Re: Examine my code!

Re: Examine my code!

[Dale Stanbrough]

Robert Dewar writes:

"A somewhat cryptic post :-)
 
 Dale Stanbrough: >No!
 Dale"


:-) Ah, the toubles you have when you use good tools :-) My threaded 
newsreader quite ably showed this as a followup to a post entitled
"Examine my code!". Ah well, no more subject line jokes i suppose!

Dale :-)

Re: No!

[Gautier]

> Yes!

Me too.

Re: No!

[Oliver Kellogg]

<segmentation fault, core dumped>

Re: Examine my code!

[Nick Roberts]

The answer is to use 

   Get_Line(Command);

instead of just

   get(command)

Reason why: things getting stuck in the buffer.

Nick.

Re: Examine my code!

[Centaury]

You can't get_line a variable, can you?
-- 
Please mail replies to :
utopian@pl.jaring.my

Re: Examine my code!

[Keith Thompson]

In <01bc4d85$4c127b40$f4f582c1@xhv46.dial.pipex.com> "Nick Roberts" <Nick.Roberts@dial.pipex.com> writes:
> The answer is to use 
> 
>    Get_Line(Command);
> 
> instead of just
> 
>    get(command)
> 
> Reason why: things getting stuck in the buffer.

That's the problem, but it's not the solution.

In the previously posted program, Command is an object of an enumeration
type.  There is no Get_Line procedure for enumeration types (though it
wouldn't be difficult to write one).

Take a look at the original program and consider exactly what happens
if the statement Get(Command) raises an exception.

The solution, of course, is left as an exercise for the student.

(I welcome e-mail feedback from educators on whether this is the right
level of response for this kind of question.)

-- 
Keith Thompson (The_Other_Keith) kst@sd.aonix.com <http://www.aonix.com> <*>
TeleSo^H^H^H^H^H^H Alsy^H^H^H^H Thomson Softw^H^H^H^H^H^H^H^H^H^H^H^H^H Aonix
5040 Shoreham Place, San Diego, CA, USA, 92122-5989
"Humor is such a subjective thing." -- Cartagia

Text input on a console

[Nick Roberts]

Keith Thompson <kst@sd.aonix.com> wrote in article
<E961o5.AuL@thomsoft.com>...
> In <01bc4d85$4c127b40$f4f582c1@xhv46.dial.pipex.com> "Nick Roberts"
<Nick.Roberts@dial.pipex.com> writes:
> > The answer is to use 
> > 
> >    Get_Line(Command);
> > 
> > instead of just
> > 
> >    get(command)
> > 
> > Reason why: things getting stuck in the buffer.
> 
> That's the problem, but it's not the solution.
> 
> In the previously posted program, Command is an object of an enumeration
> type.  There is no Get_Line procedure for enumeration types (though it
> wouldn't be difficult to write one).


I blunder again. Sorry (the little yellow tablets strike again).

However, this does raise another point, of more general interest. The
behaviour of a program when inputting text from a console is relatively
undefined in Ada (as in most languages). It may be impracticable to try to
formally define things more closely (i.e. in the RM), but it may be
sensible to informally define things a little bit more than the RM,
somehow.

When GET is used to read a string, or other type, from the console, should
the program wait for a carriage-return ('cooked' or 'buffered' mode) or
just get the characters directly ('raw' mode)? Should special characters
(like backspace, control-D, control-Z, ESC, etc.) be automatically
supported? Should there be some agreed parameters in the FORM parameter to
specify this sort of thing? (And, in which case, what do we do about
STANDARD_INPUT?)

Comments welcome.

Nick.

Re: Text input on a console

[Jerry van Dijk]

In article <01bc524d$a0fe21a0$28f982c1@xhv46.dial.pipex.com> Nick.Roberts@dial.pipex.com writes:

>However, this does raise another point, of more general interest. The
>behaviour of a program when inputting text from a console is relatively
>undefined in Ada (as in most languages). It may be impracticable to try to
>formally define things more closely (i.e. in the RM), but it may be
>sensible to informally define things a little bit more than the RM,
>somehow.

"Define informally" ?

Although I sympathize with the idea, how would one go about this ?

>When GET is used to read a string, or other type, from the console, should
>the program wait for a carriage-return ('cooked' or 'buffered' mode) or
>just get the characters directly ('raw' mode)?

Like Get vs Get_Immediate ?

The problem is of course how to go about it in a portable way as this is
purely OS dependend.

>Should special characters
>(like backspace, control-D, control-Z, ESC, etc.) be automatically
>supported?

As these are part of Latin-1 that shouldn't be much of a problem, but
how about Ctrl-Home or Gold-F1 ? Or are you talking about the Line Editing
features of the OS ?

> Should there be some agreed parameters in the FORM parameter to
>specify this sort of thing? (And, in which case, what do we do about
>STANDARD_INPUT?)

One solution might be the way C compilers for DOS solved it: using a special
Console I/O (conio) package. This could be in a special Annex, applicable
only to platforms that support terminal based or interactive I/O.

Since the IS Annex comes close to this, I wonder if such a package
(which would do wonders for the portability of console apps) has been
considered.

A good candidate would be a binding to termios, if it were supported
on all platforms (GNAT doesn't, it even can't agree on the implementation
of Get_Immediate between different ports...)

>Comments welcome.

You asked for it :-)

--

-- Jerry van Dijk       | Leiden, Holland
-- Business Consultant  | Team Ada
-- Ordina Finance       | jdijk@acm.org

Examine my code!

[Odo Wolbers]

> SkipLine;
> 
> in this expression to throw away the Inputline like this:
> 
> exception
>      when data_error=>
>           put("Invalid command! Type help.");
>           newline;
>              SkipLine;
> end;  

Oops, where are the underscores?

I meant

Skip_Line;

and

New_Line;

ciao
Odo