Date: Fri, 25 Jan 2002 13:27:39 -0500 From: Mike Barcroft <mike@FreeBSD.ORG> To: "Tim J. Robbins" <tim@robbins.dropbear.id.au> Cc: freebsd-standards@FreeBSD.ORG Subject: Re: split(1) -a option patch Message-ID: <20020125132739.B92720@espresso.q9media.com> In-Reply-To: <20020124210956.A13091@descent.robbins.dropbear.id.au>; from tim@robbins.dropbear.id.au on Thu, Jan 24, 2002 at 09:09:56PM %2B1100 References: <20020124210956.A13091@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:
> This patch implements -a option of split command to control the number
> of letters used to generate filename suffix.
Could you please update your patch to -CURRENT sources. There is
atleast one change to split.c that hasn't been MFC'd.
> diff -ru split.old/split.1 split/split.1
> --- split.old/split.1 Thu Jan 24 21:07:03 2002
> +++ split/split.1 Thu Jan 24 21:06:29 2002
> @@ -40,6 +40,7 @@
> .Nd split a file into pieces
> .Sh SYNOPSIS
> .Nm
> +.Op Fl a Ar suffix_length
> .Op Fl b Ar byte_count[k|m]
> .Op Fl l Ar line_count
> .Op Fl p Ar pattern
> @@ -54,6 +55,12 @@
> .Pp
> The options are as follows:
> .Bl -tag -width Ds
> +.It Fl a
> +Use
> +.Ar suffix_length
> +letters to form the suffix of the file name. If
> +.Fl a
> +is not specified, two letters are used as the suffix.
> .It Fl b
> Create smaller files
> .Ar byte_count
Okay.
> diff -ru split.old/split.c split/split.c
> --- split.old/split.c Thu Jan 24 21:07:03 2002
> +++ split/split.c Thu Jan 24 21:05:07 2002
> @@ -69,6 +69,7 @@
> char fname[MAXPATHLEN]; /* File name prefix. */
> regex_t rgx;
> int pflag;
> +int sufflen; /* File name suffix length. */
>
> void newfile __P((void));
> void split1 __P((void));
> @@ -83,7 +84,7 @@
> int ch;
> char *ep, *p;
>
> - while ((ch = getopt(argc, argv, "-0123456789b:l:p:")) != -1)
> + while ((ch = getopt(argc, argv, "-0123456789b:l:p:a:")) != -1)
These options need to be kept in alphabetical order.
> switch (ch) {
> case '0': case '1': case '2': case '3': case '4':
> case '5': case '6': case '7': case '8': case '9':
> @@ -130,6 +131,11 @@
> errx(EX_USAGE,
> "%s: illegal line count", optarg);
> break;
> + case 'a': /* Suffix length */
Similarly, the case clauses need to be kept in alphabetical order.
> + if ((sufflen = strtol(optarg, &ep, 10)) <= 0 || *ep)
Return value of strtol() doesn't match type of sufflen.
> + err(EX_USAGE,
> + "%s: illegal suffix length", optarg);
> + break;
> default:
> usage();
> }
> @@ -147,6 +153,9 @@
> if (*argv != NULL)
> usage();
>
> + if (strlen(fname) + sufflen >= sizeof (fname))
There shouldn't be a space between `sizeof' and `(fname)'.
This comparison will probably need some casting to prevent comparing
signed and unsigned diagnostics.
> + err(EX_USAGE, "suffix is too long");
> +
Gratuitous vertical whitespace.
> if (pflag && (numlines != 0 || bytecnt != 0))
> usage();
>
> @@ -277,6 +286,7 @@
> static long fnum;
> static int defname;
> static char *fpnt;
> + long maxfiles, i, tfnum;
This belongs above the other types, since `long' is larger than the
other types. Also, `i' comes before `m' in the alphabet.
>
> if (ofd == -1) {
> if (fname[0] == '\0') {
> @@ -289,19 +299,24 @@
> }
> ofd = fileno(stdout);
> }
> +
> + /* maxfiles = 26^sufflen, but don't use libm */
Sentences end with a period.
> + for (maxfiles = 1, i = 0; i < sufflen; i++)
> + maxfiles *= 26;
> +
Extra checking is needed here, since a 32-bit signed integer overflows
at a sufflen of 7.
> /*
> * Hack to increase max files; original code wandered through
> - * magic characters. Maximum files is 3 * 26 * 26 == 2028
> + * magic characters.
> */
> -#define MAXFILES 676
> - if (fnum == MAXFILES) {
> + if (fnum == maxfiles) {
> if (!defname || fname[0] == 'z')
> errx(EX_DATAERR, "too many files");
> ++fname[0];
> fnum = 0;
> }
> - fpnt[0] = fnum / 26 + 'a';
> - fpnt[1] = fnum % 26 + 'a';
> + for (tfnum = fnum, i = sufflen; i != 0; i--, tfnum /= 26)
> + fpnt[i - 1] = tfnum % 26 + 'a';
> + fpnt[sufflen] = '\0';
Can you explain the purpose of this change? I don't think I
completely understand it.
> ++fnum;
> if (!freopen(fname, "w", stdout))
> err(EX_IOERR, "%s", fname);
> @@ -312,6 +327,7 @@
> usage()
> {
> (void)fprintf(stderr,
> -"usage: split [-b byte_count] [-l line_count] [-p pattern] [file [prefix]]\n");
> +"usage: split [-a sufflen] [-b byte_count] [-l line_count] [-p pattern] \n"
> +"\t\t[file [prefix]]\n");
This change breaks K&R. Add a new fprintf().
> exit(EX_USAGE);
> }
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?20020125132739.B92720>