1 /* cut - remove parts of lines of files
2 Copyright (C) 1997-2013 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 3 of the License, or
8 (at your option) any later version.
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, see <http://www.gnu.org/licenses/>. */
18 /* Written by David Ihnat. */
20 /* POSIX changes, bug fixes, long-named options, and cleanup
21 by David MacKenzie <djm@gnu.ai.mit.edu>.
23 Rewrite cut_fields and cut_bytes -- Jim Meyering. */
30 #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"
44 proper_name ("David M. Ihnat"), \
45 proper_name ("David MacKenzie"), \
46 proper_name ("Jim Meyering")
48 #define FATAL_ERROR(Message) \
51 error (0, 0, (Message)); \
52 usage (EXIT_FAILURE); \
56 /* Append LOW, HIGH to the list RP of range pairs, allocating additional
57 space if necessary. Update local variable N_RP. When allocating,
58 update global variable N_RP_ALLOCATED. */
60 #define ADD_RANGE_PAIR(rp, low, high) \
63 if (low == 0 || high == 0) \
64 FATAL_ERROR (_("fields and positions are numbered from 1")); \
65 if (n_rp >= n_rp_allocated) \
67 (rp) = X2NREALLOC (rp, &n_rp_allocated); \
69 rp[n_rp].lo = (low); \
70 rp[n_rp].hi = (high); \
81 /* This buffer is used to support the semantics of the -s option
82 (or lack of same) when the specified field list includes (does
83 not include) the first field. In both of those cases, the entire
84 first field must be read into this buffer to determine whether it
85 is followed by a delimiter or a newline before any of it may be
86 output. Otherwise, cut_fields can do the job without using this
88 static char *field_1_buffer;
90 /* The number of bytes allocated for FIELD_1_BUFFER. */
91 static size_t field_1_bufsize;
93 /* The largest field or byte index used as an endpoint of a closed
94 or degenerate range specification; this doesn't include the starting
95 index of right-open-ended ranges. For example, with either range spec
96 '2-5,9-', '2-3,5,9-' this variable would be set to 5. */
97 static size_t max_range_endpoint;
99 /* If nonzero, this is the index of the first field in a range that goes
101 static size_t eol_range_start;
103 /* This is a bit vector.
104 In byte mode, which bytes to output.
105 In field mode, which DELIM-separated fields to output.
106 Both bytes and fields are numbered starting with 1,
107 so the zeroth bit of this array is unused.
108 A field or byte K has been selected if
109 (K <= MAX_RANGE_ENDPOINT and is_printable_field(K))
110 || (EOL_RANGE_START > 0 && K >= EOL_RANGE_START). */
111 static unsigned char *printable_field;
117 /* Output characters that are in the given bytes. */
120 /* Output the given delimeter-separated fields. */
124 static enum operating_mode operating_mode;
126 /* If true do not output lines containing no delimeter characters.
127 Otherwise, all such lines are printed. This option is valid only
129 static bool suppress_non_delimited;
131 /* If nonzero, print all bytes, characters, or fields _except_
132 those that were specified. */
133 static bool complement;
135 /* The delimeter character for field mode. */
136 static unsigned char delim;
138 /* True if the --output-delimiter=STRING option was specified. */
139 static bool output_delimiter_specified;
141 /* The length of output_delimiter_string. */
142 static size_t output_delimiter_length;
144 /* The output field separator string. Defaults to the 1-character
145 string consisting of the input delimiter. */
146 static char *output_delimiter_string;
148 /* True if we have ever read standard input. */
149 static bool have_read_stdin;
151 #define HT_RANGE_START_INDEX_INITIAL_CAPACITY 31
153 /* The set of range-start indices. For example, given a range-spec list like
154 '-b1,3-5,4-9,15-', the following indices will be recorded here: 1, 3, 15.
155 Note that although '4' looks like a range-start index, it is in the middle
156 of the '3-5' range, so it doesn't count.
157 This table is created/used IFF output_delimiter_specified is set. */
158 static Hash_table *range_start_ht;
160 /* For long options that have no equivalent short option, use a
161 non-character as a pseudo short option, starting with CHAR_MAX + 1. */
164 OUTPUT_DELIMITER_OPTION = CHAR_MAX + 1,
168 static struct option const longopts[] =
170 {"bytes", required_argument, NULL, 'b'},
171 {"characters", required_argument, NULL, 'c'},
172 {"fields", required_argument, NULL, 'f'},
173 {"delimiter", required_argument, NULL, 'd'},
174 {"only-delimited", no_argument, NULL, 's'},
175 {"output-delimiter", required_argument, NULL, OUTPUT_DELIMITER_OPTION},
176 {"complement", no_argument, NULL, COMPLEMENT_OPTION},
177 {GETOPT_HELP_OPTION_DECL},
178 {GETOPT_VERSION_OPTION_DECL},
185 if (status != EXIT_SUCCESS)
190 Usage: %s OPTION... [FILE]...\n\
194 Print selected parts of lines from each FILE to standard output.\n\
198 Mandatory arguments to long options are mandatory for short options too.\n\
201 -b, --bytes=LIST select only these bytes\n\
202 -c, --characters=LIST select only these characters\n\
203 -d, --delimiter=DELIM use DELIM instead of TAB for field delimiter\n\
206 -f, --fields=LIST select only these fields; also print any line\n\
207 that contains no delimiter character, unless\n\
208 the -s option is specified\n\
212 --complement complement the set of selected bytes, characters\n\
216 -s, --only-delimited do not print lines not containing delimiters\n\
217 --output-delimiter=STRING use STRING as the output delimiter\n\
218 the default is to use the input delimiter\n\
220 fputs (HELP_OPTION_DESCRIPTION, stdout);
221 fputs (VERSION_OPTION_DESCRIPTION, stdout);
224 Use one, and only one of -b, -c or -f. Each LIST is made up of one\n\
225 range, or many ranges separated by commas. Selected input is written\n\
226 in the same order that it is read, and is written exactly once.\n\
229 Each range is one of:\n\
231 N N'th byte, character or field, counted from 1\n\
232 N- from N'th byte, character or field, to end of line\n\
233 N-M from N'th to M'th (included) byte, character or field\n\
234 -M from first to M'th (included) byte, character or field\n\
236 With no FILE, or when FILE is -, read standard input.\n\
238 emit_ancillary_info ();
244 mark_range_start (size_t i)
246 /* Record the fact that 'i' is a range-start index. */
247 void *ent_from_table = hash_insert (range_start_ht, (void*) i);
248 if (ent_from_table == NULL)
250 /* Insertion failed due to lack of memory. */
253 assert ((size_t) ent_from_table == i);
257 mark_printable_field (size_t i)
259 size_t n = i / CHAR_BIT;
260 printable_field[n] |= (1 << (i % CHAR_BIT));
264 is_printable_field (size_t i)
266 size_t n = i / CHAR_BIT;
267 return (printable_field[n] >> (i % CHAR_BIT)) & 1;
271 hash_int (const void *x, size_t tablesize)
274 uintptr_t y = (uintptr_t) x;
276 size_t y = (size_t) x;
278 return y % tablesize;
282 hash_compare_ints (void const *x, void const *y)
284 return (x == y) ? true : false;
288 is_range_start_index (size_t i)
290 return hash_lookup (range_start_ht, (void *) i) ? true : false;
293 /* Return nonzero if the K'th field or byte is printable.
294 When returning nonzero, if RANGE_START is non-NULL,
295 set *RANGE_START to true if K is the beginning of a range, and to
299 print_kth (size_t k, bool *range_start)
302 = ((0 < eol_range_start && eol_range_start <= k)
303 || (k <= max_range_endpoint && is_printable_field (k)));
305 bool is_selected = k_selected ^ complement;
306 if (range_start && is_selected)
307 *range_start = is_range_start_index (k);
312 /* Comparison function for qsort to order the list of
313 struct range_pairs. */
315 compare_ranges (const void *a, const void *b)
317 int a_start = ((const struct range_pair *) a)->lo;
318 int b_start = ((const struct range_pair *) b)->lo;
319 return a_start < b_start ? -1 : a_start > b_start;
322 /* Given the list of field or byte range specifications FIELDSTR, set
323 MAX_RANGE_ENDPOINT and allocate and initialize the PRINTABLE_FIELD
324 array. If there is a right-open-ended range, set EOL_RANGE_START
325 to its starting index. FIELDSTR should be composed of one or more
326 numbers or ranges of numbers, separated by blanks or commas.
327 Incomplete ranges may be given: '-m' means '1-m'; 'n-' means 'n'
328 through end of line. Return true if FIELDSTR contains at least
329 one field specification, false otherwise. */
331 /* FIXME-someday: What if the user wants to cut out the 1,000,000-th
332 field of some huge input file? This function shouldn't have to
333 allocate a table of a million bits just so we can test every
334 field < 10^6 with an array dereference. Instead, consider using
335 an adaptive approach: if the range of selected fields is too large,
336 but only a few fields/byte-offsets are actually selected, use a
337 hash table. If the range of selected fields is too large, and
338 too many are selected, then resort to using the range-pairs (the
339 'rp' array) directly. */
342 set_fields (const char *fieldstr)
344 size_t initial = 1; /* Value of first number in a range. */
345 size_t value = 0; /* If nonzero, a number being accumulated. */
346 bool lhs_specified = false;
347 bool rhs_specified = false;
348 bool dash_found = false; /* True if a '-' is found in this field. */
349 bool field_found = false; /* True if at least one field spec
350 has been processed. */
352 struct range_pair *rp = NULL;
354 size_t n_rp_allocated = 0;
356 bool in_digits = false;
358 /* Collect and store in RP the range end points.
359 It also sets EOL_RANGE_START if appropriate. */
363 if (*fieldstr == '-')
366 /* Starting a range. */
368 FATAL_ERROR (_("invalid byte, character or field list"));
372 if (lhs_specified && !value)
373 FATAL_ERROR (_("fields and positions are numbered from 1"));
375 initial = (lhs_specified ? value : 1);
378 else if (*fieldstr == ','
379 || isblank (to_uchar (*fieldstr)) || *fieldstr == '\0')
382 /* Ending the string, or this field/byte sublist. */
387 if (!lhs_specified && !rhs_specified)
388 FATAL_ERROR (_("invalid range with no endpoint: -"));
390 /* A range. Possibilities: -n, m-n, n-.
391 In any case, 'initial' contains the start of the range. */
394 /* 'n-'. From 'initial' to end of line. If we've already
395 seen an M- range, ignore subsequent N- unless N < M. */
396 if (eol_range_start == 0 || initial < eol_range_start)
397 eol_range_start = initial;
402 /* 'm-n' or '-n' (1-n). */
404 FATAL_ERROR (_("invalid decreasing range"));
406 /* Is there already a range going to end of line? */
407 if (eol_range_start != 0)
409 /* Yes. Is the new sequence already contained
410 in the old one? If so, no processing is
412 if (initial < eol_range_start)
414 /* No, the new sequence starts before the
415 old. Does the old range going to end of line
416 extend into the new range? */
417 if (eol_range_start <= value)
419 /* Yes. Simply move the end of line marker. */
420 eol_range_start = initial;
424 /* No. A simple range, before and disjoint from
425 the range going to end of line. Fill it. */
426 ADD_RANGE_PAIR (rp, initial, value);
429 /* In any case, some fields were selected. */
435 /* There is no range going to end of line. */
436 ADD_RANGE_PAIR (rp, initial, value);
444 /* A simple field number, not a range. */
445 ADD_RANGE_PAIR (rp, value, value);
450 if (*fieldstr == '\0')
456 lhs_specified = false;
457 rhs_specified = false;
459 else if (ISDIGIT (*fieldstr))
461 /* Record beginning of digit string, in case we have to
462 complain about it. */
463 static char const *num_start;
464 if (!in_digits || !num_start)
465 num_start = fieldstr;
473 /* Detect overflow. */
474 if (!DECIMAL_DIGIT_ACCUMULATE (value, *fieldstr - '0', size_t))
476 /* In case the user specified -c$(echo 2^64|bc),22,
477 complain only about the first number. */
478 /* Determine the length of the offending number. */
479 size_t len = strspn (num_start, "0123456789");
480 char *bad_num = xstrndup (num_start, len);
481 if (operating_mode == byte_mode)
483 _("byte offset %s is too large"), quote (bad_num));
486 _("field number %s is too large"), quote (bad_num));
494 FATAL_ERROR (_("invalid byte, character or field list"));
497 max_range_endpoint = 0;
498 for (i = 0; i < n_rp; i++)
500 if (rp[i].hi > max_range_endpoint)
501 max_range_endpoint = rp[i].hi;
504 /* Allocate an array large enough so that it may be indexed by
505 the field numbers corresponding to all finite ranges
506 (i.e. '2-6' or '-4', but not '5-') in FIELDSTR. */
508 if (max_range_endpoint)
509 printable_field = xzalloc (max_range_endpoint / CHAR_BIT + 1);
511 qsort (rp, n_rp, sizeof (rp[0]), compare_ranges);
513 /* Set the array entries corresponding to integers in the ranges of RP. */
514 for (i = 0; i < n_rp; i++)
516 /* Ignore any range that is subsumed by the to-EOL range. */
517 if (eol_range_start && eol_range_start <= rp[i].lo)
520 /* Record the range-start indices, i.e., record each start
521 index that is not part of any other (lo..hi] range. */
522 size_t rsi_candidate = complement ? rp[i].hi + 1 : rp[i].lo;
523 if (output_delimiter_specified
524 && !is_printable_field (rsi_candidate))
525 mark_range_start (rsi_candidate);
527 for (size_t j = rp[i].lo; j <= rp[i].hi; j++)
528 mark_printable_field (j);
531 if (output_delimiter_specified
534 && max_range_endpoint && !is_printable_field (eol_range_start))
535 mark_range_start (eol_range_start);
542 /* Read from stream STREAM, printing to standard output any selected bytes. */
545 cut_bytes (FILE *stream)
547 size_t byte_idx; /* Number of bytes in the line so far. */
548 /* Whether to begin printing delimiters between ranges for the current line.
549 Set after we've begun printing data corresponding to the first range. */
550 bool print_delimiter;
553 print_delimiter = false;
556 int c; /* Each character from the file. */
564 print_delimiter = false;
575 bool *rs = output_delimiter_specified ? &range_start : NULL;
576 if (print_kth (++byte_idx, rs))
578 if (rs && *rs && print_delimiter)
580 fwrite (output_delimiter_string, sizeof (char),
581 output_delimiter_length, stdout);
583 print_delimiter = true;
590 /* Read from stream STREAM, printing to standard output any selected fields. */
593 cut_fields (FILE *stream)
596 size_t field_idx = 1;
597 bool found_any_selected_field = false;
598 bool buffer_first_field;
606 /* To support the semantics of the -s flag, we may have to buffer
607 all of the first field to determine whether it is 'delimited.'
608 But that is unnecessary if all non-delimited lines must be printed
609 and the first field has been selected, or if non-delimited lines
610 must be suppressed and the first field has *not* been selected.
611 That is because a non-delimited line has exactly one field. */
612 buffer_first_field = (suppress_non_delimited ^ !print_kth (1, NULL));
616 if (field_idx == 1 && buffer_first_field)
621 len = getndelim2 (&field_1_buffer, &field_1_bufsize, 0,
622 GETNLINE_NO_LIMIT, delim, '\n', stream);
625 free (field_1_buffer);
626 field_1_buffer = NULL;
627 if (ferror (stream) || feof (stream))
633 assert (n_bytes != 0);
635 /* If the first field extends to the end of line (it is not
636 delimited) and we are printing all non-delimited lines,
638 if (to_uchar (field_1_buffer[n_bytes - 1]) != delim)
640 if (suppress_non_delimited)
646 fwrite (field_1_buffer, sizeof (char), n_bytes, stdout);
647 /* Make sure the output line is newline terminated. */
648 if (field_1_buffer[n_bytes - 1] != '\n')
653 if (print_kth (1, NULL))
655 /* Print the field, but not the trailing delimiter. */
656 fwrite (field_1_buffer, sizeof (char), n_bytes - 1, stdout);
657 found_any_selected_field = true;
664 if (print_kth (field_idx, NULL))
666 if (found_any_selected_field)
668 fwrite (output_delimiter_string, sizeof (char),
669 output_delimiter_length, stdout);
671 found_any_selected_field = true;
673 while ((c = getc (stream)) != delim && c != '\n' && c != EOF)
680 while ((c = getc (stream)) != delim && c != '\n' && c != EOF)
699 else if (c == '\n' || c == EOF)
701 if (found_any_selected_field
702 || !(suppress_non_delimited && field_idx == 1))
707 found_any_selected_field = false;
713 cut_stream (FILE *stream)
715 if (operating_mode == byte_mode)
721 /* Process file FILE to standard output.
722 Return true if successful. */
725 cut_file (char const *file)
729 if (STREQ (file, "-"))
731 have_read_stdin = true;
736 stream = fopen (file, "r");
739 error (0, errno, "%s", file);
744 fadvise (stream, FADVISE_SEQUENTIAL);
750 error (0, errno, "%s", file);
753 if (STREQ (file, "-"))
754 clearerr (stream); /* Also clear EOF. */
755 else if (fclose (stream) == EOF)
757 error (0, errno, "%s", file);
764 main (int argc, char **argv)
768 bool delim_specified = false;
769 char *spec_list_string IF_LINT ( = NULL);
771 initialize_main (&argc, &argv);
772 set_program_name (argv[0]);
773 setlocale (LC_ALL, "");
774 bindtextdomain (PACKAGE, LOCALEDIR);
775 textdomain (PACKAGE);
777 atexit (close_stdout);
779 operating_mode = undefined_mode;
781 /* By default, all non-delimited lines are printed. */
782 suppress_non_delimited = false;
785 have_read_stdin = false;
787 while ((optc = getopt_long (argc, argv, "b:c:d:f:ns", longopts, NULL)) != -1)
793 /* Build the byte list. */
794 if (operating_mode != undefined_mode)
795 FATAL_ERROR (_("only one type of list may be specified"));
796 operating_mode = byte_mode;
797 spec_list_string = optarg;
801 /* Build the field list. */
802 if (operating_mode != undefined_mode)
803 FATAL_ERROR (_("only one type of list may be specified"));
804 operating_mode = field_mode;
805 spec_list_string = optarg;
810 /* Interpret -d '' to mean 'use the NUL byte as the delimiter.' */
811 if (optarg[0] != '\0' && optarg[1] != '\0')
812 FATAL_ERROR (_("the delimiter must be a single character"));
814 delim_specified = true;
817 case OUTPUT_DELIMITER_OPTION:
818 output_delimiter_specified = true;
819 /* Interpret --output-delimiter='' to mean
820 'use the NUL byte as the delimiter.' */
821 output_delimiter_length = (optarg[0] == '\0'
822 ? 1 : strlen (optarg));
823 output_delimiter_string = xstrdup (optarg);
830 suppress_non_delimited = true;
833 case COMPLEMENT_OPTION:
837 case_GETOPT_HELP_CHAR;
839 case_GETOPT_VERSION_CHAR (PROGRAM_NAME, AUTHORS);
842 usage (EXIT_FAILURE);
846 if (operating_mode == undefined_mode)
847 FATAL_ERROR (_("you must specify a list of bytes, characters, or fields"));
849 if (delim_specified && operating_mode != field_mode)
850 FATAL_ERROR (_("an input delimiter may be specified only\
851 when operating on fields"));
853 if (suppress_non_delimited && operating_mode != field_mode)
854 FATAL_ERROR (_("suppressing non-delimited lines makes sense\n\
855 \tonly when operating on fields"));
857 if (output_delimiter_specified)
859 range_start_ht = hash_initialize (HT_RANGE_START_INDEX_INITIAL_CAPACITY,
861 hash_compare_ints, NULL);
862 if (range_start_ht == NULL)
867 if (! set_fields (spec_list_string))
869 if (operating_mode == field_mode)
870 FATAL_ERROR (_("missing list of fields"));
872 FATAL_ERROR (_("missing list of positions"));
875 if (!delim_specified)
878 if (output_delimiter_string == NULL)
880 static char dummy[2];
883 output_delimiter_string = dummy;
884 output_delimiter_length = 1;
890 for (ok = true; optind < argc; optind++)
891 ok &= cut_file (argv[optind]);
894 hash_free (range_start_ht);
896 if (have_read_stdin && fclose (stdin) == EOF)
898 error (0, errno, "-");
902 exit (ok ? EXIT_SUCCESS : EXIT_FAILURE);