Bob,
Really good question. I know there are a number of different
definitions of 'BNF', but was expecting that we could pick one as a
reference, preferably one that has good software tools to go with it.
I'm hoping to avoid defining our own interpretation of BNF, as the task
of 'defining a language' is how we got here in the first place ...
... and that's the point where my computer science fails me. It does
like we need a syntax that includes both OR and XOR, as I expect that
we'll need to describe both.
Walter & I can do some research on BNF alternatives & tools, but not
until some time next week.
Do you agree we should try to find a BNF reference can all work from?
Todd.
Todd Westerhoff
VP, Software Products
SiSoft
6 Clock Tower Place, Suite 250
Maynard, MA 01754
(978) 461-0449 x24
twesterh@xxxxxxxxxx
www.sisoft.com
On 2/24/2010 11:54 AM, Bob Ross wrote:
Todd and Walter:
Thanks, these formal descriptions are helpful starts and also provide
a cross-check and basis for detailed questions.
I am not aware of any BNF standard since I have seen many different
styles.
I assume that "|" is exclusive OR since 'True False' is a violation.
GetWave_Exists (Value True|False)|(Default True|False)
Is Default AND Value a violation?
Bob
Todd Westerhoff wrote:
Walter,
I think you're on to something here.
The main issue I keep hearing is that we have trouble spec'ing the
.AMI file syntax for parser development, whether it be the IBIS
parser or EDA vendor tools. Having a formal definition of the .AMI
syntax in BNF is a great idea and, as far as I'm concerned, directly
addresses many of the problems we've had.
I took a shot at cleaning up the syntax you'd defined - fixed a few
of the typos and added whitespace for readability. I know the
computers don't care, but I think whitespace can make it easier for
us humans to review and discuss.
I think the reserved parameter stuff is reasonably clean, but I
suspect the model-specific parameter definitions still need work I'm
familiar with BNF in concept but am most certainly not an expert. If
others agree this is the way to go, let's enlist the help of someone
with more BNF experience and get this sorted out.
Todd.
Todd Westerhoff
VP, Software Products
SiSoft
6 Clock Tower Place, Suite 250
Maynard, MA 01754
(978) 461-0449 x24
twesterh@xxxxxxxxxx
www.sisoft.com
On 2/24/2010 8:35 AM, Walter Katz wrote:
All,
The parameter tree was initially used for the content of the strings
in Parameters_In and Parameters_Out. The modified BNF in the IBIS
5.0 AMI specification is an accurate representation and therefore an
accurate method to parse the strings in Parameters_In and
Parameters_Out.
The .ami file is also in a parameter tree format, but has additional
rules. I do not know how to capture these additional rules in a
formal BNF that can be used to programmatically parse a .ami file. I
did put the following document together than may very well be
misnames a BNF, but I do think graphically documents the parser
rules for the update .ami syntax.
This follows up on the ideas that Bob suggested in his e-mail to the
group on AMI Quick Reference.
I think it is useful as is, any suggestions on how to make it more
readable, or BNF like would be much appreciated.
Walter
Walter Katz
303.449-2308
Mobile 720.333-1107
wkatz@xxxxxxxxxx
www.sisoft.com
---------------------------------------------------------------------
IBIS Macro website : http://www.eda.org/pub/ibis/macromodel_wip/
IBIS Macro reflector: //www.freelists.org/list/ibis-macro
To unsubscribe send an email:
To: ibis-macro-request@xxxxxxxxxxxxx
Subject: unsubscribe
