From: "Arthur J. O'Dwyer" <ajo AT andrew DOT cmu DOT edu>
Newsgroups: comp.os.msdos.djgpp
Subject: Re: Bug in command-line globbing
Date: Fri, 13 Dec 2002 12:26:14 -0500 (EST)
Organization: Carnegie Mellon, Pittsburgh, PA
Lines: 87
Message-ID: <Pine.GSO.4.44L-027.0212131159570.4357-100000@unix4.andrew.cmu.edu>
NNTP-Posting-Host: smtp5.andrew.cmu.edu
Mime-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
X-Trace: bb3.andrew.cmu.edu 1039800374 16047 128.2.10.85 (13 Dec 2002 17:26:14 GMT)
X-Complaints-To: advisor AT andrew DOT cmu DOT edu
NNTP-Posting-Date: 13 Dec 2002 17:26:14 GMT
In-Reply-To: <4634-Fri13Dec2002111638+0200-eliz@is.elta.co.il>
To: djgpp AT delorie DOT com
DJ-Gateway: from newsgroup comp.os.msdos.djgpp
Reply-To: djgpp AT delorie DOT com
Errors-To: nobody AT delorie DOT com
X-Mailing-List: djgpp AT delorie DOT com
X-Unsubscribes-To: listserv AT delorie DOT com
Precedence: bulk


On Fri, 13 Dec 2002, Eli Zaretskii wrote:
>
> > From: "Arthur J. O'Dwyer" <ajo AT andrew DOT cmu DOT edu>
> > Newsgroups: comp.os.msdos.djgpp
> > Date: Thu, 12 Dec 2002 22:29:02 -0500 (EST)
> >
> > Notice that the "globbing" doesn't complain about mismatched quotes,
> > and makes some argument values impossible to construct.
>
> Examples, please!  Some cases of mismatched quotes are actually the
> shell's fault.  In any case, DJGPP behaves like DOS/Windows programs:
> they don't generally complain about mismatched quotes.

True.  On the other hand, DJGPP behaves like Unix programs: they
generally perform filename globbing.  So when I want a literal * or ?
in my argument list, I have to use quotes.  In DOS, I hardly ever
use quotes, because they're not needed except for including whitespace
in an argument (which is rarely useful, IME).  So when I use quotes,
I am usually in my mental "Unix mode", and I expect the quotes to
behave as closely to Unix as possible.

Wht do you mean, "Some cases of mismatched quotes are actually the
shell's fault"?  How much globbing (by which I mean generic "argument
parsing") is done by the shell (by which I mean COMMAND.COM)?  My
experience from assembler programming is that COMMAND.COM doesn't
do any parsing of arguments at all; it's entirely up to the client.

> > Hence, I suggest that it is broken by design.
>
> I think you are mistaken, but please feel free to explain why you
> think it's broken.  Specific examples will probably help.

My original example was (essentially) trying to get the following test to
execute successfully:

int main (int argc, char **argv)
{
    if (strcmp("*\\", argv[1])==0 && strcmp("hello", argv[2])==0) {
       printf("Good!\n");
    }
}

You need to quote the * character, but you can't quote the \ character
because it will try to escape the closing quote, producing *" or *'.
A. Sinan Unur points out that

% test.exe "*"\ hello

works, but it looks like an incredibly flimsy hack to me, and I
wouldn't be surprised if it stopped working in future versions of
DJGPP (because the next guy who looks at the globbing code might not
know this hack is out there).


What I really want is to design a 'tr'-like program which takes as
arguments a list of hyphen-options, followed by one or two regexes,
followed optionally by a (list of) filename(s).

  tr  [ -cds ]  string1 [ string2 ] [ filename ]

Inside the regexes, I'd like \ to be an escape character, for example,
to specify the deletion of periods, question marks, and backslashes from
all text files in the current directory one could write:

  tr  -d  .?\\  *.txt

where the first \ escapes the second \.
Unfortunately, without quotes the first argument globs to a list of
directory names, with quotes it globs to .?\" or .?\', and if I turn
off globbing altogether then the second argument does not expand.

In bash or ksh (as the nice folks in comp.unix.misc have informed
me), the following command line does what I want:

  tr  -d  '.?\\'  *.txt

With DJGPP in MS-DOS, the best I can do is

  tr  -d  ".?"\  *.txt

and I've already explained that this just looks bad.

Thanks,
-Arthur