Date: Sun, 3 Feb 2002 17:45:11 -0500 From: Mike Barcroft <mike@FreeBSD.ORG> To: "Tim J. Robbins" <tim@robbins.dropbear.id.au> Cc: freebsd-standards@FreeBSD.ORG Subject: Re: pwd -L option Message-ID: <20020203174511.A6496@espresso.q9media.com> In-Reply-To: <20020202210237.A2326@descent.robbins.dropbear.id.au>; from tim@robbins.dropbear.id.au on Sat, Feb 02, 2002 at 09:02:37PM %2B1100 References: <20020130181638.A8510@descent.robbins.dropbear.id.au> <20020202032917.K10222@espresso.q9media.com> <20020202210237.A2326@descent.robbins.dropbear.id.au>
next in thread | previous in thread | raw e-mail | index | archive | help
Tim J. Robbins <tim@robbins.dropbear.id.au> writes:
> Here's a new patch that I hope solves the problems you found with my
> previous one. I have just a few comments, though...
Almost... :)
> On Sat, Feb 02, 2002 at 03:29:17AM -0500, Mike Barcroft wrote:
>
> > Style bug: ANSI-style prototypes. This is a problem because the rest
> > of the file uses K&R-style.
>
> usage()'s prototype is also incorrect, then. It is:
> void usage(void);
>
> It used to have a __P() around its argument list in 4.4BSD. Even then,
> the 1st ed. K&R book says nothing about "void", which is usage()'s return
> type. Is the so-called K&R style that is supposed to be maintained
> documented anywhere?
Oops, sorry. As Garance pointed out, Warner has apparently already
"fixed" pwd(1) to use ANSI-style prototypes.
> Index: pwd/pwd.1
> ===================================================================
> RCS file: /home/ncvs/src/bin/pwd/pwd.1,v
> retrieving revision 1.15
> diff -u -r1.15 pwd.1
> --- pwd/pwd.1 2001/08/15 09:09:36 1.15
> +++ pwd/pwd.1 2002/02/02 10:02:31
> @@ -43,6 +43,7 @@
> .Nd return working directory name
> .Sh SYNOPSIS
> .Nm
> +.Op Fl LP
This should be `L | P'. Similarly, usage() requires this.
> .Sh DESCRIPTION
> .Nm Pwd
> writes the absolute pathname of the current working directory to
> @@ -54,17 +55,35 @@
> Consult the
> .Xr builtin 1
> manual page.
> +.Pp
> +The options are as follows:
> +.Bl -tag -width indent
> +.It Fl L
> +Display the logical current working directory.
> +.It Fl P
> +Display the physical current working directory (all symbolic links resolved).
> +This is the default.
> +.Pp
> +.El
> +Only one of
> +.Fl L
> +and
> +.Fl P
> +may be specified.
This is redundant; the usage will now explain this.
> +.Sh ENVIRONMENT
> +Environment variables used by
> +.Nm :
> +.Bl -tag -width PWD
> +.It Ev PWD
> +Logical current working directory.
Missing `.El'. It looks like it accidentally ended up above.
> .Sh DIAGNOSTICS
> .Ex -std
> .Sh STANDARDS
> The
> .Nm
> utility is expected to be
> -.St -p1003.2
> +.St -p1003.1-2001
> compatible.
Let's be assertive here and say that this utility does conform.
> -The
> -.Fl L
> -flag is not supported.
> .Sh SEE ALSO
> .Xr builtin 1 ,
> .Xr cd 1 ,
> @@ -80,3 +99,9 @@
> However, it can give a different answer in the rare case
> that the current directory or a containing directory was moved after
> the shell descended into it.
> +.Pp
> +The
> +.Fl L
> +option does not work unless the
> +.Ev PWD
> +variable is exported by the shell.
> Index: pwd/pwd.c
> ===================================================================
> RCS file: /home/ncvs/src/bin/pwd/pwd.c,v
> retrieving revision 1.14
> diff -u -r1.14 pwd.c
> --- pwd/pwd.c 2002/02/02 06:48:10 1.14
> +++ pwd/pwd.c 2002/02/02 10:02:31
> @@ -45,31 +45,38 @@
> "$FreeBSD: src/bin/pwd/pwd.c,v 1.14 2002/02/02 06:48:10 imp Exp $";
> #endif /* not lint */
>
> +#include <sys/types.h>
> +#include <sys/stat.h>
> +
> #include <err.h>
> +#include <errno.h>
> #include <limits.h>
> #include <stdio.h>
> #include <stdlib.h>
> #include <unistd.h>
> #include <sys/param.h>
>
> +extern char *__progname;
> +
> +static char *getcwd_logical(void);
> void usage(void);
>
> int
> main(int argc, char *argv[])
> {
> + int Lflag, Pflag;
> int ch;
> char *p;
> char buf[PATH_MAX];
>
> - /*
> - * Flags for pwd are a bit strange. The POSIX 1003.2B/D9 document
> - * has an optional -P flag for physical, which is what this program
> - * will produce by default. The logical flag, -L, should fail, as
> - * there's no way to display a logical path after forking.
> - */
> - while ((ch = getopt(argc, argv, "P")) != -1)
> + Lflag = Pflag = 0;
> + while ((ch = getopt(argc, argv, "LP")) != -1)
This isn't quite right yet. realpath(1) should never call getopt().
It expects no options, so `realpath -L' should return a valid path
ending with `/-L'.
> switch (ch) {
> + case 'L':
> + ++Lflag;
I prefer `Lflag = 1;'.
> + break;
> case 'P':
> + ++Pflag;
Similarly, here.
> break;
> case '?':
> default:
> @@ -78,27 +85,52 @@
> argc -= optind;
> argv += optind;
>
> - if (argc == 1) {
> + if (strcmp(__progname, "realpath") == 0) {
> + if (argc != 1 || Lflag || Pflag)
> + usage();
See above.
> p = realpath(argv[0], buf);
> if (p == NULL)
> err(1, "%s", argv[0]);
> (void)printf("%s\n", p);
> - } else if (argc == 0) {
> - p = getcwd(NULL, (size_t)0);
> + } else {
> + if (argc != 0 || (Lflag && Pflag))
> + usage();
> + p = Lflag ? getcwd_logical() : getcwd(NULL, 0);
> if (p == NULL)
> err(1, ".");
> (void)printf("%s\n", p);
> - } else {
> - usage();
> }
>
> exit(0);
> }
>
> void
> -usage(void)
> +usage()
Undo this. See comment above. (Sorry for the confusion.)
> +{
Style bug: additional vertical space needed here.
> + if (strcmp (__progname, "realpath") == 0)
^^^ Style bug: unneeded space.
> + (void)fprintf(stderr, "usage: realpath [file ...]\n");
> + else
> + (void)fprintf(stderr, "usage: pwd [-LP]\n");
See above.
> + exit(1);
> +}
> +
> +static char *
> +getcwd_logical()
> {
> + struct stat log, phy;
> + char *pwd;
> +
> + /*
> + * Check that $PWD is an absolute logical pathname referring to
> + * the current working directory.
> + */
> + if ((pwd = getenv("PWD")) != NULL && *pwd == '/') {
> + if (stat(pwd, &log) == -1 || stat(".", &phy) == -1)
> + return (NULL);
> + if (log.st_dev == phy.st_dev && log.st_ino == phy.st_ino)
> + return (pwd);
> + }
Much better.
>
> - (void)fprintf(stderr, "usage: pwd\n");
> - exit(1);
> + errno = ENOENT;
> + return (NULL);
> }
I should be able to commit this after you get these problems fixed.
Best regards,
Mike Barcroft
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-standards" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20020203174511.A6496>