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\
197 emit_mandatory_arg_note ();
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 emit_ancillary_info ();
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 lhs_specified
= false;
346 bool rhs_specified
= false;
347 bool dash_found
= false; /* True if a '-' is found in this field. */
348 bool field_found
= false; /* True if at least one field spec
349 has been processed. */
351 struct range_pair
*rp
= NULL
;
353 size_t n_rp_allocated
= 0;
355 bool in_digits
= false;
357 /* Collect and store in RP the range end points.
358 It also sets EOL_RANGE_START if appropriate. */
362 if (*fieldstr
== '-')
365 /* Starting a range. */
367 FATAL_ERROR (_("invalid byte, character or field list"));
371 if (lhs_specified
&& !value
)
372 FATAL_ERROR (_("fields and positions are numbered from 1"));
374 initial
= (lhs_specified
? value
: 1);
377 else if (*fieldstr
== ','
378 || isblank (to_uchar (*fieldstr
)) || *fieldstr
== '\0')
381 /* Ending the string, or this field/byte sublist. */
386 if (!lhs_specified
&& !rhs_specified
)
387 FATAL_ERROR (_("invalid range with no endpoint: -"));
389 /* A range. Possibilities: -n, m-n, n-.
390 In any case, 'initial' contains the start of the range. */
393 /* 'n-'. From 'initial' to end of line. If we've already
394 seen an M- range, ignore subsequent N- unless N < M. */
395 if (eol_range_start
== 0 || initial
< eol_range_start
)
396 eol_range_start
= initial
;
401 /* 'm-n' or '-n' (1-n). */
403 FATAL_ERROR (_("invalid decreasing range"));
405 /* Is there already a range going to end of line? */
406 if (eol_range_start
!= 0)
408 /* Yes. Is the new sequence already contained
409 in the old one? If so, no processing is
411 if (initial
< eol_range_start
)
413 /* No, the new sequence starts before the
414 old. Does the old range going to end of line
415 extend into the new range? */
416 if (eol_range_start
<= value
)
418 /* Yes. Simply move the end of line marker. */
419 eol_range_start
= initial
;
423 /* No. A simple range, before and disjoint from
424 the range going to end of line. Fill it. */
425 ADD_RANGE_PAIR (rp
, initial
, value
);
428 /* In any case, some fields were selected. */
434 /* There is no range going to end of line. */
435 ADD_RANGE_PAIR (rp
, initial
, value
);
443 /* A simple field number, not a range. */
444 ADD_RANGE_PAIR (rp
, value
, value
);
449 if (*fieldstr
== '\0')
455 lhs_specified
= false;
456 rhs_specified
= false;
458 else if (ISDIGIT (*fieldstr
))
460 /* Record beginning of digit string, in case we have to
461 complain about it. */
462 static char const *num_start
;
463 if (!in_digits
|| !num_start
)
464 num_start
= fieldstr
;
472 /* Detect overflow. */
473 if (!DECIMAL_DIGIT_ACCUMULATE (value
, *fieldstr
- '0', size_t))
475 /* In case the user specified -c$(echo 2^64|bc),22,
476 complain only about the first number. */
477 /* Determine the length of the offending number. */
478 size_t len
= strspn (num_start
, "0123456789");
479 char *bad_num
= xstrndup (num_start
, len
);
480 if (operating_mode
== byte_mode
)
482 _("byte offset %s is too large"), quote (bad_num
));
485 _("field number %s is too large"), quote (bad_num
));
493 FATAL_ERROR (_("invalid byte, character or field list"));
496 max_range_endpoint
= 0;
497 for (i
= 0; i
< n_rp
; i
++)
499 if (rp
[i
].hi
> max_range_endpoint
)
500 max_range_endpoint
= rp
[i
].hi
;
503 /* Allocate an array large enough so that it may be indexed by
504 the field numbers corresponding to all finite ranges
505 (i.e. '2-6' or '-4', but not '5-') in FIELDSTR. */
507 if (max_range_endpoint
)
508 printable_field
= xzalloc (max_range_endpoint
/ CHAR_BIT
+ 1);
510 qsort (rp
, n_rp
, sizeof (rp
[0]), compare_ranges
);
512 /* Set the array entries corresponding to integers in the ranges of RP. */
513 for (i
= 0; i
< n_rp
; i
++)
515 /* Ignore any range that is subsumed by the to-EOL range. */
516 if (eol_range_start
&& eol_range_start
<= rp
[i
].lo
)
519 /* Record the range-start indices, i.e., record each start
520 index that is not part of any other (lo..hi] range. */
521 size_t rsi_candidate
= complement
? rp
[i
].hi
+ 1 : rp
[i
].lo
;
522 if (output_delimiter_specified
523 && !is_printable_field (rsi_candidate
))
524 mark_range_start (rsi_candidate
);
526 for (size_t j
= rp
[i
].lo
; j
<= rp
[i
].hi
; j
++)
527 mark_printable_field (j
);
530 if (output_delimiter_specified
533 && max_range_endpoint
&& !is_printable_field (eol_range_start
))
534 mark_range_start (eol_range_start
);
541 /* Read from stream STREAM, printing to standard output any selected bytes. */
544 cut_bytes (FILE *stream
)
546 size_t byte_idx
; /* Number of bytes in the line so far. */
547 /* Whether to begin printing delimiters between ranges for the current line.
548 Set after we've begun printing data corresponding to the first range. */
549 bool print_delimiter
;
552 print_delimiter
= false;
555 int c
; /* Each character from the file. */
563 print_delimiter
= false;
574 bool *rs
= output_delimiter_specified
? &range_start
: NULL
;
575 if (print_kth (++byte_idx
, rs
))
577 if (rs
&& *rs
&& print_delimiter
)
579 fwrite (output_delimiter_string
, sizeof (char),
580 output_delimiter_length
, stdout
);
582 print_delimiter
= true;
589 /* Read from stream STREAM, printing to standard output any selected fields. */
592 cut_fields (FILE *stream
)
595 size_t field_idx
= 1;
596 bool found_any_selected_field
= false;
597 bool buffer_first_field
;
605 /* To support the semantics of the -s flag, we may have to buffer
606 all of the first field to determine whether it is 'delimited.'
607 But that is unnecessary if all non-delimited lines must be printed
608 and the first field has been selected, or if non-delimited lines
609 must be suppressed and the first field has *not* been selected.
610 That is because a non-delimited line has exactly one field. */
611 buffer_first_field
= (suppress_non_delimited
^ !print_kth (1, NULL
));
615 if (field_idx
== 1 && buffer_first_field
)
620 len
= getndelim2 (&field_1_buffer
, &field_1_bufsize
, 0,
621 GETNLINE_NO_LIMIT
, delim
, '\n', stream
);
624 free (field_1_buffer
);
625 field_1_buffer
= NULL
;
626 if (ferror (stream
) || feof (stream
))
632 assert (n_bytes
!= 0);
634 /* If the first field extends to the end of line (it is not
635 delimited) and we are printing all non-delimited lines,
637 if (to_uchar (field_1_buffer
[n_bytes
- 1]) != delim
)
639 if (suppress_non_delimited
)
645 fwrite (field_1_buffer
, sizeof (char), n_bytes
, stdout
);
646 /* Make sure the output line is newline terminated. */
647 if (field_1_buffer
[n_bytes
- 1] != '\n')
652 if (print_kth (1, NULL
))
654 /* Print the field, but not the trailing delimiter. */
655 fwrite (field_1_buffer
, sizeof (char), n_bytes
- 1, stdout
);
656 found_any_selected_field
= true;
663 if (print_kth (field_idx
, NULL
))
665 if (found_any_selected_field
)
667 fwrite (output_delimiter_string
, sizeof (char),
668 output_delimiter_length
, stdout
);
670 found_any_selected_field
= true;
672 while ((c
= getc (stream
)) != delim
&& c
!= '\n' && c
!= EOF
)
679 while ((c
= getc (stream
)) != delim
&& c
!= '\n' && c
!= EOF
)
698 else if (c
== '\n' || c
== EOF
)
700 if (found_any_selected_field
701 || !(suppress_non_delimited
&& field_idx
== 1))
706 found_any_selected_field
= false;
712 cut_stream (FILE *stream
)
714 if (operating_mode
== byte_mode
)
720 /* Process file FILE to standard output.
721 Return true if successful. */
724 cut_file (char const *file
)
728 if (STREQ (file
, "-"))
730 have_read_stdin
= true;
735 stream
= fopen (file
, "r");
738 error (0, errno
, "%s", file
);
743 fadvise (stream
, FADVISE_SEQUENTIAL
);
749 error (0, errno
, "%s", file
);
752 if (STREQ (file
, "-"))
753 clearerr (stream
); /* Also clear EOF. */
754 else if (fclose (stream
) == EOF
)
756 error (0, errno
, "%s", file
);
763 main (int argc
, char **argv
)
767 bool delim_specified
= false;
768 char *spec_list_string
IF_LINT ( = NULL
);
770 initialize_main (&argc
, &argv
);
771 set_program_name (argv
[0]);
772 setlocale (LC_ALL
, "");
773 bindtextdomain (PACKAGE
, LOCALEDIR
);
774 textdomain (PACKAGE
);
776 atexit (close_stdout
);
778 operating_mode
= undefined_mode
;
780 /* By default, all non-delimited lines are printed. */
781 suppress_non_delimited
= false;
784 have_read_stdin
= false;
786 while ((optc
= getopt_long (argc
, argv
, "b:c:d:f:ns", longopts
, NULL
)) != -1)
792 /* Build the byte list. */
793 if (operating_mode
!= undefined_mode
)
794 FATAL_ERROR (_("only one type of list may be specified"));
795 operating_mode
= byte_mode
;
796 spec_list_string
= optarg
;
800 /* Build the field list. */
801 if (operating_mode
!= undefined_mode
)
802 FATAL_ERROR (_("only one type of list may be specified"));
803 operating_mode
= field_mode
;
804 spec_list_string
= optarg
;
809 /* Interpret -d '' to mean 'use the NUL byte as the delimiter.' */
810 if (optarg
[0] != '\0' && optarg
[1] != '\0')
811 FATAL_ERROR (_("the delimiter must be a single character"));
813 delim_specified
= true;
816 case OUTPUT_DELIMITER_OPTION
:
817 output_delimiter_specified
= true;
818 /* Interpret --output-delimiter='' to mean
819 'use the NUL byte as the delimiter.' */
820 output_delimiter_length
= (optarg
[0] == '\0'
821 ? 1 : strlen (optarg
));
822 output_delimiter_string
= xstrdup (optarg
);
829 suppress_non_delimited
= true;
832 case COMPLEMENT_OPTION
:
836 case_GETOPT_HELP_CHAR
;
838 case_GETOPT_VERSION_CHAR (PROGRAM_NAME
, AUTHORS
);
841 usage (EXIT_FAILURE
);
845 if (operating_mode
== undefined_mode
)
846 FATAL_ERROR (_("you must specify a list of bytes, characters, or fields"));
848 if (delim_specified
&& operating_mode
!= field_mode
)
849 FATAL_ERROR (_("an input delimiter may be specified only\
850 when operating on fields"));
852 if (suppress_non_delimited
&& operating_mode
!= field_mode
)
853 FATAL_ERROR (_("suppressing non-delimited lines makes sense\n\
854 \tonly when operating on fields"));
856 if (output_delimiter_specified
)
858 range_start_ht
= hash_initialize (HT_RANGE_START_INDEX_INITIAL_CAPACITY
,
860 hash_compare_ints
, NULL
);
861 if (range_start_ht
== NULL
)
866 if (! set_fields (spec_list_string
))
868 if (operating_mode
== field_mode
)
869 FATAL_ERROR (_("missing list of fields"));
871 FATAL_ERROR (_("missing list of positions"));
874 if (!delim_specified
)
877 if (output_delimiter_string
== NULL
)
879 static char dummy
[2];
882 output_delimiter_string
= dummy
;
883 output_delimiter_length
= 1;
889 for (ok
= true; optind
< argc
; optind
++)
890 ok
&= cut_file (argv
[optind
]);
893 hash_free (range_start_ht
);
895 if (have_read_stdin
&& fclose (stdin
) == EOF
)
897 error (0, errno
, "-");
901 exit (ok
? EXIT_SUCCESS
: EXIT_FAILURE
);