1 /* cut - remove parts of lines of files
2 Copyright (C) 1997-2006 Free Software Foundation, Inc.
3 Copyright (C) 1984 David M. Ihnat
5 This program is free software; you can redistribute it and/or modify
6 it under the terms of the GNU General Public License as published by
7 the Free Software Foundation; either version 2, or (at your option)
10 This program is distributed in the hope that it will be useful,
11 but WITHOUT ANY WARRANTY; without even the implied warranty of
12 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
13 GNU General Public License for more details.
15 You should have received a copy of the GNU General Public License
16 along with this program; if not, write to the Free Software Foundation,
17 Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */
19 /* Written by David Ihnat. */
21 /* POSIX changes, bug fixes, long-named options, and cleanup
22 by David MacKenzie <djm@gnu.ai.mit.edu>.
24 Rewrite cut_fields and cut_bytes -- Jim Meyering. */
31 #include <sys/types.h>
35 #include "getndelim2.h"
40 /* The official name of this program (e.g., no `g' prefix). */
41 #define PROGRAM_NAME "cut"
43 #define AUTHORS "David Ihnat", "David MacKenzie", "Jim Meyering"
45 #define FATAL_ERROR(Message) \
48 error (0, 0, (Message)); \
49 usage (EXIT_FAILURE); \
53 /* Append LOW, HIGH to the list RP of range pairs, allocating additional
54 space if necessary. Update local variable N_RP. When allocating,
55 update global variable N_RP_ALLOCATED. */
57 #define ADD_RANGE_PAIR(rp, low, high) \
60 if (n_rp >= n_rp_allocated) \
62 (rp) = X2NREALLOC (rp, &n_rp_allocated); \
64 rp[n_rp].lo = (low); \
65 rp[n_rp].hi = (high); \
76 /* This buffer is used to support the semantics of the -s option
77 (or lack of same) when the specified field list includes (does
78 not include) the first field. In both of those cases, the entire
79 first field must be read into this buffer to determine whether it
80 is followed by a delimiter or a newline before any of it may be
81 output. Otherwise, cut_fields can do the job without using this
83 static char *field_1_buffer;
85 /* The number of bytes allocated for FIELD_1_BUFFER. */
86 static size_t field_1_bufsize;
88 /* The largest field or byte index used as an endpoint of a closed
89 or degenerate range specification; this doesn't include the starting
90 index of right-open-ended ranges. For example, with either range spec
91 `2-5,9-', `2-3,5,9-' this variable would be set to 5. */
92 static size_t max_range_endpoint;
94 /* If nonzero, this is the index of the first field in a range that goes
96 static size_t eol_range_start;
98 /* This is a bit vector.
99 In byte mode, which bytes to output.
100 In field mode, which DELIM-separated fields to output.
101 Both bytes and fields are numbered starting with 1,
102 so the zeroth bit of this array is unused.
103 A field or byte K has been selected if
104 (K <= MAX_RANGE_ENDPOINT and is_printable_field(K))
105 || (EOL_RANGE_START > 0 && K >= EOL_RANGE_START). */
106 static unsigned char *printable_field;
112 /* Output characters that are in the given bytes. */
115 /* Output the given delimeter-separated fields. */
119 /* The name this program was run with. */
122 static enum operating_mode operating_mode;
124 /* If true do not output lines containing no delimeter characters.
125 Otherwise, all such lines are printed. This option is valid only
127 static bool suppress_non_delimited;
129 /* If nonzero, print all bytes, characters, or fields _except_
130 those that were specified. */
131 static bool complement;
133 /* The delimeter character for field mode. */
134 static unsigned char delim;
136 /* True if the --output-delimiter=STRING option was specified. */
137 static bool output_delimiter_specified;
139 /* The length of output_delimiter_string. */
140 static size_t output_delimiter_length;
142 /* The output field separator string. Defaults to the 1-character
143 string consisting of the input delimiter. */
144 static char *output_delimiter_string;
146 /* True if we have ever read standard input. */
147 static bool have_read_stdin;
149 #define HT_RANGE_START_INDEX_INITIAL_CAPACITY 31
151 /* The set of range-start indices. For example, given a range-spec list like
152 `-b1,3-5,4-9,15-', the following indices will be recorded here: 1, 3, 15.
153 Note that although `4' looks like a range-start index, it is in the middle
154 of the `3-5' range, so it doesn't count.
155 This table is created/used IFF output_delimiter_specified is set. */
156 static Hash_table *range_start_ht;
158 /* For long options that have no equivalent short option, use a
159 non-character as a pseudo short option, starting with CHAR_MAX + 1. */
162 OUTPUT_DELIMITER_OPTION = CHAR_MAX + 1,
166 static struct option const longopts[] =
168 {"bytes", required_argument, NULL, 'b'},
169 {"characters", required_argument, NULL, 'c'},
170 {"fields", required_argument, NULL, 'f'},
171 {"delimiter", required_argument, NULL, 'd'},
172 {"only-delimited", no_argument, NULL, 's'},
173 {"output-delimiter", required_argument, NULL, OUTPUT_DELIMITER_OPTION},
174 {"complement", no_argument, NULL, COMPLEMENT_OPTION},
175 {GETOPT_HELP_OPTION_DECL},
176 {GETOPT_VERSION_OPTION_DECL},
183 if (status != EXIT_SUCCESS)
184 fprintf (stderr, _("Try `%s --help' for more information.\n"),
189 Usage: %s [OPTION]... [FILE]...\n\
193 Print selected parts of lines from each FILE to standard output.\n\
197 Mandatory arguments to long options are mandatory for short options too.\n\
200 -b, --bytes=LIST select only these bytes\n\
201 -c, --characters=LIST select only these characters\n\
202 -d, --delimiter=DELIM use DELIM instead of TAB for field delimiter\n\
205 -f, --fields=LIST select only these fields; also print any line\n\
206 that contains no delimiter character, unless\n\
207 the -s option is specified\n\
211 --complement complement the set of selected bytes, characters\n\
215 -s, --only-delimited do not print lines not containing delimiters\n\
216 --output-delimiter=STRING use STRING as the output delimiter\n\
217 the default is to use the input delimiter\n\
219 fputs (HELP_OPTION_DESCRIPTION, stdout);
220 fputs (VERSION_OPTION_DESCRIPTION, stdout);
223 Use one, and only one of -b, -c or -f. Each LIST is made up of one\n\
224 range, or many ranges separated by commas. Selected input is written\n\
225 in the same order that it is read, and is written exactly once.\n\
228 Each range is one of:\n\
230 N N'th byte, character or field, counted from 1\n\
231 N- from N'th byte, character or field, to end of line\n\
232 N-M from N'th to M'th (included) byte, character or field\n\
233 -M from first to M'th (included) byte, character or field\n\
235 With no FILE, or when FILE is -, read standard input.\n\
237 printf (_("\nReport bugs to <%s>.\n"), PACKAGE_BUGREPORT);
243 mark_range_start (size_t i)
245 /* Record the fact that `i' is a range-start index. */
246 void *ent_from_table = hash_insert (range_start_ht, (void*) i);
247 if (ent_from_table == NULL)
249 /* Insertion failed due to lack of memory. */
252 assert ((size_t) ent_from_table == i);
256 mark_printable_field (size_t i)
258 size_t n = i / CHAR_BIT;
259 printable_field[n] |= (1 << (i % CHAR_BIT));
263 is_printable_field (size_t i)
265 size_t n = i / CHAR_BIT;
266 return (printable_field[n] >> (i % CHAR_BIT)) & 1;
270 hash_int (const void *x, size_t tablesize)
273 uintptr_t y = (uintptr_t) x;
275 size_t y = (size_t) x;
277 return y % tablesize;
281 hash_compare_ints (void const *x, void const *y)
283 return (x == y) ? true : false;
287 is_range_start_index (size_t i)
289 return hash_lookup (range_start_ht, (void *) i) ? true : false;
292 /* Return nonzero if the K'th field or byte is printable.
293 When returning nonzero, if RANGE_START is non-NULL,
294 set *RANGE_START to true if K is the beginning of a range, and to
298 print_kth (size_t k, bool *range_start)
301 = ((0 < eol_range_start && eol_range_start <= k)
302 || (k <= max_range_endpoint && is_printable_field (k)));
304 bool is_selected = k_selected ^ complement;
305 if (range_start && is_selected)
306 *range_start = is_range_start_index (k);
311 /* Comparison function for qsort to order the list of
312 struct range_pairs. */
314 compare_ranges (const void *a, const void *b)
316 int a_start = ((const struct range_pair *) a)->lo;
317 int b_start = ((const struct range_pair *) b)->lo;
318 return a_start < b_start ? -1 : a_start > b_start;
321 /* Given the list of field or byte range specifications FIELDSTR, set
322 MAX_RANGE_ENDPOINT and allocate and initialize the PRINTABLE_FIELD
323 array. If there is a right-open-ended range, set EOL_RANGE_START
324 to its starting index. FIELDSTR should be composed of one or more
325 numbers or ranges of numbers, separated by blanks or commas.
326 Incomplete ranges may be given: `-m' means `1-m'; `n-' means `n'
327 through end of line. Return true if FIELDSTR contains at least
328 one field specification, false otherwise. */
330 /* FIXME-someday: What if the user wants to cut out the 1,000,000-th
331 field of some huge input file? This function shouldn't have to
332 allocate a table of a million bits just so we can test every
333 field < 10^6 with an array dereference. Instead, consider using
334 an adaptive approach: if the range of selected fields is too large,
335 but only a few fields/byte-offsets are actually selected, use a
336 hash table. If the range of selected fields is too large, and
337 too many are selected, then resort to using the range-pairs (the
338 `rp' array) directly. */
341 set_fields (const char *fieldstr)
343 size_t initial = 1; /* Value of first number in a range. */
344 size_t value = 0; /* If nonzero, a number being accumulated. */
345 bool dash_found = false; /* True if a '-' is found in this field. */
346 bool field_found = false; /* True if at least one field spec
347 has been processed. */
349 struct range_pair *rp = NULL;
351 size_t n_rp_allocated = 0;
353 bool in_digits = false;
355 /* Collect and store in RP the range end points.
356 It also sets EOL_RANGE_START if appropriate. */
360 if (*fieldstr == '-')
363 /* Starting a range. */
365 FATAL_ERROR (_("invalid byte or field list"));
377 else if (*fieldstr == ',' || isblank (*fieldstr) || *fieldstr == '\0')
380 /* Ending the string, or this field/byte sublist. */
385 /* A range. Possibilites: -n, m-n, n-.
386 In any case, `initial' contains the start of the range. */
389 /* `n-'. From `initial' to end of line. */
390 eol_range_start = initial;
395 /* `m-n' or `-n' (1-n). */
397 FATAL_ERROR (_("invalid byte or field list"));
399 /* Is there already a range going to end of line? */
400 if (eol_range_start != 0)
402 /* Yes. Is the new sequence already contained
403 in the old one? If so, no processing is
405 if (initial < eol_range_start)
407 /* No, the new sequence starts before the
408 old. Does the old range going to end of line
409 extend into the new range? */
410 if (eol_range_start <= value)
412 /* Yes. Simply move the end of line marker. */
413 eol_range_start = initial;
417 /* No. A simple range, before and disjoint from
418 the range going to end of line. Fill it. */
419 ADD_RANGE_PAIR (rp, initial, value);
422 /* In any case, some fields were selected. */
428 /* There is no range going to end of line. */
429 ADD_RANGE_PAIR (rp, initial, value);
437 /* A simple field number, not a range. */
438 ADD_RANGE_PAIR (rp, value, value);
443 if (*fieldstr == '\0')
450 else if (ISDIGIT (*fieldstr))
452 /* Record beginning of digit string, in case we have to
453 complain about it. */
454 static char const *num_start;
455 if (!in_digits || !num_start)
456 num_start = fieldstr;
459 /* Detect overflow. */
460 if (!DECIMAL_DIGIT_ACCUMULATE (value, *fieldstr - '0', size_t))
462 /* In case the user specified -c4294967296,22,
463 complain only about the first number. */
464 /* Determine the length of the offending number. */
465 size_t len = strspn (num_start, "0123456789");
466 char *bad_num = xstrndup (num_start, len);
467 if (operating_mode == byte_mode)
469 _("byte offset %s is too large"), quote (bad_num));
472 _("field number %s is too large"), quote (bad_num));
480 FATAL_ERROR (_("invalid byte or field list"));
483 max_range_endpoint = 0;
484 for (i = 0; i < n_rp; i++)
486 if (rp[i].hi > max_range_endpoint)
487 max_range_endpoint = rp[i].hi;
490 /* Allocate an array large enough so that it may be indexed by
491 the field numbers corresponding to all finite ranges
492 (i.e. `2-6' or `-4', but not `5-') in FIELDSTR. */
494 printable_field = xzalloc (max_range_endpoint / CHAR_BIT + 1);
496 qsort (rp, n_rp, sizeof (rp[0]), compare_ranges);
498 /* Set the array entries corresponding to integers in the ranges of RP. */
499 for (i = 0; i < n_rp; i++)
502 size_t rsi_candidate;
504 /* Record the range-start indices, i.e., record each start
505 index that is not part of any other (lo..hi] range. */
506 rsi_candidate = complement ? rp[i].hi + 1 : rp[i].lo;
507 if (output_delimiter_specified
508 && !is_printable_field (rsi_candidate))
509 mark_range_start (rsi_candidate);
511 for (j = rp[i].lo; j <= rp[i].hi; j++)
512 mark_printable_field (j);
515 if (output_delimiter_specified
517 && eol_range_start && !is_printable_field (eol_range_start))
518 mark_range_start (eol_range_start);
525 /* Read from stream STREAM, printing to standard output any selected bytes. */
528 cut_bytes (FILE *stream)
530 size_t byte_idx; /* Number of bytes in the line so far. */
531 /* Whether to begin printing delimiters between ranges for the current line.
532 Set after we've begun printing data corresponding to the first range. */
533 bool print_delimiter;
536 print_delimiter = false;
539 int c; /* Each character from the file. */
547 print_delimiter = false;
558 bool *rs = output_delimiter_specified ? &range_start : NULL;
559 if (print_kth (++byte_idx, rs))
561 if (rs && *rs && print_delimiter)
563 fwrite (output_delimiter_string, sizeof (char),
564 output_delimiter_length, stdout);
566 print_delimiter = true;
573 /* Read from stream STREAM, printing to standard output any selected fields. */
576 cut_fields (FILE *stream)
579 size_t field_idx = 1;
580 bool found_any_selected_field = false;
581 bool buffer_first_field;
589 /* To support the semantics of the -s flag, we may have to buffer
590 all of the first field to determine whether it is `delimited.'
591 But that is unnecessary if all non-delimited lines must be printed
592 and the first field has been selected, or if non-delimited lines
593 must be suppressed and the first field has *not* been selected.
594 That is because a non-delimited line has exactly one field. */
595 buffer_first_field = (suppress_non_delimited ^ !print_kth (1, NULL));
599 if (field_idx == 1 && buffer_first_field)
604 len = getndelim2 (&field_1_buffer, &field_1_bufsize, 0,
605 GETNLINE_NO_LIMIT, delim, '\n', stream);
608 free (field_1_buffer);
609 if (ferror (stream) || feof (stream))
615 assert (n_bytes != 0);
617 /* If the first field extends to the end of line (it is not
618 delimited) and we are printing all non-delimited lines,
620 if (to_uchar (field_1_buffer[n_bytes - 1]) != delim)
622 if (suppress_non_delimited)
628 fwrite (field_1_buffer, sizeof (char), n_bytes, stdout);
629 /* Make sure the output line is newline terminated. */
630 if (field_1_buffer[n_bytes - 1] != '\n')
635 if (print_kth (1, NULL))
637 /* Print the field, but not the trailing delimiter. */
638 fwrite (field_1_buffer, sizeof (char), n_bytes - 1, stdout);
639 found_any_selected_field = true;
646 if (print_kth (field_idx, NULL))
648 if (found_any_selected_field)
650 fwrite (output_delimiter_string, sizeof (char),
651 output_delimiter_length, stdout);
653 found_any_selected_field = true;
655 while ((c = getc (stream)) != delim && c != '\n' && c != EOF)
662 while ((c = getc (stream)) != delim && c != '\n' && c != EOF)
681 else if (c == '\n' || c == EOF)
683 if (found_any_selected_field
684 || !(suppress_non_delimited && field_idx == 1))
689 found_any_selected_field = false;
695 cut_stream (FILE *stream)
697 if (operating_mode == byte_mode)
703 /* Process file FILE to standard output.
704 Return true if successful. */
707 cut_file (char const *file)
711 if (STREQ (file, "-"))
713 have_read_stdin = true;
718 stream = fopen (file, "r");
721 error (0, errno, "%s", file);
730 error (0, errno, "%s", file);
733 if (STREQ (file, "-"))
734 clearerr (stream); /* Also clear EOF. */
735 else if (fclose (stream) == EOF)
737 error (0, errno, "%s", file);
744 main (int argc, char **argv)
748 bool delim_specified = false;
749 char *spec_list_string IF_LINT(= NULL);
751 initialize_main (&argc, &argv);
752 program_name = argv[0];
753 setlocale (LC_ALL, "");
754 bindtextdomain (PACKAGE, LOCALEDIR);
755 textdomain (PACKAGE);
757 atexit (close_stdout);
759 operating_mode = undefined_mode;
761 /* By default, all non-delimited lines are printed. */
762 suppress_non_delimited = false;
765 have_read_stdin = false;
767 while ((optc = getopt_long (argc, argv, "b:c:d:f:ns", longopts, NULL)) != -1)
773 /* Build the byte list. */
774 if (operating_mode != undefined_mode)
775 FATAL_ERROR (_("only one type of list may be specified"));
776 operating_mode = byte_mode;
777 spec_list_string = optarg;
781 /* Build the field list. */
782 if (operating_mode != undefined_mode)
783 FATAL_ERROR (_("only one type of list may be specified"));
784 operating_mode = field_mode;
785 spec_list_string = optarg;
790 /* Interpret -d '' to mean `use the NUL byte as the delimiter.' */
791 if (optarg[0] != '\0' && optarg[1] != '\0')
792 FATAL_ERROR (_("the delimiter must be a single character"));
794 delim_specified = true;
797 case OUTPUT_DELIMITER_OPTION:
798 output_delimiter_specified = true;
799 /* Interpret --output-delimiter='' to mean
800 `use the NUL byte as the delimiter.' */
801 output_delimiter_length = (optarg[0] == '\0'
802 ? 1 : strlen (optarg));
803 output_delimiter_string = xstrdup (optarg);
810 suppress_non_delimited = true;
813 case COMPLEMENT_OPTION:
817 case_GETOPT_HELP_CHAR;
819 case_GETOPT_VERSION_CHAR (PROGRAM_NAME, AUTHORS);
822 usage (EXIT_FAILURE);
826 if (operating_mode == undefined_mode)
827 FATAL_ERROR (_("you must specify a list of bytes, characters, or fields"));
829 if (delim != '\0' && operating_mode != field_mode)
830 FATAL_ERROR (_("an input delimiter may be specified only\
831 when operating on fields"));
833 if (suppress_non_delimited && operating_mode != field_mode)
834 FATAL_ERROR (_("suppressing non-delimited lines makes sense\n\
835 \tonly when operating on fields"));
837 if (output_delimiter_specified)
839 range_start_ht = hash_initialize (HT_RANGE_START_INDEX_INITIAL_CAPACITY,
841 hash_compare_ints, NULL);
842 if (range_start_ht == NULL)
847 if (! set_fields (spec_list_string))
849 if (operating_mode == field_mode)
850 FATAL_ERROR (_("missing list of fields"));
852 FATAL_ERROR (_("missing list of positions"));
855 if (!delim_specified)
858 if (output_delimiter_string == NULL)
860 static char dummy[2];
863 output_delimiter_string = dummy;
864 output_delimiter_length = 1;
870 for (ok = true; optind < argc; optind++)
871 ok &= cut_file (argv[optind]);
874 hash_free (range_start_ht);
876 if (have_read_stdin && fclose (stdin) == EOF)
878 error (0, errno, "-");
882 exit (ok ? EXIT_SUCCESS : EXIT_FAILURE);