/* split.c -- split a file into pieces.
- Copyright (C) 1988, 1991 Free Software Foundation, Inc.
+ Copyright (C) 88, 91, 1995-2001 Free Software Foundation, Inc.
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
- along with this program; if not, write to the Free Software
- Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. */
+ along with this program; if not, write to the Free Software Foundation,
+ Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */
\f
/* By tege@sics.se, with rms.
* Implement -t CHAR or -t REGEX to specify break characters other
than newline. */
+#include <config.h>
+
#include <stdio.h>
#include <getopt.h>
-#include <ctype.h>
#include <sys/types.h>
+
#include "system.h"
+#include "closeout.h"
+#include "error.h"
+#include "full-write.h"
+#include "safe-read.h"
+#include "xstrtol.h"
-char *xmalloc ();
-void error ();
+/* The official name of this program (e.g., no `g' prefix). */
+#define PROGRAM_NAME "split"
-static int convint ();
-static int isdigits ();
-static int stdread ();
-static void line_bytes_split ();
-static void bytes_split ();
-static void cwrite ();
-static void lines_split ();
-static void next_file_name ();
+#define AUTHORS N_ ("Torbjorn Granlund and Richard M. Stallman")
-/* Name under which this program was invoked. */
+/* The name this program was run with. */
char *program_name;
/* Base name of output files. */
/* Pointer to the end of OUTFILE. */
static char *outfile_end;
-/* Status for outfile name generation. */
-static unsigned outfile_count = -1;
-static unsigned outfile_name_limit = 25 * 26;
-static unsigned outfile_name_generation = 1;
-
/* Name of input file. May be "-". */
static char *infile;
/* Descriptor on which output file is open. */
static int output_desc;
-\f
-static void
-usage (reason)
- char *reason;
-{
- if (reason != NULL)
- fprintf (stderr, "%s: %s\n", program_name, reason);
- fprintf (stderr, "\
-Usage: %s [-lines] [-l lines] [-b bytes[bkm]] [-C bytes[bkm]]\n\
- [--lines=lines] [--bytes=bytes[bkm]] [--line-bytes=bytes[bkm]]\n\
- [infile [outfile-prefix]]\n",
- program_name);
- exit (2);
-}
-\f
+
+/* If nonzero, print a diagnostic on standard error just before each
+ output file is opened. */
+static int verbose;
+
static struct option const longopts[] =
{
{"bytes", required_argument, NULL, 'b'},
{"lines", required_argument, NULL, 'l'},
{"line-bytes", required_argument, NULL, 'C'},
+ {"verbose", no_argument, NULL, 2},
+ {GETOPT_HELP_OPTION_DECL},
+ {GETOPT_VERSION_OPTION_DECL},
{NULL, 0, NULL, 0}
};
void
-main (argc, argv)
- int argc;
- char *argv[];
+usage (int status)
{
- struct stat stat_buf;
- int num; /* numeric argument from command line */
- enum
- {
- type_undef, type_bytes, type_byteslines, type_lines, type_digits
- } split_type = type_undef;
- int in_blk_size; /* optimal block size of input file device */
- char *buf; /* file i/o buffer */
- int accum = 0;
- char *outbase;
- int c;
- int digits_optind = 0;
-
- program_name = argv[0];
-
- /* Parse command line options. */
-
- infile = "-";
- outbase = "x";
-
- while (1)
- {
- /* This is the argv-index of the option we will read next. */
- int this_optind = optind ? optind : 1;
-
- c = getopt_long (argc, argv, "0123456789b:l:C:", longopts, (int *) 0);
- if (c == EOF)
- break;
-
- switch (c)
- {
- case 'b':
- if (split_type != type_undef)
- usage ("cannot split in more than one way");
- split_type = type_bytes;
- if (convint (optarg, &accum) == -1)
- usage ("invalid number of bytes");
- break;
-
- case 'l':
- if (split_type != type_undef)
- usage ("cannot split in more than one way");
- split_type = type_lines;
- if (!isdigits (optarg))
- usage ("invalid number of lines");
- accum = atoi (optarg);
- break;
-
- case 'C':
- if (split_type != type_undef)
- usage ("cannot split in more than one way");
- split_type = type_byteslines;
- if (convint (optarg, &accum) == -1)
- usage ("invalid number of bytes");
- break;
-
- case '0':
- case '1':
- case '2':
- case '3':
- case '4':
- case '5':
- case '6':
- case '7':
- case '8':
- case '9':
- if (split_type != type_undef && split_type != type_digits)
- usage ("cannot split in more than one way");
- if (digits_optind != 0 && digits_optind != this_optind)
- accum = 0; /* More than one number given; ignore other. */
- digits_optind = this_optind;
- split_type = type_digits;
- accum = accum * 10 + c - '0';
- break;
-
- default:
- usage ((char *)0);
- }
- }
-
- /* Handle default case. */
- if (split_type == type_undef)
+ if (status != 0)
+ fprintf (stderr, _("Try `%s --help' for more information.\n"),
+ program_name);
+ else
{
- split_type = type_lines;
- accum = 1000;
+ printf (_("\
+Usage: %s [OPTION] [INPUT [PREFIX]]\n\
+"),
+ program_name);
+ printf (_("\
+Output fixed-size pieces of INPUT to PREFIXaa, PREFIXab, ...; default\n\
+PREFIX is `x'. With no INPUT, or when INPUT is -, read standard input.\n\
+\n\
+Mandatory arguments to long options are mandatory for short options too.\n\
+ -b, --bytes=SIZE put SIZE bytes per output file\n\
+ -C, --line-bytes=SIZE put at most SIZE bytes of lines per output file\n\
+ -l, --lines=NUMBER put NUMBER lines per output file\n\
+ -NUMBER same as -l NUMBER\n\
+ --verbose print a diagnostic to standard error just\n\
+ before each output file is opened\n\
+ --help display this help and exit\n\
+ --version output version information and exit\n\
+\n\
+SIZE may have a multiplier suffix: b for 512, k for 1K, m for 1 Meg.\n\
+"));
+ puts (_("\nReport bugs to <bug-textutils@gnu.org>."));
}
+ exit (status == 0 ? EXIT_SUCCESS : EXIT_FAILURE);
+}
- if (accum < 1)
- usage ("invalid number");
- num = accum;
-
- /* Get out the filename arguments. */
-
- if (optind < argc)
- infile = argv[optind++];
-
- if (optind < argc)
- outbase = argv[optind++];
+/* Compute the next sequential output file name suffix and store it
+ into the string `outfile' at the position pointed to by `outfile_mid'. */
- if (optind < argc)
- usage ("too many arguments");
+static void
+next_file_name (void)
+{
+ static unsigned n_digits = 2;
+ char *p;
- /* Open the input file. */
- if (!strcmp (infile, "-"))
- input_desc = 0;
- else
+ /* Change any suffix of `z's to `a's. */
+ for (p = outfile_end - 1; *p == 'z'; p--)
{
- input_desc = open (infile, O_RDONLY);
- if (input_desc < 0)
- error (1, errno, "%s", infile);
+ *p = 'a';
}
- /* No output file is open now. */
- output_desc = -1;
-
- /* Copy the output file prefix so we can add suffixes to it.
- 26**29 is certainly enough output files! */
+ /* Increment the rightmost non-`z' character that was present before the
+ above z/a substitutions. There is guaranteed to be such a character. */
+ ++(*p);
- outfile = xmalloc (strlen (outbase) + 30);
- strcpy (outfile, outbase);
- outfile_mid = outfile + strlen (outfile);
- outfile_end = outfile_mid + 2;
- bzero (outfile_mid, 30);
- outfile_mid[0] = 'a';
- outfile_mid[1] = 'a' - 1; /* first call to next_file_name makes it an 'a' */
-
- /* Get the optimal block size of input device and make a buffer. */
-
- if (fstat (input_desc, &stat_buf) < 0)
- error (1, errno, "%s", infile);
- in_blk_size = ST_BLKSIZE (stat_buf);
-
- buf = xmalloc (in_blk_size + 1);
-
- switch (split_type)
+ /* If the result of that increment operation yielded a `z' and there
+ are only `z's to the left of it, then append two more `a' characters
+ to the end and add 1 (-1 + 2) to the number of digits (we're taking
+ out this `z' and adding two `a's). */
+ if (*p == 'z' && p == outfile_mid)
{
- case type_digits:
- case type_lines:
- lines_split (num, buf, in_blk_size);
- break;
-
- case type_bytes:
- bytes_split (num, buf, in_blk_size);
- break;
-
- case type_byteslines:
- line_bytes_split (num);
- break;
-
- default:
- abort ();
+ ++n_digits;
+ ++outfile_mid;
+ *outfile_end++ = 'a';
+ *outfile_end++ = 'a';
}
-
- if (close (input_desc) < 0)
- error (1, errno, "%s", infile);
- if (output_desc >= 0 && close (output_desc) < 0)
- error (1, errno, "%s", outfile);
-
- exit (0);
}
-/* Return nonzero if the string STR is composed entirely of decimal digits. */
+/* Write BYTES bytes at BP to an output file.
+ If NEW_FILE_FLAG is nonzero, open the next output file.
+ Otherwise add to the same output file already in use. */
-static int
-isdigits (str)
- char *str;
+static void
+cwrite (int new_file_flag, const char *bp, int bytes)
{
- do
+ if (new_file_flag)
{
- if (!isdigit (*str))
- return 0;
- str++;
+ if (output_desc >= 0 && close (output_desc) < 0)
+ error (EXIT_FAILURE, errno, "%s", outfile);
+
+ next_file_name ();
+ if (verbose)
+ fprintf (stderr, _("creating file `%s'\n"), outfile);
+ output_desc = open (outfile,
+ O_WRONLY | O_CREAT | O_TRUNC | O_BINARY, 0666);
+ if (output_desc < 0)
+ error (EXIT_FAILURE, errno, "%s", outfile);
}
- while (*str);
- return 1;
+ if (full_write (output_desc, bp, bytes) != bytes)
+ error (EXIT_FAILURE, errno, "%s", outfile);
}
-/* Put the value of the number in STR into *VAL.
- STR can specify a positive integer, optionally ending in `k'
- to mean kilo or `m' to mean mega.
- Return 0 if STR is valid, -1 if not. */
+/* Read NCHARS bytes from the input file into BUF.
+ Return the number of bytes successfully read.
+ If this is less than NCHARS, do not call `stdread' again. */
static int
-convint (str, val)
- char *str;
- int *val;
+stdread (char *buf, int nchars)
{
- int multiplier = 1;
- int arglen = strlen (str);
+ int n_read;
+ int to_be_read = nchars;
- if (arglen > 1)
+ while (to_be_read)
{
- switch (str[arglen - 1])
- {
- case 'b':
- multiplier = 512;
- str[arglen - 1] = '\0';
- break;
- case 'k':
- multiplier = 1024;
- str[arglen - 1] = '\0';
- break;
- case 'm':
- multiplier = 1048576;
- str[arglen - 1] = '\0';
- break;
- }
+ n_read = safe_read (input_desc, buf, to_be_read);
+ if (n_read < 0)
+ return -1;
+ if (n_read == 0)
+ break;
+ to_be_read -= n_read;
+ buf += n_read;
}
- if (!isdigits (str))
- return -1;
- *val = atoi (str) * multiplier;
- return 0;
+ return nchars - to_be_read;
}
-\f
+
/* Split into pieces of exactly NCHARS bytes.
Use buffer BUF, whose size is BUFSIZE. */
static void
-bytes_split (nchars, buf, bufsize)
- int nchars;
- char *buf;
- int bufsize;
+bytes_split (int nchars, char *buf, int bufsize)
{
int n_read;
int new_file_flag = 1;
{
n_read = stdread (buf, bufsize);
if (n_read < 0)
- error (1, errno, "%s", infile);
+ error (EXIT_FAILURE, errno, "%s", infile);
bp_out = buf;
to_read = n_read;
for (;;)
}
while (n_read == bufsize);
}
-\f
+
/* Split into pieces of exactly NLINES lines.
Use buffer BUF, whose size is BUFSIZE. */
static void
-lines_split (nlines, buf, bufsize)
- int nlines;
- char *buf;
- int bufsize;
+lines_split (int nlines, char *buf, int bufsize)
{
int n_read;
char *bp, *bp_out, *eob;
{
n_read = stdread (buf, bufsize);
if (n_read < 0)
- error (1, errno, "%s", infile);
+ error (EXIT_FAILURE, errno, "%s", infile);
bp = bp_out = buf;
eob = bp + n_read;
*eob = '\n';
where lines longer than NCHARS bytes occur. */
static void
-line_bytes_split (nchars)
- int nchars;
+line_bytes_split (int nchars)
{
int n_read;
char *bp;
n_read = stdread (buf + n_buffered, nchars - n_buffered);
if (n_read < 0)
- error (1, errno, "%s", infile);
+ error (EXIT_FAILURE, errno, "%s", infile);
n_buffered += n_read;
if (n_buffered != nchars)
cwrite (1, buf, bp - buf);
/* Discard the chars we just output; move rest of chunk
- down to be the start of the next chunk. */
+ down to be the start of the next chunk. Source and
+ destination probably overlap. */
n_buffered -= bp - buf;
if (n_buffered > 0)
- bcopy (bp, buf, n_buffered);
+ memmove (buf, bp, n_buffered);
}
while (!eof);
free (buf);
}
-\f
-/* Write BYTES bytes at BP to an output file.
- If NEW_FILE_FLAG is nonzero, open the next output file.
- Otherwise add to the same output file already in use. */
-static void
-cwrite (new_file_flag, bp, bytes)
- int new_file_flag;
- char *bp;
- int bytes;
+int
+main (int argc, char **argv)
{
- if (new_file_flag)
+ struct stat stat_buf;
+ int num; /* numeric argument from command line */
+ enum
{
- if (output_desc >= 0 && close (output_desc) < 0)
- error (1, errno, "%s", outfile);
+ type_undef, type_bytes, type_byteslines, type_lines, type_digits
+ } split_type = type_undef;
+ int in_blk_size; /* optimal block size of input file device */
+ char *buf; /* file i/o buffer */
+ int accum = 0;
+ char *outbase;
+ int c;
+ int digits_optind = 0;
- next_file_name ();
- output_desc = open (outfile, O_WRONLY | O_CREAT | O_TRUNC, 0666);
- if (output_desc < 0)
- error (1, errno, "%s", outfile);
- }
- if (write (output_desc, bp, bytes) < 0)
- error (1, errno, "%s", outfile);
-}
+ program_name = argv[0];
+ setlocale (LC_ALL, "");
+ bindtextdomain (PACKAGE, LOCALEDIR);
+ textdomain (PACKAGE);
-/* Read NCHARS bytes from the input file into BUF.
- Return the number of bytes successfully read.
- If this is less than NCHARS, do not call `stdread' again. */
+ atexit (close_stdout);
-static int
-stdread (buf, nchars)
- char *buf;
- int nchars;
-{
- int n_read;
- int to_be_read = nchars;
+ /* Parse command line options. */
- while (to_be_read)
+ infile = "-";
+ outbase = "x";
+
+ while (1)
{
- n_read = read (input_desc, buf, to_be_read);
- if (n_read < 0)
- return -1;
- if (n_read == 0)
+ /* This is the argv-index of the option we will read next. */
+ int this_optind = optind ? optind : 1;
+ long int tmp_long;
+
+ c = getopt_long (argc, argv, "0123456789vb:l:C:", longopts, (int *) 0);
+ if (c == EOF)
break;
- to_be_read -= n_read;
- buf += n_read;
+
+ switch (c)
+ {
+ case 0:
+ break;
+
+ case 'b':
+ if (split_type != type_undef)
+ {
+ error (0, 0, _("cannot split in more than one way"));
+ usage (EXIT_FAILURE);
+ }
+ split_type = type_bytes;
+ if (xstrtol (optarg, NULL, 10, &tmp_long, "bkm") != LONGINT_OK
+ || tmp_long < 0 || tmp_long > INT_MAX)
+ {
+ error (0, 0, _("%s: invalid number of bytes"), optarg);
+ usage (EXIT_FAILURE);
+ }
+ accum = (int) tmp_long;
+ break;
+
+ case 'l':
+ if (split_type != type_undef)
+ {
+ error (0, 0, _("cannot split in more than one way"));
+ usage (EXIT_FAILURE);
+ }
+ split_type = type_lines;
+ if (xstrtol (optarg, NULL, 10, &tmp_long, "") != LONGINT_OK
+ || tmp_long < 0 || tmp_long > INT_MAX)
+ {
+ error (0, 0, _("%s: invalid number of lines"), optarg);
+ usage (EXIT_FAILURE);
+ }
+ accum = (int) tmp_long;
+ break;
+
+ case 'C':
+ if (split_type != type_undef)
+ {
+ error (0, 0, _("cannot split in more than one way"));
+ usage (EXIT_FAILURE);
+ }
+
+ split_type = type_byteslines;
+ if (xstrtol (optarg, NULL, 10, &tmp_long, "bkm") != LONGINT_OK
+ || tmp_long < 0 || tmp_long > INT_MAX)
+ {
+ error (0, 0, _("%s: invalid number of bytes"), optarg);
+ usage (EXIT_FAILURE);
+ }
+ accum = (int) tmp_long;
+ break;
+
+ case '0':
+ case '1':
+ case '2':
+ case '3':
+ case '4':
+ case '5':
+ case '6':
+ case '7':
+ case '8':
+ case '9':
+ if (split_type != type_undef && split_type != type_digits)
+ {
+ error (0, 0, _("cannot split in more than one way"));
+ usage (EXIT_FAILURE);
+ }
+ if (digits_optind != 0 && digits_optind != this_optind)
+ accum = 0; /* More than one number given; ignore other. */
+ digits_optind = this_optind;
+ split_type = type_digits;
+ accum = accum * 10 + c - '0';
+ break;
+
+ case 2:
+ verbose = 1;
+ break;
+
+ case_GETOPT_HELP_CHAR;
+
+ case_GETOPT_VERSION_CHAR (PROGRAM_NAME, AUTHORS);
+
+ default:
+ usage (EXIT_FAILURE);
+ }
}
- return nchars - to_be_read;
-}
-/* Compute the next sequential output file name suffix and store it
- into the string `outfile' at the position pointed to by `outfile_mid'. */
+ /* Handle default case. */
+ if (split_type == type_undef)
+ {
+ split_type = type_lines;
+ accum = 1000;
+ }
-static void
-next_file_name ()
-{
- int x;
- char *ne;
+ if (accum < 1)
+ {
+ error (0, 0, _("invalid number"));
+ usage (EXIT_FAILURE);
+ }
+ num = accum;
+
+ /* Get out the filename arguments. */
+
+ if (optind < argc)
+ infile = argv[optind++];
+
+ if (optind < argc)
+ outbase = argv[optind++];
- outfile_count++;
- if (outfile_count < outfile_name_limit)
+ if (optind < argc)
{
- for (ne = outfile_end - 1; ; ne--)
- {
- x = *ne;
- if (x != 'z')
- break;
- *ne = 'a';
- }
- *ne = x + 1;
- return;
+ error (0, 0, _("too many arguments"));
+ usage (EXIT_FAILURE);
}
- outfile_count = 0;
- outfile_name_limit *= 26;
- outfile_name_generation++;
- *outfile_mid++ = 'z';
- for (x = 0; x <= outfile_name_generation; x++)
- outfile_mid[x] = 'a';
- outfile_end += 2;
+ /* Open the input file. */
+ if (STREQ (infile, "-"))
+ input_desc = 0;
+ else
+ {
+ input_desc = open (infile, O_RDONLY);
+ if (input_desc < 0)
+ error (EXIT_FAILURE, errno, "%s", infile);
+ }
+ /* Binary I/O is safer when bytecounts are used. */
+ SET_BINARY (input_desc);
+
+ /* No output file is open now. */
+ output_desc = -1;
+
+ /* Copy the output file prefix so we can add suffixes to it.
+ 26**29 is certainly enough output files! */
+
+ outfile = xmalloc (strlen (outbase) + 30);
+ strcpy (outfile, outbase);
+ outfile_mid = outfile + strlen (outfile);
+ outfile_end = outfile_mid + 2;
+ memset (outfile_mid, 0, 30);
+ outfile_mid[0] = 'a';
+ outfile_mid[1] = 'a' - 1; /* first call to next_file_name makes it an 'a' */
+
+ /* Get the optimal block size of input device and make a buffer. */
+
+ if (fstat (input_desc, &stat_buf) < 0)
+ error (EXIT_FAILURE, errno, "%s", infile);
+ in_blk_size = ST_BLKSIZE (stat_buf);
+
+ buf = xmalloc (in_blk_size + 1);
+
+ switch (split_type)
+ {
+ case type_digits:
+ case type_lines:
+ lines_split (num, buf, in_blk_size);
+ break;
+
+ case type_bytes:
+ bytes_split (num, buf, in_blk_size);
+ break;
+
+ case type_byteslines:
+ line_bytes_split (num);
+ break;
+
+ default:
+ abort ();
+ }
+
+ if (close (input_desc) < 0)
+ error (EXIT_FAILURE, errno, "%s", infile);
+ if (output_desc >= 0 && close (output_desc) < 0)
+ error (EXIT_FAILURE, errno, "%s", outfile);
+
+ exit (EXIT_SUCCESS);
}