From mboxrd@z Thu Jan  1 00:00:00 1970
X-Spam-Checker-Version: SpamAssassin 3.4.6 (2021-04-09) on
	ip-172-31-65-14.ec2.internal
X-Spam-Level: 
X-Spam-Status: No, score=-1.9 required=3.0 tests=BAYES_00,T_SCC_BODY_TEXT_LINE
	autolearn=ham autolearn_force=no version=3.4.6
Path: eternal-september.org!reader01.eternal-september.org!aioe.org!Yr2L9qaRshuL+yFKTxsDBA.user.46.165.242.75.POSTED!not-for-mail
From: Simon Wright <simon@pushface.org>
Newsgroups: comp.lang.ada
Subject: Re: Compiler error (2) ?
Date: Mon, 26 Sep 2022 09:12:05 +0100
Organization: Aioe.org NNTP Server
Message-ID: <lyedvy1rmy.fsf@pushface.org>
References: <7085edb5-688c-410c-b401-23ff3ca2a38fn@googlegroups.com>
Mime-Version: 1.0
Content-Type: text/plain
Injection-Info: gioia.aioe.org; logging-data="62277"; posting-host="Yr2L9qaRshuL+yFKTxsDBA.user.gioia.aioe.org"; mail-complaints-to="abuse@aioe.org";
User-Agent: Gnus/5.13 (Gnus v5.13) Emacs/28.1 (darwin)
Cancel-Lock: sha1:J+5Q9HsgC3mV9maFQ4NSEZkFfQk=
X-Notice: Filtered by postfilter v. 0.9.2
Xref: reader01.eternal-september.org comp.lang.ada:64413
List-Id: <comp.lang.ada>

I'm not sure what compiler error you mean.

Agreed that the code as stands reports a strange issue,

    11.    S_Names: constant S_Names_P.Vector := [for I in 3..N => S0];
                                                 |
        >>> error: expected type "Standard.Integer"
        >>> error: found type "Ada.Containers.Count_Type"

which does look like a compiler error (even if it's only misreporting a
valid issue, I'm not sure).

What's more interesting is what happens if you make your suggested
change and it compiles. When you run it,

   $ ./reinert_2 

   raised CONSTRAINT_ERROR : test1a.s_names_p.Replace_Element: Index is
   out of range

Running this under the debugger, with 'catch exception',

   Catchpoint 1, CONSTRAINT_ERROR (test1a.s_names_p.Replace_Element: Index is out of range) at 0x0000000100008ae5 in test1a () at reinert_2.adb:15
   15	   s_names: constant s_names_p.Vector := [for i in 3..n => S0];

So, what's Index?

   (gdb) p index
   No definition of "index" in current context.

Looing at the backtrace,

   (gdb) bt
   [...]
   #4  0x0000000100011711 in test1a.s_names_p.replace_element (container=..., 
       index=6, new_item=s0)
       at /opt/gcc-12.1.0/lib/gcc/x86_64-apple-darwin15/12.1.0/adainclude/a-convec.adb:2522
   #5  0x0000000100008ae5 in test1a () at reinert_2.adb:15

Frame 4 is in the runtime; maybe we can see what's going on,

   (gdb) fr 4
   #4  0x0000000100011711 in test1a.s_names_p.replace_element (container=..., 
       index=6, new_item=s0)
       at /opt/gcc-12.1.0/lib/gcc/x86_64-apple-darwin15/12.1.0/adainclude/a-convec.adb:2522
   2522	         raise Constraint_Error with "Index is out of range";
   (gdb) p index
   $1 = 6

Hmm. What does the source say?

   (gdb) l
   2517	   is
   2518	   begin
   2519	      TE_Check (Container.TC);
   2520	
   2521	      if Checks and then Index > Container.Last then
   2522	         raise Constraint_Error with "Index is out of range";
   2523	      end if;
   2524	
   2525	      Container.Elements.EA (Index) := New_Item;
   2526	   end Replace_Element;

So Index (6) is more than Container.Last:

   (gdb) p container
   $2 = (elements => 0x600000008000, last => 5, tc => (busy => 0, lock => 0))
   (gdb)

6 is certainly more than 5. Where does 5 come from?

Index is Positive. We were hoping to have our container with indices
supporting values up to 7, but the first index of a container has to be
Index_Type'First i.e. 1; the compiler has looked at the length of the
aggregate (5) and created an empty Vector s_names of that length.

Clearly you can't do slices in this context!