pat@grebyn.com (Pat Bahn) (08/13/89)
At long last, My summary of responses on McCabe Complexity and C PDL
Date: Wed, 2 Aug 89 15:48:18 PDT
From: haven!Franz.COM!keb (Kerstin Barley)
Organization: Franz Inc., Berkeley
What are you doing programming in C? Real developers utilize
the power of Lisp...
:-) keb
Date: Thu, 3 Aug 89 21:10:58 CDT
From: haven!Central.Sun.COM!boone (David Boone Sun-Central Area SE)
Subject: Re: C PDL AND MCCABES COMPLEXITY MEASURES
Organization: Sun Microsystems Inc.
McCabe Associaters has exactly the software your
are looking for.
--david
Date: Wed, 2 Aug 89 16:54 GMT
From: David Kelly <haven!CUNYVM.CUNY.EDU!DPMKELLY%cs.tcd.ie>
Subject: reply to metrics query in comp.lang.c
To: pat@grebyn.com
X-Vms-To: IN%"pat@grebyn.com"
RE: C PDL AND MCCABE'S COMPLEXITY MEASURES
Greetings from Dublin ...
A software metrics analyser called SCMT (Software Complexity Metrics
Tool) has been written at Oregon State University by C. Cook and M. Nanja. They
describe it in ACM SIGSoft (July '87). It analyses C, producing McCabe's
cyclometric complexity metric and numerous Halstead's metrics, and was
originally written in C to run in a UNIX environment. The source code is freely
available and has been modified here to run on an IBM PC and (I think) VMS. As
a degree project here I wrote a metrics analyser for C (in LISP) and agreed
with SCMT's methods of calculating v(G) but was rather dubious about some of
the counting rules employed to gain Halstead's metrics. There is another
complexity measure derived from the amount of nesting within a subroutine - if
you think it might be useful, I think I can dig out a reference.
I presume that by PDL you mean *Programme Development Language*. Sorry,
I can't help here. However, my wonderful brilliant super metrics analyser (??)
did allow the analysis of partially written code. For instance, a programmer
could submit a skeleton of code (e.g., conditions just written as comments) and
the analyser would return a preliminary Halstead evaluation of it (length,
volume, estimated no. of bugs). The big idea was that this would allow bad
algorithms/badly written subroutines to be identified at an early point in the
life cycle, thus facilitating rapid prototyping. I have not had a a chance to
try it out yet ...
From: haven!devvax.Jpl.Nasa.Gov!leem (Lee Mellinger)
Organization: Jet Propulsion Laboratory, Pasadena, CA.
Cc:
Macabe and Associates - Tom Macabe of course, has just such a tool
which will take PDL as input and produce complexity, flowgraphs, test
values.
From: Warren Harrison <haven!RELAY.CS.NET!warren%jove.cs.pdx.edu>
Organization: Dept. of Computer Science, Portland State University; Portland OR
Cc:
I don't know of any formal C PDL's, but a couple of outfits have commercial
tools for McCabe's Cyclomatic Complexity. You might want to try McCabe's
company, McCabe & Associates in Columbia Maryland (301-596-3080) or SET
Laboratories here in Portland (503-289-4758).
From: haven!CUNYVM.CUNY.EDU!ssdken%JARTHUR.CLAREMONT.EDU
Subject: Re: C PDL AND MCCABES COMPLEXITY MEASURES
Sender: haven!CUNYVM.CUNY.EDU!ssdken%JARTHUR.CLAREMONT.EDU
Organization: Software Systems Design, Claremont, CA
X-Envelope-To: pat@grebyn.com
My employer (Software Systems Design) offers an existing Ada based PDL,
which can be used with C. We develop in C but use our PDL for design and
documentation. The PDL is called ADADL. ADADL also reports on Quality
Metrics. THe McCabe complexity is one of these reports. Documentation
completeness and the variations between design and the implementation.
ADADL is part of a family of tools:
DOCGEN - automatically generates military standard documentation.
TESTGEN - Test Case Development and Test Coverage Analysis
ASE - Ada Syntax Editor
QUALGEN - Over 60 Quality Metrics, with graphics, statistics, etc...
We are busy creating CDADL (our PDL for C). Following CDADL we will
port the rest of our tools to support C. CDADL should be available this
year, ADADL will provide a way to PDL now, and migrate to CDADL painlessly
when it is ready.
I did not see you address in the new article. Please call to ask for our
information brochures, and to discuss what you need from a C PDL.
Thanks,
Ken Nelson (Principle Engineer)
Software Systems Design
3267 Padua Av, Claremont, CA 91711
(714) 625-6147
P.S. You can ask for me or Tom Radi who is the president of the
company.
Date: Tue, 01 Aug 89 13:05:42 -0700
From: Steve Caine <haven!cfg.com!shc>
I'm sending off some info this afternoon. If you have questions after
looking at it, give me a call at (818) 449-3070.
Btw, GTE Government Systems has something like 15 copies of the VMS
version of PDL/81 -- mostly in Needham Heights. GTE, as a whole,
has something like 35 copies. They've been using it, and it's
predecessor, since 1975.
Steve. shc@cfg.com or ...!uunet!cfg!shc
--
=============================================================================
Pat @ grebyn.com | If the human mind was simple enough to understand,
301-948-8142 | We'd be too simple to understand it.
=============================================================================