1 /* split.c -- split a file into pieces.
2 Copyright (C) 88, 91, 1995-2002 Free Software Foundation, Inc.
4 This program is free software; you can redistribute it and/or modify
5 it under the terms of the GNU General Public License as published by
6 the Free Software Foundation; either version 2, or (at your option)
9 This program is distributed in the hope that it will be useful,
10 but WITHOUT ANY WARRANTY; without even the implied warranty of
11 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
12 GNU General Public License for more details.
14 You should have received a copy of the GNU General Public License
15 along with this program; if not, write to the Free Software Foundation,
16 Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */
18 /* By tege@sics.se, with rms.
21 * Implement -t CHAR or -t REGEX to specify break characters other
28 #include <sys/types.h>
34 #include "full-write.h"
36 #include "safe-read.h"
39 /* The official name of this program (e.g., no `g' prefix). */
40 #define PROGRAM_NAME "split"
42 #define AUTHORS N_ ("Torbjorn Granlund and Richard M. Stallman")
44 #define DEFAULT_SUFFIX_LENGTH 2
46 /* The name this program was run with. */
49 /* Base name of output files. */
50 static char const *outbase;
52 /* Name of output files. */
55 /* Pointer to the end of the prefix in OUTFILE.
56 Suffixes are inserted here. */
57 static char *outfile_mid;
59 /* Length of OUTFILE's suffix. */
60 static size_t suffix_length = DEFAULT_SUFFIX_LENGTH;
62 /* Name of input file. May be "-". */
65 /* Descriptor on which input file is open. */
66 static int input_desc;
68 /* Descriptor on which output file is open. */
69 static int output_desc;
71 /* If nonzero, print a diagnostic on standard error just before each
72 output file is opened. */
75 static struct option const longopts[] =
77 {"bytes", required_argument, NULL, 'b'},
78 {"lines", required_argument, NULL, 'l'},
79 {"line-bytes", required_argument, NULL, 'C'},
80 {"suffix-length", required_argument, NULL, 'a'},
81 {"verbose", no_argument, NULL, 2},
82 {GETOPT_HELP_OPTION_DECL},
83 {GETOPT_VERSION_OPTION_DECL},
91 fprintf (stderr, _("Try `%s --help' for more information.\n"),
96 Usage: %s [OPTION] [INPUT [PREFIX]]\n\
100 Output fixed-size pieces of INPUT to PREFIXaa, PREFIXab, ...; default\n\
101 PREFIX is `x'. With no INPUT, or when INPUT is -, read standard input.\n\
105 Mandatory arguments to long options are mandatory for short options too.\n\
107 fprintf (stdout, _("\
108 -a, --suffix-length=N use suffixes of length N (default %d)\n\
109 -b, --bytes=SIZE put SIZE bytes per output file\n\
110 -C, --line-bytes=SIZE put at most SIZE bytes of lines per output file\n\
111 -l, --lines=NUMBER put NUMBER lines per output file\n\
112 "), DEFAULT_SUFFIX_LENGTH);
114 --verbose print a diagnostic to standard error just\n\
115 before each output file is opened\n\
117 fputs (HELP_OPTION_DESCRIPTION, stdout);
118 fputs (VERSION_OPTION_DESCRIPTION, stdout);
121 SIZE may have a multiplier suffix: b for 512, k for 1K, m for 1 Meg.\n\
123 printf (_("\nReport bugs to <%s>.\n"), PACKAGE_BUGREPORT);
125 exit (status == 0 ? EXIT_SUCCESS : EXIT_FAILURE);
128 /* Compute the next sequential output file name and store it into the
132 next_file_name (void)
136 /* Allocate and initialize the first file name. */
138 size_t outbase_length = strlen (outbase);
139 size_t outfile_length = outbase_length + suffix_length;
140 if (outfile_length + 1 < outbase_length)
142 outfile = xmalloc (outfile_length + 1);
143 outfile_mid = outfile + outbase_length;
144 memcpy (outfile, outbase, outbase_length);
145 memset (outfile_mid, 'a', suffix_length);
146 outfile[outfile_length] = 0;
148 #if ! _POSIX_NO_TRUNC && HAVE_PATHCONF && defined _PC_NAME_MAX
149 /* POSIX requires that if the output file name is too long for
150 its directory, `split' must fail without creating any files.
151 This must be checked for explicitly on operating systems that
152 silently truncate file names. */
154 char *dir = dir_name (outfile);
155 long name_max = pathconf (dir, _PC_NAME_MAX);
156 if (0 <= name_max && name_max < base_len (base_name (outfile)))
157 error (EXIT_FAILURE, ENAMETOOLONG, "%s", outfile);
164 /* Increment the suffix in place, if possible. */
167 for (p = outfile_mid + suffix_length; outfile_mid < p; *--p = 'a')
170 error (EXIT_FAILURE, 0, _("Output file suffixes exhausted"));
174 /* Write BYTES bytes at BP to an output file.
175 If NEW_FILE_FLAG is nonzero, open the next output file.
176 Otherwise add to the same output file already in use. */
179 cwrite (int new_file_flag, const char *bp, int bytes)
183 if (output_desc >= 0 && close (output_desc) < 0)
184 error (EXIT_FAILURE, errno, "%s", outfile);
188 fprintf (stderr, _("creating file `%s'\n"), outfile);
189 output_desc = open (outfile,
190 O_WRONLY | O_CREAT | O_TRUNC | O_BINARY, 0666);
192 error (EXIT_FAILURE, errno, "%s", outfile);
194 if (full_write (output_desc, bp, bytes) != (size_t) bytes)
195 error (EXIT_FAILURE, errno, "%s", outfile);
198 /* Read NCHARS bytes from the input file into BUF.
199 Return the number of bytes successfully read.
200 If this is less than NCHARS, do not call `stdread' again. */
203 stdread (char *buf, int nchars)
206 int to_be_read = nchars;
210 n_read = safe_read (input_desc, buf, to_be_read);
215 to_be_read -= n_read;
218 return nchars - to_be_read;
221 /* Split into pieces of exactly NCHARS bytes.
222 Use buffer BUF, whose size is BUFSIZE. */
225 bytes_split (int nchars, char *buf, int bufsize)
228 int new_file_flag = 1;
230 int to_write = nchars;
235 n_read = stdread (buf, bufsize);
237 error (EXIT_FAILURE, errno, "%s", infile);
242 if (to_read < to_write)
244 if (to_read) /* do not write 0 bytes! */
246 cwrite (new_file_flag, bp_out, to_read);
254 cwrite (new_file_flag, bp_out, to_write);
262 while (n_read == bufsize);
265 /* Split into pieces of exactly NLINES lines.
266 Use buffer BUF, whose size is BUFSIZE. */
269 lines_split (int nlines, char *buf, int bufsize)
272 char *bp, *bp_out, *eob;
273 int new_file_flag = 1;
278 n_read = stdread (buf, bufsize);
280 error (EXIT_FAILURE, errno, "%s", infile);
286 while (*bp++ != '\n')
287 ; /* this semicolon takes most of the time */
290 if (eob != bp_out) /* do not write 0 bytes! */
292 cwrite (new_file_flag, bp_out, eob - bp_out);
300 cwrite (new_file_flag, bp_out, bp - bp_out);
307 while (n_read == bufsize);
310 /* Split into pieces that are as large as possible while still not more
311 than NCHARS bytes, and are split on line boundaries except
312 where lines longer than NCHARS bytes occur. */
315 line_bytes_split (int nchars)
321 char *buf = (char *) xmalloc (nchars);
325 /* Fill up the full buffer size from the input file. */
327 n_read = stdread (buf + n_buffered, nchars - n_buffered);
329 error (EXIT_FAILURE, errno, "%s", infile);
331 n_buffered += n_read;
332 if (n_buffered != nchars)
335 /* Find where to end this chunk. */
336 bp = buf + n_buffered;
337 if (n_buffered == nchars)
339 while (bp > buf && bp[-1] != '\n')
343 /* If chunk has no newlines, use all the chunk. */
345 bp = buf + n_buffered;
347 /* Output the chars as one output file. */
348 cwrite (1, buf, bp - buf);
350 /* Discard the chars we just output; move rest of chunk
351 down to be the start of the next chunk. Source and
352 destination probably overlap. */
353 n_buffered -= bp - buf;
355 memmove (buf, bp, n_buffered);
361 #define FAIL_ONLY_ONE_WAY() \
364 error (0, 0, _("cannot split in more than one way")); \
365 usage (EXIT_FAILURE); \
370 main (int argc, char **argv)
372 struct stat stat_buf;
373 int num; /* numeric argument from command line */
376 type_undef, type_bytes, type_byteslines, type_lines, type_digits
377 } split_type = type_undef;
378 int in_blk_size; /* optimal block size of input file device */
379 char *buf; /* file i/o buffer */
382 int digits_optind = 0;
384 program_name = argv[0];
385 setlocale (LC_ALL, "");
386 bindtextdomain (PACKAGE, LOCALEDIR);
387 textdomain (PACKAGE);
389 atexit (close_stdout);
391 /* Parse command line options. */
398 /* This is the argv-index of the option we will read next. */
399 int this_optind = optind ? optind : 1;
402 c = getopt_long (argc, argv, "0123456789C:a:b:l:", longopts, NULL);
414 if (xstrtoul (optarg, NULL, 10, &tmp, "") != LONGINT_OK
417 error (0, 0, _("%s: invalid suffix length"), optarg);
418 usage (EXIT_FAILURE);
425 if (split_type != type_undef)
426 FAIL_ONLY_ONE_WAY ();
427 split_type = type_bytes;
428 if (xstrtol (optarg, NULL, 10, &tmp_long, "bkm") != LONGINT_OK
429 || tmp_long < 0 || tmp_long > INT_MAX)
431 error (0, 0, _("%s: invalid number of bytes"), optarg);
432 usage (EXIT_FAILURE);
434 accum = (int) tmp_long;
438 if (split_type != type_undef)
439 FAIL_ONLY_ONE_WAY ();
440 split_type = type_lines;
441 if (xstrtol (optarg, NULL, 10, &tmp_long, "") != LONGINT_OK
442 || tmp_long < 0 || tmp_long > INT_MAX)
444 error (0, 0, _("%s: invalid number of lines"), optarg);
445 usage (EXIT_FAILURE);
447 accum = (int) tmp_long;
451 if (split_type != type_undef)
452 FAIL_ONLY_ONE_WAY ();
453 split_type = type_byteslines;
454 if (xstrtol (optarg, NULL, 10, &tmp_long, "bkm") != LONGINT_OK
455 || tmp_long < 0 || tmp_long > INT_MAX)
457 error (0, 0, _("%s: invalid number of bytes"), optarg);
458 usage (EXIT_FAILURE);
460 accum = (int) tmp_long;
473 if (split_type != type_undef && split_type != type_digits)
474 FAIL_ONLY_ONE_WAY ();
475 if (digits_optind != 0 && digits_optind != this_optind)
476 accum = 0; /* More than one number given; ignore other. */
477 digits_optind = this_optind;
478 split_type = type_digits;
479 accum = accum * 10 + c - '0';
486 case_GETOPT_HELP_CHAR;
488 case_GETOPT_VERSION_CHAR (PROGRAM_NAME, AUTHORS);
491 usage (EXIT_FAILURE);
495 if (digits_optind && 200112 <= posix2_version ())
497 error (0, 0, _("`-%d' option is obsolete; use `-l %d'"), accum, accum);
498 usage (EXIT_FAILURE);
501 /* Handle default case. */
502 if (split_type == type_undef)
504 split_type = type_lines;
510 error (0, 0, _("invalid number"));
511 usage (EXIT_FAILURE);
515 /* Get out the filename arguments. */
518 infile = argv[optind++];
521 outbase = argv[optind++];
525 error (0, 0, _("too many arguments"));
526 usage (EXIT_FAILURE);
529 /* Open the input file. */
530 if (STREQ (infile, "-"))
534 input_desc = open (infile, O_RDONLY);
536 error (EXIT_FAILURE, errno, "%s", infile);
538 /* Binary I/O is safer when bytecounts are used. */
539 SET_BINARY (input_desc);
541 /* No output file is open now. */
544 /* Get the optimal block size of input device and make a buffer. */
546 if (fstat (input_desc, &stat_buf) < 0)
547 error (EXIT_FAILURE, errno, "%s", infile);
548 in_blk_size = ST_BLKSIZE (stat_buf);
550 buf = xmalloc (in_blk_size + 1);
556 lines_split (num, buf, in_blk_size);
560 bytes_split (num, buf, in_blk_size);
563 case type_byteslines:
564 line_bytes_split (num);
571 if (close (input_desc) < 0)
572 error (EXIT_FAILURE, errno, "%s", infile);
573 if (output_desc >= 0 && close (output_desc) < 0)
574 error (EXIT_FAILURE, errno, "%s", outfile);