comp.lang.ada
 help / color / mirror / Atom feed
* list of errors whose detection is required by RM
@ 2011-08-18 16:07 Baptiste Fouques
  2011-08-18 20:41 ` Randy Brukardt
  0 siblings, 1 reply; 2+ messages in thread
From: Baptiste Fouques @ 2011-08-18 16:07 UTC (permalink / raw)


Hi all,

I am looking for a referenced list of errors whose detection is
required by the Ada Standard. (either 95 or 2005).

The RM, §1.1.3 is clear on the point that compliant compiler is
required to detect every error specified in the standard (RM, that
is). The next chapter in the RM gives a classification of errors,
mainly by the detection point, and the expected behaviour after the
detection.

Then, all over the RM chapters, there is the specification of expected
checks at any relevant point. By the way, those checks are not linked
to the error classification.

I am looking for a single list of those checks.

Given this list, it is easy to demonstrate that no other tool than a
compliant compiler is required to prove the absence of given type of
errors (depending on the list). I can't find such a list.

If you know where I can find this list, please give me the link ...

Thanks.



^ permalink raw reply	[flat|nested] 2+ messages in thread

* Re: list of errors whose detection is required by RM
  2011-08-18 16:07 list of errors whose detection is required by RM Baptiste Fouques
@ 2011-08-18 20:41 ` Randy Brukardt
  0 siblings, 0 replies; 2+ messages in thread
From: Randy Brukardt @ 2011-08-18 20:41 UTC (permalink / raw)


[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 2664 bytes --]

"Baptiste Fouques" <bateast@bat.fr.eu.org> wrote in message 
news:26839fca-37be-4c24-bb04-57fb527e8b66@s2g2000vby.googlegroups.com...
>Hi all,
>
>I am looking for a referenced list of errors whose detection is
>required by the Ada Standard. (either 95 or 2005).
>
>The RM, �1.1.3 is clear on the point that compliant compiler is
>required to detect every error specified in the standard (RM, that
>is). The next chapter in the RM gives a classification of errors,
>mainly by the detection point, and the expected behaviour after the
>detection.
>
>Then, all over the RM chapters, there is the specification of expected
>checks at any relevant point. By the way, those checks are not linked
>to the error classification.
>
>I am looking for a single list of those checks.
>
>Given this list, it is easy to demonstrate that no other tool than a
>compliant compiler is required to prove the absence of given type of
>errors (depending on the list). I can't find such a list.
>
>If you know where I can find this list, please give me the link ...

So far as I know, no such list exists. It would be helpful in verifying the 
coverage (or lack thereof) of the ACATS test suite. [Which, BTW, is a 
redundancy: the "TS" in ACATS stands for "Test Suite"; but if I just write 
"ACATS" half of the people don't know what I mean.]

Probably the closest thing that exists is the Test Objective spreadsheets 
that I created for the Ada 2005 test suite, in order to gauge coverage. 
Those list all of the objectives for particular sections, which include 
checks (both compile-time and run-time) that need to be tested. But those 
only cover a small fraction of the language (about 12% of the core) - there 
hasn't been enough time or money to do more. You can find those objectives 
at http://www.ada-auth.org/acats.html -- look for the link to "Test 
Objective Files" at the very bottom of the page. [There are newer versions 
of these files that have never been posted - ask if you want the newer 
versions.]

Note that what you are asking for is a *very* long list. For Ada 2005, I 
have objectives created for 36 clauses out of 292 core clauses. This 
resulted in 875 total objectives. Extrapolating over the entire standard, 
that would result in 7100 objectives for the core alone (for Ada 2012, there 
are 468 clauses, giving an estimate of 11,300 objectives for the entire 
standard). Not all of these objectives relate to "checks" of course, but a 
large fraction do (most of rest relate to run-time semantics, for instance 
checking that the correct limb of a case statement is selected).

                                                   Randy.




Thanks. 





^ permalink raw reply	[flat|nested] 2+ messages in thread

end of thread, other threads:[~2011-08-18 20:41 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2011-08-18 16:07 list of errors whose detection is required by RM Baptiste Fouques
2011-08-18 20:41 ` Randy Brukardt

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