unofficial mirror of bug-gnu-emacs@gnu.org 
 help / color / mirror / code / Atom feed
From: Eli Zaretskii <eliz@gnu.org>
To: Paul Eggert <eggert@cs.ucla.edu>,
	Stefan Monnier <monnier@iro.umontreal.ca>
Cc: luangruo@yahoo.com, 73307@debbugs.gnu.org, wiz@gatalith.at
Subject: bug#73307: Fix ctype(3) usage
Date: Wed, 18 Sep 2024 14:45:25 +0300	[thread overview]
Message-ID: <86r09h9odm.fsf@gnu.org> (raw)
In-Reply-To: <89deb908-92a6-489f-bab9-116556ce8ce4@cs.ucla.edu> (message from Paul Eggert on Tue, 17 Sep 2024 17:05:29 -0700)

> Date: Tue, 17 Sep 2024 17:05:29 -0700
> Cc: 73307-done@debbugs.gnu.org, Thomas Klausner <wiz@gatalith.at>,
>  Po Lu <luangruo@yahoo.com>
> From: Paul Eggert <eggert@cs.ucla.edu>
> 
> On 2024-09-17 04:52, Eli Zaretskii wrote:
> > However, I'm not sure this is the right fix, the function is
> > defined with argument type of 'int'.  Paul, any comments?
> 
> Although that patch was an improvement it still had problems, as it 
> incorrectly assumed the string does not end in a multibyte space, and 
> that Emacs's locale matches the system's.
> 
> Emacs itself should not use <ctype.h> unless it knows the string is 
> unibyte and the system locale matches Emacs's. I scanned through its 
> source code looking for all problematic instances of <ctype.h> that have 
> crept in (except I didn't scan the MS-Windows code, where you're the 
> expert), and found five other places where ctype.h was obviously 
> misused. I installed the attached to fix these glitches and am boldly 
> closing this the report.
> 
> I can't easily test patch 0003, which fixes Android-specific code. 
> Although I think it's an improvement, in unlikely cases I suspect it 
> still doesn't exactly match what the Android kernel does with #! lines. 
> I don't know whether that matters. I'll CC this to Po Lu (my goto person 
> for Android) as a heads-up.

Thanks.

This part doesn't look right to me:

> -  {
> -    char *s = SSDATA (prompt);
> -    ptrdiff_t len = strlen (s);
> -    if ((len > 0) && !isspace (s[len - 1]))
> -      prompt = CALLN (Fconcat, prompt, build_string (" "));
> -  }
> -  prompt = CALLN (Fconcat, prompt, Vyes_or_no_prompt);
> +  ptrdiff_t promptlen = SCHARS (prompt);
> +  bool prompt_ends_in_nonspace
> +    = (0 < promptlen
> +       && (SYNTAX (XFIXNAT (Faref (prompt, make_fixnum (promptlen - 1))))
> +	   != Swhitespace));
> +  AUTO_STRING (space_string, " ");
> +  prompt = CALLN (Fconcat, prompt,
> +		  prompt_ends_in_nonspace ? space_string : empty_unibyte_string,
> +		  Vyes_or_no_prompt);

You are testing characters for whitespace syntax, which is AFAIU
subject to buffer-local syntax tables.  Thus, a strange enough setting
of buffer syntax could make this code return unexpected results.

Stefan, am I right?

Why can't we use something closer to the original code, which doesn't
depend on buffer-local customizations?  In particular, what's the
problem with using c_isspace (and what do you mean by "multibyte
space"?)






  parent reply	other threads:[~2024-09-18 11:45 UTC|newest]

Thread overview: 10+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-09-16 21:25 bug#73307: Fix ctype(3) usage Thomas Klausner
2024-09-17 11:52 ` Eli Zaretskii
2024-09-17 11:55   ` Thomas Klausner
2024-09-18  0:05   ` Paul Eggert
2024-09-18  2:13     ` Po Lu via Bug reports for GNU Emacs, the Swiss army knife of text editors
2024-09-18 11:45     ` Eli Zaretskii [this message]
2024-09-18 15:18       ` Stefan Monnier via Bug reports for GNU Emacs, the Swiss army knife of text editors
2024-09-18 15:35       ` Paul Eggert
2024-09-18 16:11         ` Eli Zaretskii
2024-09-18 16:22           ` Paul Eggert

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

  List information: https://www.gnu.org/software/emacs/

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=86r09h9odm.fsf@gnu.org \
    --to=eliz@gnu.org \
    --cc=73307@debbugs.gnu.org \
    --cc=eggert@cs.ucla.edu \
    --cc=luangruo@yahoo.com \
    --cc=monnier@iro.umontreal.ca \
    --cc=wiz@gatalith.at \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
Code repositories for project(s) associated with this public inbox

	https://git.savannah.gnu.org/cgit/emacs.git

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).