Re: <ctype.h> toLower()
From: Mike Wahler (mkwahler_at_mkwahler.net)
Date: 11/18/03
- Next message: Richard Heathfield: "Re: <ctype.h> toLower()"
- Previous message: Mike Wahler: "Re: strings"
- In reply to: ellie fant: "Re: <ctype.h> toLower()"
- Next in thread: Richard Heathfield: "Re: <ctype.h> toLower()"
- Reply: Richard Heathfield: "Re: <ctype.h> toLower()"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Date: Tue, 18 Nov 2003 19:40:43 GMT
"ellie fant" <pcrcutitout1000011@uko2.co.uk> wrote in message
news:1069183013.607789@news.minx.net.uk...
> Richard heathfield wrote:
> > No, that would be very petty. You did indeed point out an error in one
of
> my
> > articles, for which I'm grateful. Why not be grateful that I have
pointed
> > out an error in yours?
> >
>
> In reality 99% of all working code is not portable
Non-sequitur. This newsgroup is about ISO standard C, not
about any 'working code'.
> so lets stop trying to
> pretend portable code is the only code that is correct.
It's the only code which is correct in the context of
comp.lang.c. Did you not read the group's welcome message?
> Do the rules of this newsgroup state that we must only discuss portable
code
Yes. Please see:
http://www.angelfire.com/ms3/bchambless0/welcome_to_clc.html
Nonportable code is necessarily platform-dependent, and
there already exist newsgroups which discuss those platforms
and their issues.
> or is this something you have personally decided to include?
Richard (and I, and I presume most/many others who participate
here) have 'personally decided' to follow the newsgroup guidelines
stated in the link I cite above.
>Is code wrong
> because it is not portable?
It's not 'wrong', but neither is it topical for comp.lang.c
>
> So what is WRONG with OR'ing an uppercase ASCII alpha character with 20h
to
> make it lowercase?
What's wrong is that a portable C program cannot assume its
host system uses a particular character set. The only gurantee
made my the standard language about the encoding values of characters
is that the digit characters 0 - 9 have sequential values.
> A quasirandom value as you called it.- lol
A good term for referring to it imo.
> I'll tell you what's good about it. It's more efficient that using tolower
> function.
You don't know that. I suspect that for implementations on systems
that do use ASCII, the library functions having to do with changing
case do indeed take advantage of the particular attributes of the
host character set, such as the 'symmetry' of ASCII alpha character
codes. The implementation is allowed to take such advantage of
intimate knowledge of the platform, portable source code is not.
> And do you still say that it doesn't work? because it does.
No it does not. Try expanding your test data set to include
*all* possible character value inputs.
> If you can genuinely point out an error with this method I will eat cows
> turd
Here's a fork. Eat up. :-) Richard has already pointed out
your errors.
>but the fact that it is not 100% portable is not an error.
"Error" means nothing without context. It is indeed an error
to post or suggest nonstandard code in comp.lang.c, since it's
not topical here.
Despite that, even on an ASCII system, your code *does not work*.
>
> So lets see you point out a real error rather than making one up out of
the
> blue.
Try feeding your function the string:
"ABC123 $100"
-Mike
- Next message: Richard Heathfield: "Re: <ctype.h> toLower()"
- Previous message: Mike Wahler: "Re: strings"
- In reply to: ellie fant: "Re: <ctype.h> toLower()"
- Next in thread: Richard Heathfield: "Re: <ctype.h> toLower()"
- Reply: Richard Heathfield: "Re: <ctype.h> toLower()"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Relevant Pages
|
