1 /* diff - compare files line by line
3 Copyright (C) 1988, 1989, 1992, 1993, 1994, 1996, 1998, 2001, 2002,
4 2004 Free Software Foundation, Inc.
6 This file is part of GNU DIFF.
8 GNU DIFF is free software; you can redistribute it and/or modify
9 it under the terms of the GNU General Public License as published by
10 the Free Software Foundation; either version 2, or (at your option)
13 GNU DIFF is distributed in the hope that it will be useful,
14 but WITHOUT ANY WARRANTY; without even the implied warranty of
15 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
16 See the GNU General Public License for more details.
18 You should have received a copy of the GNU General Public License
19 along with GNU DIFF; see the file COPYING.
20 If not, write to the Free Software Foundation,
21 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */
32 #include <file-type.h>
35 #include <hard-locale.h>
40 #include <version-etc.h>
43 #ifndef GUTTER_WIDTH_MINIMUM
44 # define GUTTER_WIDTH_MINIMUM 3
49 char *regexps; /* chars representing disjunction of the regexps */
50 size_t len; /* chars used in `regexps' */
51 size_t size; /* size malloc'ed for `regexps'; 0 if not malloc'ed */
52 bool multiple_regexps;/* Does `regexps' represent a disjunction? */
53 struct re_pattern_buffer *buf;
56 static int compare_files (struct comparison const *, char const *, char const *);
57 static void add_regexp (struct regexp_list *, char const *);
58 static void summarize_regexp_list (struct regexp_list *);
59 static void specify_style (enum output_style);
60 static void specify_value (char const **, char const *, char const *);
61 static void try_help (char const *, char const *) __attribute__((noreturn));
62 static void check_stdout (void);
63 static void usage (void);
65 /* If comparing directories, compare their common subdirectories
67 static bool recursive;
69 /* In context diffs, show previous lines that match these regexps. */
70 static struct regexp_list function_regexp_list;
72 /* Ignore changes affecting only lines that match these regexps. */
73 static struct regexp_list ignore_regexp_list;
76 /* Use binary I/O when reading and writing data (--binary).
77 On POSIX hosts, this has no effect. */
80 enum { binary = true };
83 /* When comparing directories, if a file appears only in one
84 directory, treat it as present but empty in the other (-N).
85 Then `patch' would create the file with appropriate contents. */
88 /* When comparing directories, if a file appears only in the second
89 directory of the two, treat it as present but empty in the other
90 (--unidirectional-new-file).
91 Then `patch' would create the file with appropriate contents. */
92 static bool unidirectional_new_file;
94 /* Report files compared that are the same (-s).
95 Normally nothing is output when that happens. */
96 static bool report_identical_files;
99 /* Return a string containing the command options with which diff was invoked.
100 Spaces appear between what were separate ARGV-elements.
101 There is a space at the beginning but none at the end.
102 If there were no options, the result is an empty string.
104 Arguments: OPTIONVEC, a vector containing separate ARGV-elements, and COUNT,
105 the length of that vector. */
108 option_list (char **optionvec, int count)
115 for (i = 0; i < count; i++)
116 size += 1 + quote_system_arg ((char *) 0, optionvec[i]);
118 p = result = xmalloc (size);
120 for (i = 0; i < count; i++)
123 p += quote_system_arg (p, optionvec[i]);
131 /* Return an option value suitable for add_exclude. */
134 exclude_options (void)
136 return EXCLUDE_WILDCARDS | (ignore_file_name_case ? FNM_CASEFOLD : 0);
139 static char const shortopts[] =
140 "0123456789abBcC:dD:eEfF:hHiI:lL:nNopPqrsS:tTuU:vwW:x:X:y";
142 /* Values for long options that do not have single-letter equivalents. */
145 BINARY_OPTION = CHAR_MAX + 1,
148 HORIZON_LINES_OPTION,
149 IGNORE_FILE_NAME_CASE_OPTION,
150 INHIBIT_HUNK_MERGE_OPTION,
153 NO_IGNORE_FILE_NAME_CASE_OPTION,
155 SDIFF_MERGE_ASSIST_OPTION,
156 STRIP_TRAILING_CR_OPTION,
157 SUPPRESS_COMMON_LINES_OPTION,
161 /* These options must be in sequence. */
162 UNCHANGED_LINE_FORMAT_OPTION,
163 OLD_LINE_FORMAT_OPTION,
164 NEW_LINE_FORMAT_OPTION,
166 /* These options must be in sequence. */
167 UNCHANGED_GROUP_FORMAT_OPTION,
168 OLD_GROUP_FORMAT_OPTION,
169 NEW_GROUP_FORMAT_OPTION,
170 CHANGED_GROUP_FORMAT_OPTION
173 static char const group_format_option[][sizeof "--unchanged-group-format"] =
175 "--unchanged-group-format",
176 "--old-group-format",
177 "--new-group-format",
178 "--changed-group-format"
181 static char const line_format_option[][sizeof "--unchanged-line-format"] =
183 "--unchanged-line-format",
188 static struct option const longopts[] =
190 {"binary", 0, 0, BINARY_OPTION},
191 {"brief", 0, 0, 'q'},
192 {"changed-group-format", 1, 0, CHANGED_GROUP_FORMAT_OPTION},
193 {"context", 2, 0, 'C'},
195 {"exclude", 1, 0, 'x'},
196 {"exclude-from", 1, 0, 'X'},
197 {"expand-tabs", 0, 0, 't'},
198 {"forward-ed", 0, 0, 'f'},
199 {"from-file", 1, 0, FROM_FILE_OPTION},
200 {"help", 0, 0, HELP_OPTION},
201 {"horizon-lines", 1, 0, HORIZON_LINES_OPTION},
202 {"ifdef", 1, 0, 'D'},
203 {"ignore-all-space", 0, 0, 'w'},
204 {"ignore-blank-lines", 0, 0, 'B'},
205 {"ignore-case", 0, 0, 'i'},
206 {"ignore-file-name-case", 0, 0, IGNORE_FILE_NAME_CASE_OPTION},
207 {"ignore-matching-lines", 1, 0, 'I'},
208 {"ignore-space-change", 0, 0, 'b'},
209 {"ignore-tab-expansion", 0, 0, 'E'},
210 {"inhibit-hunk-merge", 0, 0, INHIBIT_HUNK_MERGE_OPTION},
211 {"initial-tab", 0, 0, 'T'},
212 {"label", 1, 0, 'L'},
213 {"left-column", 0, 0, LEFT_COLUMN_OPTION},
214 {"line-format", 1, 0, LINE_FORMAT_OPTION},
215 {"minimal", 0, 0, 'd'},
216 {"new-file", 0, 0, 'N'},
217 {"new-group-format", 1, 0, NEW_GROUP_FORMAT_OPTION},
218 {"new-line-format", 1, 0, NEW_LINE_FORMAT_OPTION},
219 {"no-ignore-file-name-case", 0, 0, NO_IGNORE_FILE_NAME_CASE_OPTION},
220 {"normal", 0, 0, NORMAL_OPTION},
221 {"old-group-format", 1, 0, OLD_GROUP_FORMAT_OPTION},
222 {"old-line-format", 1, 0, OLD_LINE_FORMAT_OPTION},
223 {"paginate", 0, 0, 'l'},
225 {"recursive", 0, 0, 'r'},
226 {"report-identical-files", 0, 0, 's'},
227 {"sdiff-merge-assist", 0, 0, SDIFF_MERGE_ASSIST_OPTION},
228 {"show-c-function", 0, 0, 'p'},
229 {"show-function-line", 1, 0, 'F'},
230 {"side-by-side", 0, 0, 'y'},
231 {"speed-large-files", 0, 0, 'H'},
232 {"starting-file", 1, 0, 'S'},
233 {"strip-trailing-cr", 0, 0, STRIP_TRAILING_CR_OPTION},
234 {"suppress-common-lines", 0, 0, SUPPRESS_COMMON_LINES_OPTION},
235 {"tabsize", 1, 0, TABSIZE_OPTION},
237 {"to-file", 1, 0, TO_FILE_OPTION},
238 {"unchanged-group-format", 1, 0, UNCHANGED_GROUP_FORMAT_OPTION},
239 {"unchanged-line-format", 1, 0, UNCHANGED_LINE_FORMAT_OPTION},
240 {"unidirectional-new-file", 0, 0, 'P'},
241 {"unified", 2, 0, 'U'},
242 {"version", 0, 0, 'v'},
243 {"width", 1, 0, 'W'},
248 main (int argc, char **argv)
250 int exit_status = EXIT_SUCCESS;
255 bool explicit_context = false;
257 bool show_c_function = false;
258 char const *from_file = 0;
259 char const *to_file = 0;
263 /* Do our initializations. */
265 initialize_main (&argc, &argv);
266 program_name = argv[0];
267 setlocale (LC_ALL, "");
268 textdomain (PACKAGE);
270 function_regexp_list.buf = &function_regexp;
271 ignore_regexp_list.buf = &ignore_regexp;
272 re_set_syntax (RE_SYNTAX_GREP);
273 excluded = new_exclude ();
275 prepend_default_options (getenv ("DIFF_OPTIONS"), &argc, &argv);
277 /* Decode the options. */
279 while ((c = getopt_long (argc, argv, shortopts, longopts, 0)) != -1)
296 if (! ISDIGIT (prev))
298 else if (LIN_MAX / 10 < ocontext
299 || ((ocontext = 10 * ocontext + c - '0') < 0))
308 if (ignore_white_space < IGNORE_SPACE_CHANGE)
309 ignore_white_space = IGNORE_SPACE_CHANGE;
313 ignore_blank_lines = true;
321 numval = strtoumax (optarg, &numend, 10);
323 try_help ("invalid context length `%s'", optarg);
324 if (LIN_MAX < numval)
330 specify_style (c == 'U' ? OUTPUT_UNIFIED : OUTPUT_CONTEXT);
331 if (context < numval)
333 explicit_context = true;
338 specify_style (OUTPUT_CONTEXT);
348 specify_style (OUTPUT_IFDEF);
350 static char const C_ifdef_group_formats[] =
351 "%%=%c#ifndef %s\n%%<#endif /* ! %s */\n%c#ifdef %s\n%%>#endif /* %s */\n%c#ifndef %s\n%%<#else /* %s */\n%%>#endif /* %s */\n";
352 char *b = xmalloc (sizeof C_ifdef_group_formats
353 + 7 * strlen (optarg) - 14 /* 7*"%s" */
354 - 8 /* 5*"%%" + 3*"%c" */);
355 sprintf (b, C_ifdef_group_formats,
359 optarg, optarg, optarg);
360 for (i = 0; i < sizeof group_format / sizeof *group_format; i++)
362 specify_value (&group_format[i], b, "-D");
369 specify_style (OUTPUT_ED);
373 if (ignore_white_space < IGNORE_TAB_EXPANSION)
374 ignore_white_space = IGNORE_TAB_EXPANSION;
378 specify_style (OUTPUT_FORWARD_ED);
382 add_regexp (&function_regexp_list, optarg);
386 /* Split the files into chunks for faster processing.
387 Usually does not change the result.
389 This currently has no effect. */
393 speed_large_files = true;
401 add_regexp (&ignore_regexp_list, optarg);
406 try_help ("pagination not supported on this host", 0);
409 /* Pagination requires forking and waiting, and
410 System V fork+wait does not work if SIGCHLD is ignored. */
411 signal (SIGCHLD, SIG_DFL);
417 file_label[0] = optarg;
418 else if (!file_label[1])
419 file_label[1] = optarg;
421 fatal ("too many file label options");
425 specify_style (OUTPUT_RCS);
433 /* Output in the old tradition style. */
434 specify_style (OUTPUT_NORMAL);
438 show_c_function = true;
439 add_regexp (&function_regexp_list, "^[[:alpha:]$_]");
443 unidirectional_new_file = true;
455 report_identical_files = true;
459 specify_value (&starting_file, optarg, "-S");
471 specify_style (OUTPUT_UNIFIED);
477 version_etc (stdout, "diff", PACKAGE_NAME, PACKAGE_VERSION,
478 "Paul Eggert", "Mike Haertel", "David Hayes",
479 "Richard Stallman", "Len Tower", (char *) 0);
484 ignore_white_space = IGNORE_ALL_SPACE;
488 add_exclude (excluded, optarg, exclude_options ());
492 if (add_exclude_file (add_exclude, excluded, optarg,
493 exclude_options (), '\n'))
494 pfatal_with_name (optarg);
498 specify_style (OUTPUT_SDIFF);
502 numval = strtoumax (optarg, &numend, 10);
503 if (! (0 < numval && numval <= SIZE_MAX) || *numend)
504 try_help ("invalid width `%s'", optarg);
508 fatal ("conflicting width options");
516 set_binary_mode (STDOUT_FILENO, true);
520 case FROM_FILE_OPTION:
521 specify_value (&from_file, optarg, "--from-file");
529 case HORIZON_LINES_OPTION:
530 numval = strtoumax (optarg, &numend, 10);
532 try_help ("invalid horizon length `%s'", optarg);
533 horizon_lines = MAX (horizon_lines, MIN (numval, LIN_MAX));
536 case IGNORE_FILE_NAME_CASE_OPTION:
537 ignore_file_name_case = true;
540 case INHIBIT_HUNK_MERGE_OPTION:
541 /* This option is obsolete, but accept it for backward
545 case LEFT_COLUMN_OPTION:
549 case LINE_FORMAT_OPTION:
550 specify_style (OUTPUT_IFDEF);
551 for (i = 0; i < sizeof line_format / sizeof *line_format; i++)
552 specify_value (&line_format[i], optarg, "--line-format");
555 case NO_IGNORE_FILE_NAME_CASE_OPTION:
556 ignore_file_name_case = false;
560 specify_style (OUTPUT_NORMAL);
563 case SDIFF_MERGE_ASSIST_OPTION:
564 specify_style (OUTPUT_SDIFF);
565 sdiff_merge_assist = true;
568 case STRIP_TRAILING_CR_OPTION:
569 strip_trailing_cr = true;
572 case SUPPRESS_COMMON_LINES_OPTION:
573 suppress_common_lines = true;
577 numval = strtoumax (optarg, &numend, 10);
578 if (! (0 < numval && numval <= SIZE_MAX) || *numend)
579 try_help ("invalid tabsize `%s'", optarg);
580 if (tabsize != numval)
583 fatal ("conflicting tabsize options");
589 specify_value (&to_file, optarg, "--to-file");
592 case UNCHANGED_LINE_FORMAT_OPTION:
593 case OLD_LINE_FORMAT_OPTION:
594 case NEW_LINE_FORMAT_OPTION:
595 specify_style (OUTPUT_IFDEF);
596 c -= UNCHANGED_LINE_FORMAT_OPTION;
597 specify_value (&line_format[c], optarg, line_format_option[c]);
600 case UNCHANGED_GROUP_FORMAT_OPTION:
601 case OLD_GROUP_FORMAT_OPTION:
602 case NEW_GROUP_FORMAT_OPTION:
603 case CHANGED_GROUP_FORMAT_OPTION:
604 specify_style (OUTPUT_IFDEF);
605 c -= UNCHANGED_GROUP_FORMAT_OPTION;
606 specify_value (&group_format[c], optarg, group_format_option[c]);
615 if (output_style == OUTPUT_UNSPECIFIED)
619 specify_style (OUTPUT_CONTEXT);
624 specify_style (OUTPUT_NORMAL);
627 if (output_style != OUTPUT_CONTEXT || hard_locale (LC_TIME))
630 time_format = "%Y-%m-%d %H:%M:%S.%N %z";
632 time_format = "%Y-%m-%d %H:%M:%S %z";
637 /* See POSIX 1003.1-2001 for this format. */
638 time_format = "%a %b %e %T %Y";
643 bool modern_usage = 200112 <= posix2_version ();
645 if ((output_style == OUTPUT_CONTEXT
646 || output_style == OUTPUT_UNIFIED)
647 && (context < ocontext
648 || (ocontext < context && ! explicit_context)))
653 _("`-%ld' option is obsolete; use `-%c %ld'"),
655 output_style == OUTPUT_CONTEXT ? 'C' : 'U',
656 (long int) ocontext);
665 error (0, 0, _("`-%ld' option is obsolete; omit it"),
666 (long int) ocontext);
678 /* Maximize first the half line width, and then the gutter width,
679 according to the following constraints:
681 1. Two half lines plus a gutter must fit in a line.
682 2. If the half line width is nonzero:
683 a. The gutter width is at least GUTTER_WIDTH_MINIMUM.
684 b. If tabs are not expanded to spaces,
685 a half line plus a gutter is an integral number of tabs,
686 so that tabs in the right column line up. */
688 intmax_t t = expand_tabs ? 1 : tabsize;
690 intmax_t off = (w + t + GUTTER_WIDTH_MINIMUM) / (2 * t) * t;
691 sdiff_half_width = MAX (0, MIN (off - GUTTER_WIDTH_MINIMUM, w - off)),
692 sdiff_column2_offset = sdiff_half_width ? off : w;
695 /* Make the horizon at least as large as the context, so that
696 shift_boundaries has more freedom to shift the first and last hunks. */
697 if (horizon_lines < context)
698 horizon_lines = context;
700 summarize_regexp_list (&function_regexp_list);
701 summarize_regexp_list (&ignore_regexp_list);
703 if (output_style == OUTPUT_IFDEF)
705 for (i = 0; i < sizeof line_format / sizeof *line_format; i++)
707 line_format[i] = "%l\n";
708 if (!group_format[OLD])
710 = group_format[CHANGED] ? group_format[CHANGED] : "%<";
711 if (!group_format[NEW])
713 = group_format[CHANGED] ? group_format[CHANGED] : "%>";
714 if (!group_format[UNCHANGED])
715 group_format[UNCHANGED] = "%=";
716 if (!group_format[CHANGED])
717 group_format[CHANGED] = concat (group_format[OLD],
718 group_format[NEW], "");
721 no_diff_means_no_output =
722 (output_style == OUTPUT_IFDEF ?
723 (!*group_format[UNCHANGED]
724 || (strcmp (group_format[UNCHANGED], "%=") == 0
725 && !*line_format[UNCHANGED]))
726 : (output_style != OUTPUT_SDIFF) | suppress_common_lines);
728 files_can_be_treated_as_binary =
730 & ~ (ignore_blank_lines | ignore_case | strip_trailing_cr
731 | (ignore_regexp_list.regexps || ignore_white_space)));
733 switch_string = option_list (argv + 1, optind - 1);
738 fatal ("--from-file and --to-file both specified");
740 for (; optind < argc; optind++)
742 int status = compare_files ((struct comparison *) 0,
743 from_file, argv[optind]);
744 if (exit_status < status)
745 exit_status = status;
751 for (; optind < argc; optind++)
753 int status = compare_files ((struct comparison *) 0,
754 argv[optind], to_file);
755 if (exit_status < status)
756 exit_status = status;
760 if (argc - optind != 2)
762 if (argc - optind < 2)
763 try_help ("missing operand after `%s'", argv[argc - 1]);
765 try_help ("extra operand `%s'", argv[optind + 2]);
768 exit_status = compare_files ((struct comparison *) 0,
769 argv[optind], argv[optind + 1]);
773 /* Print any messages that were saved up for last. */
774 print_message_queue ();
781 /* Append to REGLIST the regexp PATTERN. */
784 add_regexp (struct regexp_list *reglist, char const *pattern)
786 size_t patlen = strlen (pattern);
787 char const *m = re_compile_pattern (pattern, patlen, reglist->buf);
790 error (0, 0, "%s: %s", pattern, m);
793 char *regexps = reglist->regexps;
794 size_t len = reglist->len;
795 bool multiple_regexps = reglist->multiple_regexps = regexps != 0;
796 size_t newlen = reglist->len = len + 2 * multiple_regexps + patlen;
797 size_t size = reglist->size;
805 while (size <= newlen);
807 reglist->size = size;
808 reglist->regexps = regexps = xrealloc (regexps, size);
810 if (multiple_regexps)
812 regexps[len++] = '\\';
813 regexps[len++] = '|';
815 memcpy (regexps + len, pattern, patlen + 1);
819 /* Ensure that REGLIST represents the disjunction of its regexps.
820 This is done here, rather than earlier, to avoid O(N^2) behavior. */
823 summarize_regexp_list (struct regexp_list *reglist)
825 if (reglist->regexps)
827 /* At least one regexp was specified. Allocate a fastmap for it. */
828 reglist->buf->fastmap = xmalloc (1 << CHAR_BIT);
829 if (reglist->multiple_regexps)
831 /* Compile the disjunction of the regexps.
832 (If just one regexp was specified, it is already compiled.) */
833 char const *m = re_compile_pattern (reglist->regexps, reglist->len,
836 error (EXIT_TROUBLE, 0, "%s: %s", reglist->regexps, m);
842 try_help (char const *reason_msgid, char const *operand)
845 error (0, 0, _(reason_msgid), operand);
846 error (EXIT_TROUBLE, 0, _("Try `%s --help' for more information."),
855 fatal ("write failed");
856 else if (fclose (stdout) != 0)
857 pfatal_with_name (_("standard output"));
860 static char const * const option_help_msgid[] = {
861 N_("Compare files line by line."),
863 N_("-i --ignore-case Ignore case differences in file contents."),
864 N_("--ignore-file-name-case Ignore case when comparing file names."),
865 N_("--no-ignore-file-name-case Consider case when comparing file names."),
866 N_("-E --ignore-tab-expansion Ignore changes due to tab expansion."),
867 N_("-b --ignore-space-change Ignore changes in the amount of white space."),
868 N_("-w --ignore-all-space Ignore all white space."),
869 N_("-B --ignore-blank-lines Ignore changes whose lines are all blank."),
870 N_("-I RE --ignore-matching-lines=RE Ignore changes whose lines all match RE."),
871 N_("--strip-trailing-cr Strip trailing carriage return on input."),
873 N_("--binary Read and write data in binary mode."),
875 N_("-a --text Treat all files as text."),
877 N_("-c -C NUM --context[=NUM] Output NUM (default 3) lines of copied context.\n\
878 -u -U NUM --unified[=NUM] Output NUM (default 3) lines of unified context.\n\
879 --label LABEL Use LABEL instead of file name.\n\
880 -p --show-c-function Show which C function each change is in.\n\
881 -F RE --show-function-line=RE Show the most recent line matching RE."),
882 N_("-q --brief Output only whether files differ."),
883 N_("-e --ed Output an ed script."),
884 N_("--normal Output a normal diff."),
885 N_("-n --rcs Output an RCS format diff."),
886 N_("-y --side-by-side Output in two columns.\n\
887 -W NUM --width=NUM Output at most NUM (default 130) print columns.\n\
888 --left-column Output only the left column of common lines.\n\
889 --suppress-common-lines Do not output common lines."),
890 N_("-D NAME --ifdef=NAME Output merged file to show `#ifdef NAME' diffs."),
891 N_("--GTYPE-group-format=GFMT Similar, but format GTYPE input groups with GFMT."),
892 N_("--line-format=LFMT Similar, but format all input lines with LFMT."),
893 N_("--LTYPE-line-format=LFMT Similar, but format LTYPE input lines with LFMT."),
894 N_(" LTYPE is `old', `new', or `unchanged'. GTYPE is LTYPE or `changed'."),
895 N_(" GFMT may contain:\n\
896 %< lines from FILE1\n\
897 %> lines from FILE2\n\
898 %= lines common to FILE1 and FILE2\n\
899 %[-][WIDTH][.[PREC]]{doxX}LETTER printf-style spec for LETTER\n\
900 LETTERs are as follows for new group, lower case for old group:\n\
901 F first line number\n\
902 L last line number\n\
903 N number of lines = L-F+1\n\
906 N_(" LFMT may contain:\n\
907 %L contents of line\n\
908 %l contents of line, excluding any trailing newline\n\
909 %[-][WIDTH][.[PREC]]{doxX}n printf-style spec for input line number"),
910 N_(" Either GFMT or LFMT may contain:\n\
912 %c'C' the single character C\n\
913 %c'\\OOO' the character with octal code OOO"),
915 N_("-l --paginate Pass the output through `pr' to paginate it."),
916 N_("-t --expand-tabs Expand tabs to spaces in output."),
917 N_("-T --initial-tab Make tabs line up by prepending a tab."),
918 N_("--tabsize=NUM Tab stops are every NUM (default 8) print columns."),
920 N_("-r --recursive Recursively compare any subdirectories found."),
921 N_("-N --new-file Treat absent files as empty."),
922 N_("--unidirectional-new-file Treat absent first files as empty."),
923 N_("-s --report-identical-files Report when two files are the same."),
924 N_("-x PAT --exclude=PAT Exclude files that match PAT."),
925 N_("-X FILE --exclude-from=FILE Exclude files that match any pattern in FILE."),
926 N_("-S FILE --starting-file=FILE Start with FILE when comparing directories."),
927 N_("--from-file=FILE1 Compare FILE1 to all operands. FILE1 can be a directory."),
928 N_("--to-file=FILE2 Compare all operands to FILE2. FILE2 can be a directory."),
930 N_("--horizon-lines=NUM Keep NUM lines of the common prefix and suffix."),
931 N_("-d --minimal Try hard to find a smaller set of changes."),
932 N_("--speed-large-files Assume large files and many scattered small changes."),
934 N_("-v --version Output version info."),
935 N_("--help Output this help."),
937 N_("FILES are `FILE1 FILE2' or `DIR1 DIR2' or `DIR FILE...' or `FILE... DIR'."),
938 N_("If --from-file or --to-file is given, there are no restrictions on FILES."),
939 N_("If a FILE is `-', read standard input."),
940 N_("Exit status is 0 if inputs are the same, 1 if different, 2 if trouble."),
942 N_("Report bugs to <bug-gnu-utils@gnu.org>."),
949 char const * const *p;
951 printf (_("Usage: %s [OPTION]... FILES\n"), program_name);
953 for (p = option_help_msgid; *p; p++)
959 char const *msg = _(*p);
961 while ((nl = strchr (msg, '\n')))
963 int msglen = nl + 1 - msg;
964 printf (" %.*s", msglen, msg);
968 printf (" %s\n" + 2 * (*msg != ' ' && *msg != '-'), msg);
973 /* Set VAR to VALUE, reporting an OPTION error if this is a
976 specify_value (char const **var, char const *value, char const *option)
978 if (*var && strcmp (*var, value) != 0)
980 error (0, 0, _("conflicting %s option value `%s'"), option, value);
986 /* Set the output style to STYLE, diagnosing conflicts. */
988 specify_style (enum output_style style)
990 if (output_style != style)
992 output_style = style;
996 /* Set the last-modified time of *ST to be the current time. */
999 set_mtime_to_now (struct stat *st)
1003 # if HAVE_CLOCK_GETTIME && defined CLOCK_REALTIME
1004 if (clock_gettime (CLOCK_REALTIME, &st->st_mtim) == 0)
1008 # if HAVE_GETTIMEOFDAY
1010 struct timeval timeval;
1011 if (gettimeofday (&timeval, 0) == 0)
1013 st->st_mtime = timeval.tv_sec;
1014 st->st_mtim.ST_MTIM_NSEC = timeval.tv_usec * 1000;
1020 #endif /* ST_MTIM_NSEC */
1022 time (&st->st_mtime);
1025 /* Compare two files (or dirs) with parent comparison PARENT
1026 and names NAME0 and NAME1.
1027 (If PARENT is 0, then the first name is just NAME0, etc.)
1028 This is self-contained; it opens the files and closes them.
1030 Value is EXIT_SUCCESS if files are the same, EXIT_FAILURE if
1031 different, EXIT_TROUBLE if there is a problem opening them. */
1034 compare_files (struct comparison const *parent,
1038 struct comparison cmp;
1039 #define DIR_P(f) (S_ISDIR (cmp.file[f].stat.st_mode) != 0)
1041 int status = EXIT_SUCCESS;
1043 char *free0, *free1;
1045 /* If this is directory comparison, perhaps we have a file
1046 that exists only in one of the directories.
1047 If so, just print a message to that effect. */
1049 if (! ((name0 && name1)
1050 || (unidirectional_new_file && name1)
1053 char const *name = name0 == 0 ? name1 : name0;
1054 char const *dir = parent->file[name0 == 0].name;
1056 /* See POSIX 1003.1-2001 for this format. */
1057 message ("Only in %s: %s\n", dir, name);
1059 /* Return EXIT_FAILURE so that diff_dirs will return
1060 EXIT_FAILURE ("some files differ"). */
1061 return EXIT_FAILURE;
1064 memset (cmp.file, 0, sizeof cmp.file);
1065 cmp.parent = parent;
1067 /* cmp.file[f].desc markers */
1068 #define NONEXISTENT (-1) /* nonexistent file */
1069 #define UNOPENED (-2) /* unopened file (e.g. directory) */
1070 #define ERRNO_ENCODE(errno) (-3 - (errno)) /* encoded errno value */
1072 #define ERRNO_DECODE(desc) (-3 - (desc)) /* inverse of ERRNO_ENCODE */
1074 cmp.file[0].desc = name0 == 0 ? NONEXISTENT : UNOPENED;
1075 cmp.file[1].desc = name1 == 0 ? NONEXISTENT : UNOPENED;
1077 /* Now record the full name of each file, including nonexistent ones. */
1088 cmp.file[0].name = name0;
1089 cmp.file[1].name = name1;
1093 cmp.file[0].name = free0
1094 = dir_file_pathname (parent->file[0].name, name0);
1095 cmp.file[1].name = free1
1096 = dir_file_pathname (parent->file[1].name, name1);
1099 /* Stat the files. */
1101 for (f = 0; f < 2; f++)
1103 if (cmp.file[f].desc != NONEXISTENT)
1105 if (f && file_name_cmp (cmp.file[f].name, cmp.file[0].name) == 0)
1107 cmp.file[f].desc = cmp.file[0].desc;
1108 cmp.file[f].stat = cmp.file[0].stat;
1110 else if (strcmp (cmp.file[f].name, "-") == 0)
1112 cmp.file[f].desc = STDIN_FILENO;
1113 if (fstat (STDIN_FILENO, &cmp.file[f].stat) != 0)
1114 cmp.file[f].desc = ERRNO_ENCODE (errno);
1117 if (S_ISREG (cmp.file[f].stat.st_mode))
1119 off_t pos = lseek (STDIN_FILENO, (off_t) 0, SEEK_CUR);
1121 cmp.file[f].desc = ERRNO_ENCODE (errno);
1123 cmp.file[f].stat.st_size =
1124 MAX (0, cmp.file[f].stat.st_size - pos);
1127 /* POSIX 1003.1-2001 requires current time for
1129 set_mtime_to_now (&cmp.file[f].stat);
1132 else if (stat (cmp.file[f].name, &cmp.file[f].stat) != 0)
1133 cmp.file[f].desc = ERRNO_ENCODE (errno);
1137 /* Mark files as nonexistent as needed for -N and -P, if they are
1138 inaccessible empty regular files (the kind of files that 'patch'
1139 creates to indicate nonexistent backups), or if they are
1140 top-level files that do not exist but their counterparts do
1142 for (f = 0; f < 2; f++)
1143 if ((new_file || (f == 0 && unidirectional_new_file))
1144 && (cmp.file[f].desc == UNOPENED
1145 ? (S_ISREG (cmp.file[f].stat.st_mode)
1146 && ! (cmp.file[f].stat.st_mode & (S_IRWXU | S_IRWXG | S_IRWXO))
1147 && cmp.file[f].stat.st_size == 0)
1148 : (cmp.file[f].desc == ERRNO_ENCODE (ENOENT)
1150 && cmp.file[1 - f].desc == UNOPENED)))
1151 cmp.file[f].desc = NONEXISTENT;
1153 for (f = 0; f < 2; f++)
1154 if (cmp.file[f].desc == NONEXISTENT)
1156 memset (&cmp.file[f].stat, 0, sizeof cmp.file[f].stat);
1157 cmp.file[f].stat.st_mode = cmp.file[1 - f].stat.st_mode;
1160 for (f = 0; f < 2; f++)
1162 int e = ERRNO_DECODE (cmp.file[f].desc);
1166 perror_with_name (cmp.file[f].name);
1167 status = EXIT_TROUBLE;
1171 if (status == EXIT_SUCCESS && ! parent && DIR_P (0) != DIR_P (1))
1173 /* If one is a directory, and it was specified in the command line,
1174 use the file in that dir with the other file's basename. */
1176 int fnm_arg = DIR_P (0);
1177 int dir_arg = 1 - fnm_arg;
1178 char const *fnm = cmp.file[fnm_arg].name;
1179 char const *dir = cmp.file[dir_arg].name;
1180 char const *filename = cmp.file[dir_arg].name = free0
1181 = dir_file_pathname (dir, base_name (fnm));
1183 if (strcmp (fnm, "-") == 0)
1184 fatal ("cannot compare `-' to a directory");
1186 if (stat (filename, &cmp.file[dir_arg].stat) != 0)
1188 perror_with_name (filename);
1189 status = EXIT_TROUBLE;
1193 if (status != EXIT_SUCCESS)
1195 /* One of the files should exist but does not. */
1197 else if (cmp.file[0].desc == NONEXISTENT
1198 && cmp.file[1].desc == NONEXISTENT)
1200 /* Neither file "exists", so there's nothing to compare. */
1202 else if ((same_files
1203 = (cmp.file[0].desc != NONEXISTENT
1204 && cmp.file[1].desc != NONEXISTENT
1205 && 0 < same_file (&cmp.file[0].stat, &cmp.file[1].stat)
1206 && same_file_attributes (&cmp.file[0].stat,
1207 &cmp.file[1].stat)))
1208 && no_diff_means_no_output)
1210 /* The two named files are actually the same physical file.
1211 We know they are identical without actually reading them. */
1213 else if (DIR_P (0) & DIR_P (1))
1215 if (output_style == OUTPUT_IFDEF)
1216 fatal ("-D option not supported with directories");
1218 /* If both are directories, compare the files in them. */
1220 if (parent && !recursive)
1222 /* But don't compare dir contents one level down
1223 unless -r was specified.
1224 See POSIX 1003.1-2001 for this format. */
1225 message ("Common subdirectories: %s and %s\n",
1226 cmp.file[0].name, cmp.file[1].name);
1229 status = diff_dirs (&cmp, compare_files);
1231 else if ((DIR_P (0) | DIR_P (1))
1233 && (! S_ISREG (cmp.file[0].stat.st_mode)
1234 || ! S_ISREG (cmp.file[1].stat.st_mode))))
1236 if (cmp.file[0].desc == NONEXISTENT || cmp.file[1].desc == NONEXISTENT)
1238 /* We have a subdirectory that exists only in one directory. */
1240 if ((DIR_P (0) | DIR_P (1))
1243 || (unidirectional_new_file
1244 && cmp.file[0].desc == NONEXISTENT)))
1245 status = diff_dirs (&cmp, compare_files);
1249 = parent->file[cmp.file[0].desc == NONEXISTENT].name;
1251 /* See POSIX 1003.1-2001 for this format. */
1252 message ("Only in %s: %s\n", dir, name0);
1254 status = EXIT_FAILURE;
1259 /* We have two files that are not to be compared. */
1261 /* See POSIX 1003.1-2001 for this format. */
1262 message5 ("File %s is a %s while file %s is a %s\n",
1263 file_label[0] ? file_label[0] : cmp.file[0].name,
1264 file_type (&cmp.file[0].stat),
1265 file_label[1] ? file_label[1] : cmp.file[1].name,
1266 file_type (&cmp.file[1].stat));
1268 /* This is a difference. */
1269 status = EXIT_FAILURE;
1272 else if (files_can_be_treated_as_binary
1273 && S_ISREG (cmp.file[0].stat.st_mode)
1274 && S_ISREG (cmp.file[1].stat.st_mode)
1275 && cmp.file[0].stat.st_size != cmp.file[1].stat.st_size)
1277 message ("Files %s and %s differ\n",
1278 file_label[0] ? file_label[0] : cmp.file[0].name,
1279 file_label[1] ? file_label[1] : cmp.file[1].name);
1280 status = EXIT_FAILURE;
1284 /* Both exist and neither is a directory. */
1286 /* Open the files and record their descriptors. */
1288 if (cmp.file[0].desc == UNOPENED)
1289 if ((cmp.file[0].desc = open (cmp.file[0].name, O_RDONLY, 0)) < 0)
1291 perror_with_name (cmp.file[0].name);
1292 status = EXIT_TROUBLE;
1294 if (cmp.file[1].desc == UNOPENED)
1297 cmp.file[1].desc = cmp.file[0].desc;
1298 else if ((cmp.file[1].desc = open (cmp.file[1].name, O_RDONLY, 0))
1301 perror_with_name (cmp.file[1].name);
1302 status = EXIT_TROUBLE;
1306 #if HAVE_SETMODE_DOS
1308 for (f = 0; f < 2; f++)
1309 if (0 <= cmp.file[f].desc)
1310 set_binary_mode (cmp.file[f].desc, true);
1313 /* Compare the files, if no error was found. */
1315 if (status == EXIT_SUCCESS)
1316 status = diff_2_files (&cmp);
1318 /* Close the file descriptors. */
1320 if (0 <= cmp.file[0].desc && close (cmp.file[0].desc) != 0)
1322 perror_with_name (cmp.file[0].name);
1323 status = EXIT_TROUBLE;
1325 if (0 <= cmp.file[1].desc && cmp.file[0].desc != cmp.file[1].desc
1326 && close (cmp.file[1].desc) != 0)
1328 perror_with_name (cmp.file[1].name);
1329 status = EXIT_TROUBLE;
1333 /* Now the comparison has been done, if no error prevented it,
1334 and STATUS is the value this function will return. */
1336 if (status == EXIT_SUCCESS)
1338 if (report_identical_files && !DIR_P (0))
1339 message ("Files %s and %s are identical\n",
1340 file_label[0] ? file_label[0] : cmp.file[0].name,
1341 file_label[1] ? file_label[1] : cmp.file[1].name);
1345 /* Flush stdout so that the user sees differences immediately.
1346 This can hurt performance, unfortunately. */
1347 if (fflush (stdout) != 0)
1348 pfatal_with_name (_("standard output"));