1 /* split.c -- split a file into pieces.
2 Copyright (C) 88, 91, 1995-2001 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>
33 #include "full-write.h"
34 #include "safe-read.h"
37 /* The official name of this program (e.g., no `g' prefix). */
38 #define PROGRAM_NAME "split"
40 #define AUTHORS N_ ("Torbjorn Granlund and Richard M. Stallman")
42 /* The name this program was run with. */
45 /* Base name of output files. */
48 /* Pointer to the end of the prefix in OUTFILE.
49 Suffixes are inserted here. */
50 static char *outfile_mid;
52 /* Pointer to the end of OUTFILE. */
53 static char *outfile_end;
55 /* Name of input file. May be "-". */
58 /* Descriptor on which input file is open. */
59 static int input_desc;
61 /* Descriptor on which output file is open. */
62 static int output_desc;
64 /* If nonzero, print a diagnostic on standard error just before each
65 output file is opened. */
68 static struct option const longopts[] =
70 {"bytes", required_argument, NULL, 'b'},
71 {"lines", required_argument, NULL, 'l'},
72 {"line-bytes", required_argument, NULL, 'C'},
73 {"verbose", no_argument, NULL, 2},
74 {GETOPT_HELP_OPTION_DECL},
75 {GETOPT_VERSION_OPTION_DECL},
83 fprintf (stderr, _("Try `%s --help' for more information.\n"),
88 Usage: %s [OPTION] [INPUT [PREFIX]]\n\
92 Output fixed-size pieces of INPUT to PREFIXaa, PREFIXab, ...; default\n\
93 PREFIX is `x'. With no INPUT, or when INPUT is -, read standard input.\n\
95 -b, --bytes=SIZE put SIZE bytes per output file\n\
96 -C, --line-bytes=SIZE put at most SIZE bytes of lines per output file\n\
97 -l, --lines=NUMBER put NUMBER lines per output file\n\
98 -NUMBER same as -l NUMBER\n\
99 --verbose print a diagnostic to standard error just\n\
100 before each output file is opened\n\
101 --help display this help and exit\n\
102 --version output version information and exit\n\
104 SIZE may have a multiplier suffix: b for 512, k for 1K, m for 1 Meg.\n\
106 puts (_("\nReport bugs to <bug-textutils@gnu.org>."));
108 exit (status == 0 ? EXIT_SUCCESS : EXIT_FAILURE);
111 /* Compute the next sequential output file name suffix and store it
112 into the string `outfile' at the position pointed to by `outfile_mid'. */
115 next_file_name (void)
117 static unsigned n_digits = 2;
120 /* Change any suffix of `z's to `a's. */
121 for (p = outfile_end - 1; *p == 'z'; p--)
126 /* Increment the rightmost non-`z' character that was present before the
127 above z/a substitutions. There is guaranteed to be such a character. */
130 /* If the result of that increment operation yielded a `z' and there
131 are only `z's to the left of it, then append two more `a' characters
132 to the end and add 1 (-1 + 2) to the number of digits (we're taking
133 out this `z' and adding two `a's). */
134 if (*p == 'z' && p == outfile_mid)
138 *outfile_end++ = 'a';
139 *outfile_end++ = 'a';
143 /* Write BYTES bytes at BP to an output file.
144 If NEW_FILE_FLAG is nonzero, open the next output file.
145 Otherwise add to the same output file already in use. */
148 cwrite (int new_file_flag, const char *bp, int bytes)
152 if (output_desc >= 0 && close (output_desc) < 0)
153 error (EXIT_FAILURE, errno, "%s", outfile);
157 fprintf (stderr, _("creating file `%s'\n"), outfile);
158 output_desc = open (outfile,
159 O_WRONLY | O_CREAT | O_TRUNC | O_BINARY, 0666);
161 error (EXIT_FAILURE, errno, "%s", outfile);
163 if (full_write (output_desc, bp, bytes) != bytes)
164 error (EXIT_FAILURE, errno, "%s", outfile);
167 /* Read NCHARS bytes from the input file into BUF.
168 Return the number of bytes successfully read.
169 If this is less than NCHARS, do not call `stdread' again. */
172 stdread (char *buf, int nchars)
175 int to_be_read = nchars;
179 n_read = safe_read (input_desc, buf, to_be_read);
184 to_be_read -= n_read;
187 return nchars - to_be_read;
190 /* Split into pieces of exactly NCHARS bytes.
191 Use buffer BUF, whose size is BUFSIZE. */
194 bytes_split (int nchars, char *buf, int bufsize)
197 int new_file_flag = 1;
199 int to_write = nchars;
204 n_read = stdread (buf, bufsize);
206 error (EXIT_FAILURE, errno, "%s", infile);
211 if (to_read < to_write)
213 if (to_read) /* do not write 0 bytes! */
215 cwrite (new_file_flag, bp_out, to_read);
223 cwrite (new_file_flag, bp_out, to_write);
231 while (n_read == bufsize);
234 /* Split into pieces of exactly NLINES lines.
235 Use buffer BUF, whose size is BUFSIZE. */
238 lines_split (int nlines, char *buf, int bufsize)
241 char *bp, *bp_out, *eob;
242 int new_file_flag = 1;
247 n_read = stdread (buf, bufsize);
249 error (EXIT_FAILURE, errno, "%s", infile);
255 while (*bp++ != '\n')
256 ; /* this semicolon takes most of the time */
259 if (eob != bp_out) /* do not write 0 bytes! */
261 cwrite (new_file_flag, bp_out, eob - bp_out);
269 cwrite (new_file_flag, bp_out, bp - bp_out);
276 while (n_read == bufsize);
279 /* Split into pieces that are as large as possible while still not more
280 than NCHARS bytes, and are split on line boundaries except
281 where lines longer than NCHARS bytes occur. */
284 line_bytes_split (int nchars)
290 char *buf = (char *) xmalloc (nchars);
294 /* Fill up the full buffer size from the input file. */
296 n_read = stdread (buf + n_buffered, nchars - n_buffered);
298 error (EXIT_FAILURE, errno, "%s", infile);
300 n_buffered += n_read;
301 if (n_buffered != nchars)
304 /* Find where to end this chunk. */
305 bp = buf + n_buffered;
306 if (n_buffered == nchars)
308 while (bp > buf && bp[-1] != '\n')
312 /* If chunk has no newlines, use all the chunk. */
314 bp = buf + n_buffered;
316 /* Output the chars as one output file. */
317 cwrite (1, buf, bp - buf);
319 /* Discard the chars we just output; move rest of chunk
320 down to be the start of the next chunk. Source and
321 destination probably overlap. */
322 n_buffered -= bp - buf;
324 memmove (buf, bp, n_buffered);
331 main (int argc, char **argv)
333 struct stat stat_buf;
334 int num; /* numeric argument from command line */
337 type_undef, type_bytes, type_byteslines, type_lines, type_digits
338 } split_type = type_undef;
339 int in_blk_size; /* optimal block size of input file device */
340 char *buf; /* file i/o buffer */
344 int digits_optind = 0;
346 program_name = argv[0];
347 setlocale (LC_ALL, "");
348 bindtextdomain (PACKAGE, LOCALEDIR);
349 textdomain (PACKAGE);
351 atexit (close_stdout);
353 /* Parse command line options. */
360 /* This is the argv-index of the option we will read next. */
361 int this_optind = optind ? optind : 1;
364 c = getopt_long (argc, argv, "0123456789vb:l:C:", longopts, (int *) 0);
374 if (split_type != type_undef)
376 error (0, 0, _("cannot split in more than one way"));
377 usage (EXIT_FAILURE);
379 split_type = type_bytes;
380 if (xstrtol (optarg, NULL, 10, &tmp_long, "bkm") != LONGINT_OK
381 || tmp_long < 0 || tmp_long > INT_MAX)
383 error (0, 0, _("%s: invalid number of bytes"), optarg);
384 usage (EXIT_FAILURE);
386 accum = (int) tmp_long;
390 if (split_type != type_undef)
392 error (0, 0, _("cannot split in more than one way"));
393 usage (EXIT_FAILURE);
395 split_type = type_lines;
396 if (xstrtol (optarg, NULL, 10, &tmp_long, "") != LONGINT_OK
397 || tmp_long < 0 || tmp_long > INT_MAX)
399 error (0, 0, _("%s: invalid number of lines"), optarg);
400 usage (EXIT_FAILURE);
402 accum = (int) tmp_long;
406 if (split_type != type_undef)
408 error (0, 0, _("cannot split in more than one way"));
409 usage (EXIT_FAILURE);
412 split_type = type_byteslines;
413 if (xstrtol (optarg, NULL, 10, &tmp_long, "bkm") != LONGINT_OK
414 || tmp_long < 0 || tmp_long > INT_MAX)
416 error (0, 0, _("%s: invalid number of bytes"), optarg);
417 usage (EXIT_FAILURE);
419 accum = (int) tmp_long;
432 if (split_type != type_undef && split_type != type_digits)
434 error (0, 0, _("cannot split in more than one way"));
435 usage (EXIT_FAILURE);
437 if (digits_optind != 0 && digits_optind != this_optind)
438 accum = 0; /* More than one number given; ignore other. */
439 digits_optind = this_optind;
440 split_type = type_digits;
441 accum = accum * 10 + c - '0';
448 case_GETOPT_HELP_CHAR;
450 case_GETOPT_VERSION_CHAR (PROGRAM_NAME, AUTHORS);
453 usage (EXIT_FAILURE);
457 /* Handle default case. */
458 if (split_type == type_undef)
460 split_type = type_lines;
466 error (0, 0, _("invalid number"));
467 usage (EXIT_FAILURE);
471 /* Get out the filename arguments. */
474 infile = argv[optind++];
477 outbase = argv[optind++];
481 error (0, 0, _("too many arguments"));
482 usage (EXIT_FAILURE);
485 /* Open the input file. */
486 if (STREQ (infile, "-"))
490 input_desc = open (infile, O_RDONLY);
492 error (EXIT_FAILURE, errno, "%s", infile);
494 /* Binary I/O is safer when bytecounts are used. */
495 SET_BINARY (input_desc);
497 /* No output file is open now. */
500 /* Copy the output file prefix so we can add suffixes to it.
501 26**29 is certainly enough output files! */
503 outfile = xmalloc (strlen (outbase) + 30);
504 strcpy (outfile, outbase);
505 outfile_mid = outfile + strlen (outfile);
506 outfile_end = outfile_mid + 2;
507 memset (outfile_mid, 0, 30);
508 outfile_mid[0] = 'a';
509 outfile_mid[1] = 'a' - 1; /* first call to next_file_name makes it an 'a' */
511 /* Get the optimal block size of input device and make a buffer. */
513 if (fstat (input_desc, &stat_buf) < 0)
514 error (EXIT_FAILURE, errno, "%s", infile);
515 in_blk_size = ST_BLKSIZE (stat_buf);
517 buf = xmalloc (in_blk_size + 1);
523 lines_split (num, buf, in_blk_size);
527 bytes_split (num, buf, in_blk_size);
530 case type_byteslines:
531 line_bytes_split (num);
538 if (close (input_desc) < 0)
539 error (EXIT_FAILURE, errno, "%s", infile);
540 if (output_desc >= 0 && close (output_desc) < 0)
541 error (EXIT_FAILURE, errno, "%s", outfile);