Re: Loops (was Re: do { quit; } else { })

On 21.04.2025 01:29, bart wrote:

On 20/04/2025 23:36, Keith Thompson wrote:

bart <bc@freeuk.com> writes:

 

     for(p=sqliteHashFirst(&pSchema->trigHash); p; p=sqliteHashNext(p)){
         sqlite3DeleteTrigger(db, (Trigger*)sqliteHashData(p));
     }

 

>
I might write it like this:
>
     for ( p = sqliteHashFirst(&pSchema->trigHash);
           p != NULL;
           p = sqliteHashNext(p) )
     {
         sqlite3DeleteTrigger(db, (Trigger*)sqliteHashData(p));
     }

 

I have certain preferences (spaces around most operators, explicit
comparison to NULL, willingness to split long lines) that other C
programmers may or may not share.

 
I rarely see complex loops split over multiple lines

(There's obviously a lot you haven't seen.)

(mainly when they
get so long that they'd overflow the line, but they can still be complex
enough before then).
 
But the thing is, once you split it into multiple lines, then there is
little advantage over using a regular 'while' loop:

You still have the controlling elements in one place with 'for'.

[...]
 

It keeps it even more together, which you seem to like.

>
That's something you invented.  I find it ugly, and I presume you'd
agree.  The fact that you think that someone else would prefer it
indicates that you don't understand how other people think.

 
AFAIK it is legal C code, and I invented it because somebody said things
that belong together should be together in one place.

Yes, that's a commonplace. (But "one place" doesn't mean "stuffed" or
"all in one line".)

However, I have
seen actual examples like that, in for-headers that that use
comma-separated expressions.

(You can see and find arbitrary code, better or worse.)

 
 

I'd rather not write `for (ch in 0..255)` because it's a syntax error.

 
It's a syntax error because the form doesn't naturally exist in C; you'd
have to emulate using macros, which is a poor solution.

Yes, and that's his point; it's not "C".

If I want abstractions I don't (unless I have to) use "C". If I want to
have clean concepts, say iterating over an indexed array, I can use in
"C"

  for (i=0; i<N; i++)
    arr[i] = fun (arr[i]);

which is clean enough for me. But I might prefer (with other languages)
to not care about the bounds in the first place, and use

  for (i := lwb(arr) to upb(arr))    # not "C"
    arr[i] := fun (arr[i]);

if I'd want to avoid errors with the bounds. But the real thing I'd want
is a functional approach where I don't need indices at all

  iterate (arr, fun);    # not "C"

 

You have the luxury of using your own language.

 
That 'ch in 0..255' form or close equivalent is supported by that long
set of languages I listed earlier. It's not my invention, I just copied it.
 
It is just something that is desirable. Look again at the C version: it
looks off. (At least, you'd use a shorter loop index!)

There's a lot I'd like to have that I don't find in the "C" language.

If I choose to use (or have to use) "C" I use the options I have to best
solve my tasks.

[...]
 

If you don't like C-style for loop, they absolutely should not
exist in a language for which you are, if I understand correctly,
the sole implementer and the sole user.

 
But I hear so much about how wonderful it is, how multi-purpose, how
indispensible, how superior to an ordinary 'for' (admittedly from people
who don't have a choice) that I didn't want to miss out!

No one said that. - You should stop your regular misrepresentations and
also your misinterpretations! - STOP IT, stupid!

Janis