Question on Ada.Strings.Unbounded.Slice

Question on Ada.Strings.Unbounded.Slice

[Brian Nettleton @pulsar]

While building David Wheeler's "small" program (part of the Lovelace
tutorial) I came across a problem with Ada.Strings.Unbounded.Slice.

I'm getting Index_error when the High > Length(Source) + 1 (where
High, Length and Source are as defined in spec of Ada.Strings.Unbounded).

Here is a smaller program which displays the problem (i.e. this program
raises Ada.Strings.Index_Error):


     with Ada.Strings.Unbounded;
     with Ada.Text_Io;
     procedure Do_Slice is
 
       package TIO renames Ada.Text_Io;
 
     begin
       TIO.Put_Line
         ( Ada.Strings.Unbounded.Slice
           ( Source => ASU.To_Unbounded_String( "hello" ),
             Low    => 1,
             High   => 7 
           ) 
         );
     end Do_Slice;


RM A.4.4:101 states Slice "propagates Index_Error if Low > 
Length(Source) + 1."  But the RM is silent on what happens when
High > Length(Source) + 1!  Is this an oversight in the RM, or
by silence does the RM mean to imply that Index_Error is only
raised when Low exceeds the bounds of the source?  Is it implementation
dependent what happens when High > Length(Source) + 1?

-Brian Nettleton

Question on Ada.Strings.Unbounded.Slice

[Brian Nettleton @pulsar]

Doesn't look like my original posting went through.  Here's another try
(my apologies if you receive this twice).



While building David Wheeler's "small" program (part of the Lovelace
tutorial) I came across a problem with Ada.Strings.Unbounded.Slice.

I'm getting Index_error when the High > Length(Source) + 1 (where
High, Length and Source are as defined in spec of Ada.Strings.Unbounded).

Here is a smaller program which displays the problem (i.e. this program
raises Ada.Strings.Index_Error):


     with Ada.Strings.Unbounded;
     with Ada.Text_Io;
     procedure Do_Slice is
 
       package TIO renames Ada.Text_Io;
 
     begin
       TIO.Put_Line
         ( Ada.Strings.Unbounded.Slice
           ( Source => ASU.To_Unbounded_String( "hello" ),
             Low    => 1,
             High   => 7 
           ) 
         );
     end Do_Slice;


RM A.4.4:101 states Slice "propagates Index_Error if Low > 
Length(Source) + 1."  But the RM is silent on what happens when
High > Length(Source) + 1!  Is this an oversight in the RM, or
by silence does the RM mean to imply that Index_Error is only
raised when Low exceeds the bounds of the source?  Is it implementation
dependent what happens when High > Length(Source) + 1?

-Brian Nettleton

Re: Question on Ada.Strings.Unbounded.Slice

[Keith Thompson]

In <DuE7D3.194@thomsoft.com> bn@alsys.com (Brian Nettleton @pulsar) writes:
> While building David Wheeler's "small" program (part of the Lovelace
> tutorial) I came across a problem with Ada.Strings.Unbounded.Slice.
> 
> I'm getting Index_error when the High > Length(Source) + 1 (where
> High, Length and Source are as defined in spec of Ada.Strings.Unbounded).

[example deleted]

> RM A.4.4:101 states Slice "propagates Index_Error if Low > 
> Length(Source) + 1."  But the RM is silent on what happens when
> High > Length(Source) + 1!  Is this an oversight in the RM, or
> by silence does the RM mean to imply that Index_Error is only
> raised when Low exceeds the bounds of the source?  Is it implementation
> dependent what happens when High > Length(Source) + 1?

According to the current (unapproved) version of AI95-00128, "A call to
Bounded.Slice with High > Length(Source) is equivalent to a call with
High = Length(Source)."  In other words, a slice that overlaps the end
of the string is silently truncated.

Personally, I disagree with this; I think the Slice function should
correspond to the predefined slice operation for type String (raising
Index_Error when a slice would raise Constraint_Error).  At least two
existing implementations do it this way.  I've just submitted a comment
to this effect to the ada-comment mailing list; we'll see what happens.

Incidentally, the problem occurs when High > Length(Source), not just
when High > Length(Source) + 1.

AI95-00128 is one of many commentaries on the Ada 95 standard; they're
available at

    <ftp://sw-eng.falls-church.va.us/public/AdaIC/standards/95com/ada-issues>

Note that these commentaries have little or no official standing until
they've gone through an approval process.  Someone more directly involved
than I am may wish to comment further on this process.

-- 
Keith Thompson (The_Other_Keith) kst@thomsoft.com <*>
TeleSoft^H^H^H^H^H^H^H^H Alsys^H^H^H^H^H Thomson Software Products
10251 Vista Sorrento Parkway, Suite 300, San Diego, CA, USA, 92121-2718
"As the most participatory form of mass speech yet developed, the Internet
deserves the highest protection from government intrusion." -- ACLU v. Reno

Re: Question on Ada.Strings.Unbounded.Slice

[Robert A Duff]

In article <DuMvF0.B4v@thomsoft.com>, Keith Thompson <kst@thomsoft.com> wrote:
>    <ftp://sw-eng.falls-church.va.us/public/AdaIC/standards/95com/ada-issues>
>
>Note that these commentaries have little or no official standing until
>they've gone through an approval process.  Someone more directly involved
>than I am may wish to comment further on this process.

I believe there is a file in the same directory explaining the policies
and procedures, and what the various statuses mean, and so forth.

- Bob

Re: Question on Ada.Strings.Unbounded.Slice

[David Wheeler]

I believe this conversation started due to an unexpected exception
being raised in program "Small" in the parser (my local news reader has
lost the beginning of this thread, so I can't be sure).

I suspect that the current version fixes the problem, you might check
it out.

---- David