Thanks to those people who responded to my request on tools for
measuring 'C'. I got 4 responses, one of which pointed me at some PD

for this :-

It included two different methods of measuring code complexity. One
called halstead and one called mccabe.

I did not get any response about PD tools to find ambiguous or undefined 'c'
features, just references to a number of commercial products. These include
various versions of lint on various platforms and some other products, one
called C-Doc from a company called software blacksmiths :-

SOFTWARE BLACKSMITHS INC
ST.IVES WAY
MISSISSAUGA
ON
L5N4M1
(416) 858-4466

One called PC-Lint from a company called Gimpel software. Apparently
there may also be a source code version of this for other platforms.

GIMPEL SOFTWARE
TOWNSHIP LINE RD
NORRISTOWN
PA
19403
(215) 584-4261

Also one called QAC which I believe is a product of the company who
came to talk to us - Programming Research.

The version of lint on Digital's Alpha platform apparently includes an
option which looks at 32bit-64bit porting problems.

Thanks to :-

Jon

If metrics are a new game for you, I would suggest spending some time in
deciding what you want to collect.  Both Halstead and McCabe are fairly
abstract, so if you don't know what you're looking for neither of these
philosophies may provide it.  My organization uses the following as part
of our entry criteria for Fagan Inspections.

________________________________________________
|           |               |     Complexity   |
|Module Name|  Function Name| _________________|
|           |               |  McCabe   Density|
|___________|_______________|__________________|
|test1.c    |  func1        |  1        1.00   |
|           |  func2        |  2        0.67   |
|           |  Other        |  N/A      N/A    |
|___________|_______________|__________________|
|test2.c    |  Other        |  N/A      N/A    |
|___________|_______________|__________________|
|test3.c    |  func3        |  3        1.00   |
|           |  func4        |  1        1.00   |
|           |  Other        |  N/A      N/A    |
|___________|_______________|__________________|
|Totals     |               |  N/A      N/A    |
|Mean       |               |  1.75     0.92   |
|Std Dev    |               |  0.25     0.08   |
|Minimum    |               |  1.50     0.83   |
|Maximum    |               |  2.00     1.00   |
|___________|_______________|__________________|

_________________________________________________________________
|           |               |          Halstead Metrics         |
|Module Name|  Function Name| __________________________________|
|           |               |  Volume   Length   Purity   Effort|
|___________|_______________|___________________________________|
|test1.c    |  func1        |  25       9        1.42     50    |
|           |  func2        |  59       18       1.31     221   |
|           |  Other        |  N/A      N/A      N/A      N/A   |
|___________|_______________|___________________________________|
|test2.c    |  Other        |  N/A      N/A      N/A      N/A   |
|___________|_______________|___________________________________|
|test3.c    |  func3        |  183      44       1.30     1281  |
|           |  func4        |  15       6        1.67     30    |
|           |  Other        |  N/A      N/A      N/A      N/A   |
|___________|_______________|___________________________________|
|Totals     |               |  282      77       N/A      N/A   |
|Mean       |               |  141.00   38.50    1.42     395.56|
|Std Dev    |               |  57.00    11.50    0.06     259.94|
|Minimum    |               |  84       27       N/A      N/A   |
|Maximum    |               |  198      50       N/A      N/A   |
|___________|_______________|___________________________________|

Each of the above outputs were generated using an in-house PERL script.
I am sure available tools provide similar information.

The real risk behind collecting either of these metrics is transfering
their importance to the troops.  IMHO (most humble), metrics should be
used to measure something strickly for the purpose of improving future
assignments.  Realizing the Purity of a particular function does not
lend itself to reevaluation and scrutiny.

Just my $0.02.  I'll claim this post is valid in this newsgroup as
metrics evaluation is just as imperative as having good C code.

Doug