1 /* split.c -- split a file into pieces.
2 Copyright (C) 1988, 1991, 1995-2010 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 3 of the License, or
7 (at your option) any later version.
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, see <http://www.gnu.org/licenses/>. */
17 /* By tege@sics.se, with rms.
20 * Implement -t CHAR or -t REGEX to specify break characters other
27 #include <sys/types.h>
31 #include "fd-reopen.h"
33 #include "full-read.h"
34 #include "full-write.h"
39 /* The official name of this program (e.g., no `g' prefix). */
40 #define PROGRAM_NAME "split"
43 proper_name_utf8 ("Torbjorn Granlund", "Torbj\303\266rn Granlund"), \
44 proper_name ("Richard M. Stallman")
46 #define DEFAULT_SUFFIX_LENGTH 2
48 /* Base name of output files. */
49 static char const *outbase;
51 /* Name of output files. */
54 /* Pointer to the end of the prefix in OUTFILE.
55 Suffixes are inserted here. */
56 static char *outfile_mid;
58 /* Length of OUTFILE's suffix. */
59 static size_t suffix_length = DEFAULT_SUFFIX_LENGTH;
61 /* Alphabet of characters to use in suffix. */
62 static char const *suffix_alphabet = "abcdefghijklmnopqrstuvwxyz";
64 /* Name of input file. May be "-". */
67 /* Descriptor on which output file is open. */
68 static int output_desc;
70 /* If true, print a diagnostic on standard error just before each
71 output file is opened. */
74 /* For long options that have no equivalent short option, use a
75 non-character as a pseudo short option, starting with CHAR_MAX + 1. */
78 VERBOSE_OPTION = CHAR_MAX + 1
81 static struct option const longopts[] =
83 {"bytes", required_argument, NULL, 'b'},
84 {"lines", required_argument, NULL, 'l'},
85 {"line-bytes", required_argument, NULL, 'C'},
86 {"suffix-length", required_argument, NULL, 'a'},
87 {"numeric-suffixes", no_argument, NULL, 'd'},
88 {"verbose", no_argument, NULL, VERBOSE_OPTION},
89 {GETOPT_HELP_OPTION_DECL},
90 {GETOPT_VERSION_OPTION_DECL},
97 if (status != EXIT_SUCCESS)
98 fprintf (stderr, _("Try `%s --help' for more information.\n"),
103 Usage: %s [OPTION]... [INPUT [PREFIX]]\n\
107 Output fixed-size pieces of INPUT to PREFIXaa, PREFIXab, ...; default\n\
108 size is 1000 lines, and default PREFIX is `x'. With no INPUT, or when INPUT\n\
109 is -, read standard input.\n\
113 Mandatory arguments to long options are mandatory for short options too.\n\
115 fprintf (stdout, _("\
116 -a, --suffix-length=N use suffixes of length N (default %d)\n\
117 -b, --bytes=SIZE put SIZE bytes per output file\n\
118 -C, --line-bytes=SIZE put at most SIZE bytes of lines per output file\n\
119 -d, --numeric-suffixes use numeric suffixes instead of alphabetic\n\
120 -l, --lines=NUMBER put NUMBER lines per output file\n\
121 "), DEFAULT_SUFFIX_LENGTH);
123 --verbose print a diagnostic just before each\n\
124 output file is opened\n\
126 fputs (HELP_OPTION_DESCRIPTION, stdout);
127 fputs (VERSION_OPTION_DESCRIPTION, stdout);
129 emit_ancillary_info ();
134 /* Compute the next sequential output file name and store it into the
138 next_file_name (void)
140 /* Index in suffix_alphabet of each character in the suffix. */
141 static size_t *sufindex;
145 /* Allocate and initialize the first file name. */
147 size_t outbase_length = strlen (outbase);
148 size_t outfile_length = outbase_length + suffix_length;
149 if (outfile_length + 1 < outbase_length)
151 outfile = xmalloc (outfile_length + 1);
152 outfile_mid = outfile + outbase_length;
153 memcpy (outfile, outbase, outbase_length);
154 memset (outfile_mid, suffix_alphabet[0], suffix_length);
155 outfile[outfile_length] = 0;
156 sufindex = xcalloc (suffix_length, sizeof *sufindex);
158 #if ! _POSIX_NO_TRUNC && HAVE_PATHCONF && defined _PC_NAME_MAX
159 /* POSIX requires that if the output file name is too long for
160 its directory, `split' must fail without creating any files.
161 This must be checked for explicitly on operating systems that
162 silently truncate file names. */
164 char *dir = dir_name (outfile);
165 long name_max = pathconf (dir, _PC_NAME_MAX);
166 if (0 <= name_max && name_max < base_len (last_component (outfile)))
167 error (EXIT_FAILURE, ENAMETOOLONG, "%s", outfile);
174 /* Increment the suffix in place, if possible. */
176 size_t i = suffix_length;
180 outfile_mid[i] = suffix_alphabet[sufindex[i]];
184 outfile_mid[i] = suffix_alphabet[sufindex[i]];
186 error (EXIT_FAILURE, 0, _("output file suffixes exhausted"));
190 /* Write BYTES bytes at BP to an output file.
191 If NEW_FILE_FLAG is true, open the next output file.
192 Otherwise add to the same output file already in use. */
195 cwrite (bool new_file_flag, const char *bp, size_t bytes)
199 if (output_desc >= 0 && close (output_desc) < 0)
200 error (EXIT_FAILURE, errno, "%s", outfile);
204 fprintf (stdout, _("creating file %s\n"), quote (outfile));
205 output_desc = open (outfile,
206 O_WRONLY | O_CREAT | O_TRUNC | O_BINARY,
207 (S_IRUSR | S_IWUSR | S_IRGRP | S_IWGRP
208 | S_IROTH | S_IWOTH));
210 error (EXIT_FAILURE, errno, "%s", outfile);
212 if (full_write (output_desc, bp, bytes) != bytes)
213 error (EXIT_FAILURE, errno, "%s", outfile);
216 /* Split into pieces of exactly N_BYTES bytes.
217 Use buffer BUF, whose size is BUFSIZE. */
220 bytes_split (uintmax_t n_bytes, char *buf, size_t bufsize)
223 bool new_file_flag = true;
225 uintmax_t to_write = n_bytes;
230 n_read = full_read (STDIN_FILENO, buf, bufsize);
231 if (n_read < bufsize && errno)
232 error (EXIT_FAILURE, errno, "%s", infile);
237 if (to_read < to_write)
239 if (to_read) /* do not write 0 bytes! */
241 cwrite (new_file_flag, bp_out, to_read);
243 new_file_flag = false;
250 cwrite (new_file_flag, bp_out, w);
253 new_file_flag = true;
258 while (n_read == bufsize);
261 /* Split into pieces of exactly N_LINES lines.
262 Use buffer BUF, whose size is BUFSIZE. */
265 lines_split (uintmax_t n_lines, char *buf, size_t bufsize)
268 char *bp, *bp_out, *eob;
269 bool new_file_flag = true;
274 n_read = full_read (STDIN_FILENO, buf, bufsize);
275 if (n_read < bufsize && errno)
276 error (EXIT_FAILURE, errno, "%s", infile);
282 bp = memchr (bp, '\n', eob - bp + 1);
285 if (eob != bp_out) /* do not write 0 bytes! */
287 size_t len = eob - bp_out;
288 cwrite (new_file_flag, bp_out, len);
289 new_file_flag = false;
297 cwrite (new_file_flag, bp_out, bp - bp_out);
299 new_file_flag = true;
304 while (n_read == bufsize);
307 /* Split into pieces that are as large as possible while still not more
308 than N_BYTES bytes, and are split on line boundaries except
309 where lines longer than N_BYTES bytes occur.
310 FIXME: Allow N_BYTES to be any uintmax_t value, and don't require a
311 buffer of size N_BYTES, in case N_BYTES is very large. */
314 line_bytes_split (size_t n_bytes)
318 size_t n_buffered = 0;
319 char *buf = xmalloc (n_bytes);
323 /* Fill up the full buffer size from the input file. */
325 size_t to_read = n_bytes - n_buffered;
326 size_t n_read = full_read (STDIN_FILENO, buf + n_buffered, to_read);
327 if (n_read < to_read && errno)
328 error (EXIT_FAILURE, errno, "%s", infile);
330 n_buffered += n_read;
331 if (n_buffered != n_bytes)
338 /* Find where to end this chunk. */
339 bp = buf + n_buffered;
340 if (n_buffered == n_bytes)
342 while (bp > buf && bp[-1] != '\n')
346 /* If chunk has no newlines, use all the chunk. */
348 bp = buf + n_buffered;
350 /* Output the chars as one output file. */
351 cwrite (true, buf, bp - buf);
353 /* Discard the chars we just output; move rest of chunk
354 down to be the start of the next chunk. Source and
355 destination probably overlap. */
356 n_buffered -= bp - buf;
358 memmove (buf, bp, n_buffered);
364 #define FAIL_ONLY_ONE_WAY() \
367 error (0, 0, _("cannot split in more than one way")); \
368 usage (EXIT_FAILURE); \
373 main (int argc, char **argv)
375 struct stat stat_buf;
378 type_undef, type_bytes, type_byteslines, type_lines, type_digits
379 } split_type = type_undef;
380 size_t in_blk_size; /* optimal block size of input file device */
381 char *buf; /* file i/o buffer */
382 size_t page_size = getpagesize ();
384 static char const multipliers[] = "bEGKkMmPTYZ0";
386 int digits_optind = 0;
388 initialize_main (&argc, &argv);
389 set_program_name (argv[0]);
390 setlocale (LC_ALL, "");
391 bindtextdomain (PACKAGE, LOCALEDIR);
392 textdomain (PACKAGE);
394 atexit (close_stdout);
396 /* Parse command line options. */
398 infile = bad_cast ( "-");
399 outbase = bad_cast ("x");
403 /* This is the argv-index of the option we will read next. */
404 int this_optind = optind ? optind : 1;
406 c = getopt_long (argc, argv, "0123456789C:a:b:dl:", longopts, NULL);
415 if (xstrtoul (optarg, NULL, 10, &tmp, "") != LONGINT_OK
416 || SIZE_MAX / sizeof (size_t) < tmp)
418 error (0, 0, _("%s: invalid suffix length"), optarg);
419 usage (EXIT_FAILURE);
426 if (split_type != type_undef)
427 FAIL_ONLY_ONE_WAY ();
428 split_type = type_bytes;
429 if (xstrtoumax (optarg, NULL, 10, &n_units, multipliers) != LONGINT_OK
432 error (0, 0, _("%s: invalid number of bytes"), optarg);
433 usage (EXIT_FAILURE);
435 /* If input is a pipe, we could get more data than is possible
436 to write to a single file, so indicate that immediately
437 rather than having possibly future invocations fail. */
438 if (OFF_T_MAX < n_units)
439 error (EXIT_FAILURE, EFBIG,
440 _("%s: invalid number of bytes"), optarg);
445 if (split_type != type_undef)
446 FAIL_ONLY_ONE_WAY ();
447 split_type = type_lines;
448 if (xstrtoumax (optarg, NULL, 10, &n_units, "") != LONGINT_OK
451 error (0, 0, _("%s: invalid number of lines"), optarg);
452 usage (EXIT_FAILURE);
457 if (split_type != type_undef)
458 FAIL_ONLY_ONE_WAY ();
459 split_type = type_byteslines;
460 if (xstrtoumax (optarg, NULL, 10, &n_units, multipliers) != LONGINT_OK
461 || n_units == 0 || SIZE_MAX < n_units)
463 error (0, 0, _("%s: invalid number of bytes"), optarg);
464 usage (EXIT_FAILURE);
466 if (OFF_T_MAX < n_units)
467 error (EXIT_FAILURE, EFBIG,
468 _("%s: invalid number of bytes"), optarg);
481 if (split_type == type_undef)
483 split_type = type_digits;
486 if (split_type != type_undef && split_type != type_digits)
487 FAIL_ONLY_ONE_WAY ();
488 if (digits_optind != 0 && digits_optind != this_optind)
489 n_units = 0; /* More than one number given; ignore other. */
490 digits_optind = this_optind;
491 if (!DECIMAL_DIGIT_ACCUMULATE (n_units, c - '0', uintmax_t))
493 char buffer[INT_BUFSIZE_BOUND (uintmax_t)];
494 error (EXIT_FAILURE, 0,
495 _("line count option -%s%c... is too large"),
496 umaxtostr (n_units, buffer), c);
501 suffix_alphabet = "0123456789";
508 case_GETOPT_HELP_CHAR;
510 case_GETOPT_VERSION_CHAR (PROGRAM_NAME, AUTHORS);
513 usage (EXIT_FAILURE);
517 /* Handle default case. */
518 if (split_type == type_undef)
520 split_type = type_lines;
526 error (0, 0, _("invalid number of lines: 0"));
527 usage (EXIT_FAILURE);
530 /* Get out the filename arguments. */
533 infile = argv[optind++];
536 outbase = argv[optind++];
540 error (0, 0, _("extra operand %s"), quote (argv[optind]));
541 usage (EXIT_FAILURE);
544 /* Open the input file. */
545 if (! STREQ (infile, "-")
546 && fd_reopen (STDIN_FILENO, infile, O_RDONLY, 0) < 0)
547 error (EXIT_FAILURE, errno, _("cannot open %s for reading"),
550 /* Binary I/O is safer when bytecounts are used. */
551 if (O_BINARY && ! isatty (STDIN_FILENO))
552 xfreopen (NULL, "rb", stdin);
554 /* No output file is open now. */
557 /* Get the optimal block size of input device and make a buffer. */
559 if (fstat (STDIN_FILENO, &stat_buf) != 0)
560 error (EXIT_FAILURE, errno, "%s", infile);
561 in_blk_size = io_blksize (stat_buf);
563 buf = ptr_align (xmalloc (in_blk_size + 1 + page_size - 1), page_size);
569 lines_split (n_units, buf, in_blk_size);
573 bytes_split (n_units, buf, in_blk_size);
576 case type_byteslines:
577 line_bytes_split (n_units);
584 if (close (STDIN_FILENO) != 0)
585 error (EXIT_FAILURE, errno, "%s", infile);
586 if (output_desc >= 0 && close (output_desc) < 0)
587 error (EXIT_FAILURE, errno, "%s", outfile);