comp.lang.ada
 help / color / mirror / Atom feed
* Source code representation
@ 2000-11-23  0:00 Florian Weimer
  2000-11-23  0:00 ` Ted Dennison
  2000-11-27  0:00 ` Tucker Taft
  0 siblings, 2 replies; 4+ messages in thread
From: Florian Weimer @ 2000-11-23  0:00 UTC (permalink / raw)


The file-centered approach used by GNAT is close enough to what other
programming language implementations do on unixoid platforms
traditionally.  I've never used another Ada compiler, and I wonder if
other vendors choose a different model (i.e. precompilation of specs
and bodies and some kind of library repository).  Are there any other
Ada compilers which can directly use separate body/spec files in the
way GNAT mandates them?

(I'm interested in this because I'm going to try to push some
Ada-related directory into the FHS standard, and the following
question arises: Are directories containing .ads/.adb files
GNAT-specific, practically speaking, or not?)




^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: Source code representation
  2000-11-23  0:00 Source code representation Florian Weimer
@ 2000-11-23  0:00 ` Ted Dennison
  2000-11-27  0:00 ` Tucker Taft
  1 sibling, 0 replies; 4+ messages in thread
From: Ted Dennison @ 2000-11-23  0:00 UTC (permalink / raw)


Florian Weimer wrote:

> The file-centered approach used by GNAT is close enough to what other
> programming language implementations do on unixoid platforms
> traditionally.  I've never used another Ada compiler, and I wonder if
> other vendors choose a different model (i.e. precompilation of specs

The Modula-2 implemenations I've worked with all used a library approach
similar to what traditional Ada compilers use. I believe Oberon uses a
library as well. C's just file-based because its modularity support is so
weak and low-level that it has no choice.

I'd think the best way for you to get a handle on this would be to go
download the evaluation version of ObjectAda and play with it a while.
(Go to http://www.aonix.com and click on "Downloads" on the sidebar.) It
uses Tucker's Favorite Front End, which is both widely used by non-Gnat
compilers, and fairly representative of a typical approach.

--
T.E.D.

Home - mailto:dennison@telepath.com  Work - mailto:dennison@ssd.fsi.com
WWW  - http://www.telepath.com/dennison/Ted/TED.html  ICQ  - 10545591






^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: Source code representation
  2000-11-27  0:00 ` Tucker Taft
@ 2000-11-27  0:00   ` Ted Dennison
  0 siblings, 0 replies; 4+ messages in thread
From: Ted Dennison @ 2000-11-27  0:00 UTC (permalink / raw)


In article <3A228D7F.2E2EB8ED@averstar.com>,
  Tucker Taft <stt@averstar.com> wrote:

> For example, with the compilers based on
> the AverStar AdaMagic front end (Green Hills and Aonix),
> there is a "registration" step which creates a map between
> source code and Ada unit name.  This is optional if you compile
> in the "right" order, but if you want to compile in a random
> order, you need to initialize the "UNIT.MAP" file.  Note that
> both of these compilers come with their own integrated development
> environment, which may more or less hide this requirement by
> bundling it in with the notion of adding source files/directories
> to a "project" or equivalent.

Typically there is also some way to "mass register" a group of files.
Once that is done, the process is little different from the user's point
of view than it is with Gnat.

--
T.E.D.

http://www.telepath.com/~dennison/Ted/TED.html


Sent via Deja.com http://www.deja.com/
Before you buy.




^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: Source code representation
  2000-11-23  0:00 Source code representation Florian Weimer
  2000-11-23  0:00 ` Ted Dennison
@ 2000-11-27  0:00 ` Tucker Taft
  2000-11-27  0:00   ` Ted Dennison
  1 sibling, 1 reply; 4+ messages in thread
From: Tucker Taft @ 2000-11-27  0:00 UTC (permalink / raw)


Florian Weimer wrote:
> 
> The file-centered approach used by GNAT is close enough to what other
> programming language implementations do on unixoid platforms
> traditionally.  I've never used another Ada compiler, and I wonder if
> other vendors choose a different model (i.e. precompilation of specs
> and bodies and some kind of library repository).  Are there any other
> Ada compilers which can directly use separate body/spec files in the
> way GNAT mandates them?

All Ada compilers can handle source split up into separate
body/spec files.  However, some require Ada code to have
a particular filename suffix (e.g. ".a"), though that seems
rarer these days.  Because of GNAT's popularity, almost all
compilers accept the .ads/.adb suffix as well.  

I suppose part of your question might be whether you have
to take some additional step, or compile in some particular
order, with a non-GNAT compiler.  The answer is probably yes.
However, that doesn't really affect the representation of
the source.  

For example, with the compilers based on
the AverStar AdaMagic front end (Green Hills and Aonix),
there is a "registration" step which creates a map between
source code and Ada unit name.  This is optional if you compile
in the "right" order, but if you want to compile in a random
order, you need to initialize the "UNIT.MAP" file.  Note that
both of these compilers come with their own integrated development
environment, which may more or less hide this requirement by
bundling it in with the notion of adding source files/directories
to a "project" or equivalent.

> (I'm interested in this because I'm going to try to push some
> Ada-related directory into the FHS standard, and the following
> question arises: Are directories containing .ads/.adb files
> GNAT-specific, practically speaking, or not?)

GNAT-compatible source file organization is generally accepted by any 
Ada 95 compiler, though as mentioned above, you may have to
compile in a particular order or perform some kind of "registration"
step.

-- 
-Tucker Taft   stt@averstar.com   http://www.averstar.com/~stt/
Technical Director, Commercial Division, AverStar (formerly Intermetrics)
(http://www.averstar.com/services/IT_consulting.html)  Burlington, MA  USA




^ permalink raw reply	[flat|nested] 4+ messages in thread

end of thread, other threads:[~2000-11-27  0:00 UTC | newest]

Thread overview: 4+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2000-11-23  0:00 Source code representation Florian Weimer
2000-11-23  0:00 ` Ted Dennison
2000-11-27  0:00 ` Tucker Taft
2000-11-27  0:00   ` Ted Dennison

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox