Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

Automated SLOC Collection

50 views
Skip to first unread message

Greg Stine

unread,
Sep 13, 1996, 3:00:00 AM9/13/96
to

As part of our metrics program, we are collecting Source Lines Of Code (SLOC) metrics. There are actually three numbers we are looking
at: number of new lines, number of lines deleted, and number of lines modified. We would like to automate as much of this collection as
possible. We are developing our current application in C/C++ and Gupta SQLWindows. Is anyone familiar with any tools (commercial or
otherwise) which would assist with this?

Thanks.

Greg Stine
SAIC - Fairmont, WV
W.GREGO...@cpmx.saic.com

Nancy Mead

unread,
Sep 13, 1996, 3:00:00 AM9/13/96
to

Make sure that you decide how you will define lines - will it include comment
lines? Source statements that extend across more than one line? Data? Reuse?
Once you decide that then you can look for a tool.

Alternatively, you can decide that you'll define lines according to what the
tool will support.

Sorry I can't help on the tools question, but I think it's important to
clearly define the way you want to collect lines so that the data makes sense.

Doug Toppin

unread,
Sep 16, 1996, 3:00:00 AM9/16/96
to

In article <32394D...@cpmx.saic.com>, Greg Stine <W.GREGO...@cpmx.saic.com> writes:
>As part of our metrics program, we are collecting Source Lines Of Code (SLOC) metrics. There are actually three numbers we are looking
>at: number of new lines, number of lines deleted, and number of lines modified. We would like to automate as much of this collection as
>possible. We are developing our current application in C/C++ and Gupta SQLWindows. Is anyone familiar with any tools (commercial or
>otherwise) which would assist with this?

We had to do the same activity a while back.
It became fairly simple because we use the CM tool PCMS which
supports baselining files.
I was able to release the baselined software we started with (created
a year or so ago at the beginning of the project).
At the end of the project I baselined and released the final version
of the software and then compared (using 'diff' in Unix) the beginning
and ending versions.
'diff'ing 2 files will produce an indication of lines that
are not the same between the files (added/modified, deleted).
Using the diff output I interpreted lines added and modified as lines
modified.
In general, I think that using a CM tool that supports baselining
(snapshotting the versions of all of your files)
greatly facilitates measurement activities.

--
-----------------------------------------------------------------------------
Doug Toppin, Mail Stop N102
E-Systems, Falls Church Division E-Mail: dto...@melpar.esys.com
7700 Arlington Blvd Voice: 703-560-5000, x2050, RayNet 821-2050
Falls Church, VA 22046 FAX: 703-849-1614


0 new messages