I think that there has been some misunderstanding with regard to the
original question. I believe Dillon is trying to find the start bit and
number of bits in the multiplexer. The terminology of multiplexer and
mode gets a little confusing. For clarification, my understanding is
that a mode is a specific value of a multiplexer. In the case of Dirk's
example code, he was able to retrieve the number
of bits of the CHANNEL, but not the number of bits of the MULTIPLEXER.
I believe that I have run into a similar problem to
Dillon's. I want to read a NI CAN Database (*.ncd) or a Vector CAN
Database (*.dbc) programmitically. I have attached a zip containing a
VI and some *.ncd file to allow others to see my progress thus far.
This is a heavily modified version of Dirk's previously posted program.
As you can see, it is easy to get a list of messages using the "CAN Get
Names" VI. Using that list of messages, it is also easy to get a list
of channels by reusing the "CAN Get Names" VI. The list of channels
does not include any information as to whether or not they belong to a
multiplexer, but fortunately we can use the "CAN Get Property" VI to
find out which channels are mode dependent and the value of the mode
that the channel corresponds to (if it is mode dependent).
Unfortunately, this does not give us any information as to the location
or length of the multiplexer which contains the mode. Thus, I know
whether I should be looking for a multiplexer and I know what value to
watch for in the multiplexer, but I have no idea where or how to look
for that value in the CAN message.
Interestingly, by
playing around with Dirk's database that he had attached, I was able to
determine that the list of channels does not always give out the same
information. Typically, you will simply receive an 1D array strings
containing only the channel names, but the output is different in the
event that you have a database that meets the following criteria:
1. Two channels (in different modes) have the same name
2. The database contains at least two multiplexers
In
this case the output will be a 1D array of strings in which the
repeated channels (regardless of whether they are in the same
multiplexer) will be output in the format "<message
name>.<multiplexer NAME>.<mode value>.<channel
name>". The channels that have unique names are simply output as the
channel name. Now this is helpful in that by default the multiplexer
name contains the start bit of the multiplexer, but this cannot be
relied upon. The creator of the database could easily change the name
of the multiplexer and the format is only adhered to when the
previously mentioned criteria are met. Also it still does not tell us
anything about the number of bits expected in the multiplexer (remember
Dirk had found the number of bits expected for the CHANNEL).
I have found another alternative output format when the following criteria are met:
1. Two channels (in different modes) have the same name
2. Only one multiplexer is contained within the file
In
this case the output will be a 1D array of strings in which the
repeated channels will be output in the format "<message
name>.<mode value>.<channel
name>". Again, this special format is only used for the repeated
channels. Uniquely named channels are simply output as channel names. While this is not particularly useful, but I thought I should mention it.
It
is also probably worth noting that I found the previously mentioned
alternative formats by directly reading reading a *.ncd file. It
appears that by reading the channels directly from MAX we get at least
one additional format (<message name.channel name> ). Considering
that this is actually even less information and since my application
specifically requires that I directly read a CAN database without the
assistance of MAX, I have not significantly explored this avenue.
Let
me finish by reiterating the question that I am asking: How do you
programmitically determine the start bit and number of bits of the
multiplexers in a CAN database (*.ncd or *.dbc). I have attached a zip
file containg my current "testing" VI and some databases to illustrate
the results that I have described above.
Message Edited by William Griffin on 01-06-2009 04:58 PM
William Griffin
NI Certified LabVIEW Architect
NI Certified Professional Instructor
Spark Embedded
https://sparkembedded.com/
