

You didn't give python version, so my sample is assuming Python 2.7 
For your code it shouldn't matter.

tdm = 10
tdm2 = 5

y = min(3,4)
print y

for now in range(10,32):
     print now, now%tdm, now%tdm2,
     print not(now % tdm !=0 ^ now%tdm2 !=0) #bad
     print not((now % tdm !=0) ^ (now%tdm2 !=0))  #good


Your problem is one of operator precedence.  Notice that ^ has a higher 
precedence than != operator, so you need the parentheses I added in the 
following line.

What I don't understand is why you used this convoluted approach.  Why not

     print now%tdm != tdm2

For precedence rules, see:
   http://docs.python.org/2/reference/expressions.html#operator-precedence




-- 
DaveA

[toc] | [prev] | [next] | [standalone]

#47336

DaveA,

Yep, that seems to just be about it! Much easier!

Thanks for the hint! Much appreciated!!!! :)

Ron

On Thursday, June 6, 2013 5:43:11 PM UTC-7, Dave Angel wrote:
> On 06/06/2013 08:03 PM, cerr wrote:
> 
> > Hi,
> 
> >
> 
> > I have a process that I can trigger only at a certain time. Assume I have a TDM period of 10min, that means, I can only fire my trigger at the 5th minute of every 10min cycle i.e. at XX:05, XX:15, XX:25... For hat I came up with following algorithm which oly leaves the waiting while loop if minute % TDM/2 is 0 but not if minute % TDM is 0:
> 
> > 	min = datetime.datetime.now().timetuple().tm_hour*60 + datetime.datetime.now().timetuple().tm_min
> 
> > 	while not (min%tdm_timeslot != 0 ^ min%(int(tdm_timeslot/2)) != 0):
> 
> 
> 
> You might have spent three minutes and simplified this for us.  And in 
> 
> the process discovered the problem.
> 
> 
> 
> (BTW, min() is a builtin function, so it's not really a good idea to be 
> 
> shadowing it.)
> 
> 
> 
> You didn't give python version, so my sample is assuming Python 2.7 
> 
> For your code it shouldn't matter.
> 
> 
> 
> tdm = 10
> 
> tdm2 = 5
> 
> 
> 
> y = min(3,4)
> 
> print y
> 
> 
> 
> for now in range(10,32):
> 
>      print now, now%tdm, now%tdm2,
> 
>      print not(now % tdm !=0 ^ now%tdm2 !=0) #bad
> 
>      print not((now % tdm !=0) ^ (now%tdm2 !=0))  #good
> 
> 
> 
> 
> 
> Your problem is one of operator precedence.  Notice that ^ has a higher 
> 
> precedence than != operator, so you need the parentheses I added in the 
> 
> following line.
> 
> 
> 
> What I don't understand is why you used this convoluted approach.  Why not
> 
> 
> 
>      print now%tdm != tdm2
> 
> 
> 
> For precedence rules, see:
> 
>    http://docs.python.org/2/reference/expressions.html#operator-precedence
> 
> 
> 
> 
> 
> 
> 
> 
> 
> -- 
> 
> DaveA

[toc] | [prev] | [standalone]

Back to top | Article view | comp.lang.python

csiph-web