diff --git a/usr.bin/diff/diff.1 b/usr.bin/diff/diff.1 index e04aeb5d6d67..d3978158d1d1 100644 --- a/usr.bin/diff/diff.1 +++ b/usr.bin/diff/diff.1 @@ -1,743 +1,752 @@ .\" $OpenBSD: diff.1,v 1.47 2015/11/24 19:35:41 jmc Exp $ .\" .\" Copyright (c) 1980, 1990, 1993 .\" The Regents of the University of California. All rights reserved. .\" .\" Redistribution and use in source and binary forms, with or without .\" modification, are permitted provided that the following conditions .\" are met: .\" 1. Redistributions of source code must retain the above copyright .\" notice, this list of conditions and the following disclaimer. .\" 2. Redistributions in binary form must reproduce the above copyright .\" notice, this list of conditions and the following disclaimer in the .\" documentation and/or other materials provided with the distribution. .\" 3. Neither the name of the University nor the names of its contributors .\" may be used to endorse or promote products derived from this software .\" without specific prior written permission. .\" .\" THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND .\" ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE .\" IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE .\" ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE .\" FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL .\" DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS .\" OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) .\" HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT .\" LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY .\" OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF .\" SUCH DAMAGE. .\" .\" @(#)diff.1 8.1 (Berkeley) 6/30/93 .\" -.Dd March 10, 2022 +.Dd February 26, 2024 .Dt DIFF 1 .Os .Sh NAME .Nm diff .Nd differential file and directory comparator .Sh SYNOPSIS .Nm diff .Op Fl aBbdipTtw .Oo .Fl c | e | f | .Fl n | q | u | y .Oc .Op Fl -brief .Op Fl -color Ns = Ns Ar when .Op Fl -changed-group-format Ar GFMT .Op Fl -ed .Op Fl -expand-tabs .Op Fl -forward-ed .Op Fl -ignore-all-space .Op Fl -ignore-case .Op Fl -ignore-space-change .Op Fl -initial-tab .Op Fl -minimal .Op Fl -no-ignore-file-name-case .Op Fl -normal .Op Fl -rcs .Op Fl -show-c-function .Op Fl -starting-file .Op Fl -speed-large-files .Op Fl -strip-trailing-cr .Op Fl -tabsize Ar number .Op Fl -text .Op Fl -unified .Op Fl I Ar pattern | Fl -ignore-matching-lines Ar pattern .Op Fl F Ar pattern | Fl -show-function-line Ar pattern .Op Fl L Ar label | Fl -label Ar label .Ar file1 file2 .Nm diff .Op Fl aBbdilpTtw .Op Fl I Ar pattern | Fl -ignore-matching-lines Ar pattern .Op Fl F Ar pattern | Fl -show-function-line Ar pattern .Op Fl L Ar label | Fl -label Ar label .Op Fl -brief .Op Fl -color Ns = Ns Ar when .Op Fl -changed-group-format Ar GFMT .Op Fl -ed .Op Fl -expand-tabs .Op Fl -forward-ed .Op Fl -ignore-all-space .Op Fl -ignore-case .Op Fl -ignore-space-change .Op Fl -initial-tab .Op Fl -minimal .Op Fl -no-ignore-file-name-case .Op Fl -normal .Op Fl -paginate .Op Fl -rcs .Op Fl -show-c-function .Op Fl -speed-large-files .Op Fl -starting-file .Op Fl -strip-trailing-cr .Op Fl -tabsize Ar number .Op Fl -text .Fl C Ar number | -context Ar number .Ar file1 file2 .Nm diff .Op Fl aBbdiltw .Op Fl I Ar pattern | Fl -ignore-matching-lines Ar pattern .Op Fl -brief .Op Fl -color Ns = Ns Ar when .Op Fl -changed-group-format Ar GFMT .Op Fl -ed .Op Fl -expand-tabs .Op Fl -forward-ed .Op Fl -ignore-all-space .Op Fl -ignore-case .Op Fl -ignore-space-change .Op Fl -initial-tab .Op Fl -minimal .Op Fl -no-ignore-file-name-case .Op Fl -normal .Op Fl -paginate .Op Fl -rcs .Op Fl -show-c-function .Op Fl -speed-large-files .Op Fl -starting-file .Op Fl -strip-trailing-cr .Op Fl -tabsize Ar number .Op Fl -text .Fl D Ar string | Fl -ifdef Ar string .Ar file1 file2 .Nm diff .Op Fl aBbdilpTtw .Op Fl I Ar pattern | Fl -ignore-matching-lines Ar pattern .Op Fl F Ar pattern | Fl -show-function-line Ar pattern .Op Fl L Ar label | Fl -label Ar label .Op Fl -brief .Op Fl -color Ns = Ns Ar when .Op Fl -changed-group-format Ar GFMT .Op Fl -ed .Op Fl -expand-tabs .Op Fl -forward-ed .Op Fl -ignore-all-space .Op Fl -ignore-case .Op Fl -ignore-space-change .Op Fl -initial-tab .Op Fl -minimal .Op Fl -no-ignore-file-name-case .Op Fl -normal .Op Fl -paginate .Op Fl -rcs .Op Fl -show-c-function .Op Fl -speed-large-files .Op Fl -starting-file .Op Fl -strip-trailing-cr .Op Fl -tabsize Ar number .Op Fl -text .Fl U Ar number | Fl -unified Ar number .Ar file1 file2 .Nm diff .Op Fl aBbdilNPprsTtw .Oo .Fl c | e | f | .Fl n | q | u .Oc .Op Fl -brief .Op Fl -color Ns = Ns Ar when .Op Fl -changed-group-format Ar GFMT .Op Fl -context .Op Fl -ed .Op Fl -expand-tabs .Op Fl -forward-ed .Op Fl -ignore-all-space .Op Fl -ignore-case .Op Fl -ignore-space-change .Op Fl -initial-tab .Op Fl -minimal .Op Fl -new-file .Op Fl -no-ignore-file-name-case .Op Fl -normal .Op Fl -paginate .Op Fl -rcs .Op Fl -recursive .Op Fl -report-identical-files .Op Fl -show-c-function .Op Fl -speed-large-files .Op Fl -strip-trailing-cr .Op Fl -tabsize Ar number .Op Fl -text .Op Fl -unidirectional-new-file .Op Fl -unified .Op Fl I Ar pattern | Fl -ignore-matching-lines Ar pattern .Op Fl F Ar pattern | Fl -show-function-line Ar pattern .Bk -words .Op Fl L Ar label | Fl -label Ar label .Op Fl S Ar name | Fl -starting-file Ar name .Op Fl X Ar file | Fl -exclude-from Ar file .Op Fl x Ar pattern | Fl -exclude Ar pattern .Ek .Ar dir1 dir2 .Nm diff .Op Fl aBbditwW .Op Fl -color Ns = Ns Ar when .Op Fl -expand-tabs .Op Fl -ignore-all-blanks .Op Fl -ignore-blank-lines .Op Fl -ignore-case .Op Fl -minimal .Op Fl -no-ignore-file-name-case .Op Fl -strip-trailing-cr .Op Fl -suppress-common-lines .Op Fl -tabsize Ar number .Op Fl -text .Op Fl -width .Fl y | Fl -side-by-side .Ar file1 file2 .Nm diff .Op Fl -help .Op Fl -version .Sh DESCRIPTION The .Nm utility compares the contents of .Ar file1 and .Ar file2 and writes to the standard output the list of changes necessary to convert one file into the other. No output is produced if the files are identical. .Pp Output options (mutually exclusive): .Bl -tag -width Ds .It Fl C Ar number Fl -context Ar number Like .Fl c but produces a diff with .Ar number lines of context. .It Fl c Produces a diff with 3 lines of context. With .Fl c the output format is modified slightly: the output begins with identification of the files involved and their creation dates and then each change is separated by a line with fifteen .Li * Ns 's . The lines removed from .Ar file1 are marked with .Sq \&-\ \& ; those added to .Ar file2 are marked .Sq +\ \& . Lines which are changed from one file to the other are marked in both files with .Sq !\ \& . Changes which lie within 3 lines of each other are grouped together on output. .It Fl D Ar string Fl -ifdef Ar string Creates a merged version of .Ar file1 and .Ar file2 on the standard output, with C preprocessor controls included so that a compilation of the result without defining .Ar string is equivalent to compiling .Ar file1 , while defining .Ar string will yield .Ar file2 . .It Fl e -ed Produces output in a form suitable as input for the editor utility, .Xr ed 1 , which can then be used to convert file1 into file2. .Pp Extra commands are added to the output when comparing directories with .Fl e , so that the result is a .Xr sh 1 script for converting text files which are common to the two directories from their state in .Ar dir1 to their state in .Ar dir2 . .It Fl f -forward-ed Identical output to that of the .Fl e flag, but in reverse order. It cannot be digested by .Xr ed 1 . .It Fl -help This option prints a summary to stdout and exits with status 0. .It Fl n Produces a script similar to that of .Fl e , but in the opposite order and with a count of changed lines on each insert or delete command. This is the form used by rcsdiff. .It Fl q -brief Just print a line when the files differ. Does not output a list of changes. .It Fl U Ar number Fl -unified Ar number Like .Fl u but produces a diff with .Ar number lines of context. .It Fl u Produces a .Em unified diff with 3 lines of context. A unified diff is similar to the context diff produced by the .Fl c option. However, unlike with .Fl c , all lines to be changed (added and/or removed) are present in a single section. .It Fl -version This option prints a version string to stdout and exits with status 0. .It Fl y Fl -side-by-side Output in two columns with a marker between them. The marker can be one of the following: .Pp .Bl -tag -width Ds -offset indent -compact .It space Corresponding lines are identical. .It '|' Corresponding lines are different. .It '<' Files differ and only the first file contains the line. .It '>' Files differ and only the second file contains the line. .El .El .Pp Comparison options: .Bl -tag -width Ds .It Fl a -text Treat all files as ASCII text. Normally .Nm will simply print .Dq Binary files ... differ if files contain binary characters. Use of this option forces .Nm to produce a diff. .It Fl B Fl -ignore-blank-lines Causes chunks that include only blank lines to be ignored. .It Fl b -ignore-space-change Causes trailing blanks (spaces and tabs) to be ignored, and other strings of blanks to compare equal. .It Fl -color= Ns Oo Ar when Oc Color the additions green, and removals red, or the value in the .Ev DIFFCOLORS environment variable. The possible values of .Ar when are .Dq Cm never , .Dq Cm always and .Dq Cm auto . .Cm auto will use color if the output is a tty and the .Ev COLORTERM environment variable is set to a non-empty string. .It Fl d -minimal Try very hard to produce a diff as small as possible. This may consume a lot of processing power and memory when processing large files with many changes. .It Fl F Ar pattern, Fl -show-function-line Ar pattern Like .Fl p, but display the last line that matches provided pattern. .It Fl I Ar pattern Fl -ignore-matching-lines Ar pattern Ignores changes, insertions, and deletions whose lines match the extended regular expression .Ar pattern . Multiple .Fl I patterns may be specified. All lines in the change must match some pattern for the change to be ignored. See .Xr re_format 7 for more information on regular expression patterns. .It Fl i -ignore-case Ignores the case of letters. E.g., .Dq A will compare equal to .Dq a . .It Fl l -paginate Pass the output through .Xr pr 1 to paginate it. .It Fl L Ar label Fl -label Ar label Print .Ar label instead of the first (and second, if this option is specified twice) file name and time in the context or unified diff header. .It Fl p -show-c-function With unified and context diffs, show with each change the first 40 characters of the last line before the context beginning with a letter, an underscore or a dollar sign. For C and Objective-C source code following standard layout conventions, this will show the prototype of the function the change applies to. .It Fl T -initial-tab Print a tab rather than a space before the rest of the line for the normal, context or unified output formats. This makes the alignment of tabs in the line consistent. .It Fl t -expand-tabs Will expand tabs in output lines. Normal or .Fl c output adds character(s) to the front of each line which may screw up the indentation of the original source lines and make the output listing difficult to interpret. This option will preserve the original source's indentation. .It Fl w -ignore-all-blanks Is similar to .Fl b -ignore-space-change but causes whitespace (blanks and tabs) to be totally ignored. E.g., .Dq if (\ \&a == b \&) will compare equal to .Dq if(a==b) . .It Fl W Ar number Fl -width Ar number Output at most .Ar number columns when using side by side format. The default value is 130. +Note that unless +.It Fl t +was specified, +.Nm +will always align the second column to a tab stop, so values of +.Fl -width +smaller than approximately five times the value of +.Fl -tabsize +may yield surprising results. .It Fl -changed-group-format Ar GFMT Format input groups in the provided .Pp the format is a string with special keywords: .Bl -tag -width %< .It %< lines from FILE1 .It %< lines from FILE2 .El .It Fl -ignore-file-name-case ignore case when comparing file names .It Fl -no-ignore-file-name-case do not ignore case wen comparing file names (default) .It Fl -normal default diff output .It Fl -speed-large-files stub option for compatibility with GNU diff .It Fl -strip-trailing-cr strip carriage return on input files .It Fl -suppress-common-lines Do not output common lines when using the side by side format .It Fl -tabsize Ar number Number of spaces representing a tab (default 8) .El .Pp Directory comparison options: .Bl -tag -width Ds .It Fl N -new-file If a file is found in only one directory, act as if it was found in the other directory too but was of zero size. .It Fl P -unidirectional-new-file If a file is found only in .Ar dir2 , act as if it was found in .Ar dir1 too but was of zero size. .It Fl r -recursive Causes application of .Nm recursively to common subdirectories encountered. .It Fl S Ar name Fl -starting-file Ar name Re-starts a directory .Nm in the middle, beginning with file .Ar name . .It Fl s -report-identical-files Causes .Nm to report files which are the same, which are otherwise not mentioned. .It Fl X Ar file Fl -exclude-from Ar file Exclude files and subdirectories from comparison whose basenames match lines in .Ar file . Multiple .Fl X options may be specified. .It Fl x Ar pattern Fl -exclude Ar pattern Exclude files and subdirectories from comparison whose basenames match .Ar pattern . Patterns are matched using shell-style globbing via .Xr fnmatch 3 . Multiple .Fl x options may be specified. .El .Pp If both arguments are directories, .Nm sorts the contents of the directories by name, and then runs the regular file .Nm algorithm, producing a change list, on text files which are different. Binary files which differ, common subdirectories, and files which appear in only one directory are described as such. In directory mode only regular files and directories are compared. If a non-regular file such as a device special file or FIFO is encountered, a diagnostic message is printed. .Pp If only one of .Ar file1 and .Ar file2 is a directory, .Nm is applied to the non-directory file and the file contained in the directory file with a filename that is the same as the last component of the non-directory file. .Pp If either .Ar file1 or .Ar file2 is .Sq - , the standard input is used in its place. .Ss Output Style The default (without .Fl e , .Fl c , or .Fl n -rcs .\" -C options) output contains lines of these forms, where .Va XX , YY , ZZ , QQ are line numbers respective of file order. .Pp .Bl -tag -width "XX,YYcZZ,QQ" -compact .It Li XX Ns Ic a Ns Li YY At (the end of) line .Va XX of .Ar file1 , append the contents of line .Va YY of .Ar file2 to make them equal. .It Li XX Ns Ic a Ns Li YY,ZZ Same as above, but append the range of lines, .Va YY through .Va ZZ of .Ar file2 to line .Va XX of file1. .It Li XX Ns Ic d Ns Li YY At line .Va XX delete the line. The value .Va YY tells to which line the change would bring .Ar file1 in line with .Ar file2 . .It Li XX,YY Ns Ic d Ns Li ZZ Delete the range of lines .Va XX through .Va YY in .Ar file1 . .It Li XX Ns Ic c Ns Li YY Change the line .Va XX in .Ar file1 to the line .Va YY in .Ar file2 . .It Li XX,YY Ns Ic c Ns Li ZZ Replace the range of specified lines with the line .Va ZZ . .It Li XX,YY Ns Ic c Ns Li ZZ,QQ Replace the range .Va XX , Ns Va YY from .Ar file1 with the range .Va ZZ , Ns Va QQ from .Ar file2 . .El .Pp These lines resemble .Xr ed 1 subcommands to convert .Ar file1 into .Ar file2 . The line numbers before the action letters pertain to .Ar file1 ; those after pertain to .Ar file2 . Thus, by exchanging .Ic a for .Ic d and reading the line in reverse order, one can also determine how to convert .Ar file2 into .Ar file1 . As in .Xr ed 1 , identical pairs (where num1 = num2) are abbreviated as a single number. .Sh ENVIRONMENT .Bl -tag -width DIFFCOLORS .It Ev DIFFCOLORS The value of this variable is the form .Ar add Ns : Ns Ar rm , where .Ar add is the ASCII escape sequence for additions and .Ar rm is the ASCII escape sequence for deletions. If this is unset, .Nm uses green for additions and red for removals. .El .Sh FILES .Bl -tag -width /tmp/diff.XXXXXXXX -compact .It Pa /tmp/diff.XXXXXXXX Temporary file used when comparing a device or the standard input. Note that the temporary file is unlinked as soon as it is created so it will not show up in a directory listing. .El .Sh EXIT STATUS The .Nm utility exits with one of the following values: .Pp .Bl -tag -width Ds -offset indent -compact .It 0 No differences were found. .It 1 Differences were found. .It >1 An error occurred. .El .Pp The .Fl -help and .Fl -version options exit with a status of 0. .Sh EXAMPLES Compare .Pa old_dir and .Pa new_dir recursively generating an unified diff and treating files found only in one of those directories as new files: .Bd -literal -offset indent $ diff -ruN /path/to/old_dir /path/to/new_dir .Ed .Pp Same as above but excluding files matching the expressions .Dq *.h and .Dq *.c : .Bd -literal -offset indent $ diff -ruN -x '*.h' -x '*.c' /path/to/old_dir /path/to/new_dir .Ed .Pp Show a single line indicating if the files differ: .Bd -literal -offset indent $ diff -q /boot/loader.conf /boot/defaults/loader.conf Files /boot/loader.conf and /boot/defaults/loader.conf differ .Ed .Pp Assuming a file named .Pa example.txt with the following contents: .Bd -literal -offset indent FreeBSD is an operating system Linux is a kernel OpenBSD is an operating system .Ed .Pp Compare stdin with .Pa example.txt excluding from the comparison those lines containing either .Qq Linux or .Qq Open : .Bd -literal -offset indent $ echo "FreeBSD is an operating system" | diff -q -I 'Linux|Open' example.txt - .Ed .Sh SEE ALSO .Xr cmp 1 , .Xr comm 1 , .Xr diff3 1 , .Xr ed 1 , .Xr patch 1 , .Xr pr 1 , .Xr sdiff 1 .Rs .%A James W. Hunt .%A M. Douglas McIlroy .%T "An Algorithm for Differential File Comparison" .%J Computing Science Technical Report .%Q Bell Laboratories 41 .%D June 1976 .Re .Sh STANDARDS The .Nm utility is compliant with the .St -p1003.1-2008 specification. .Pp The flags .Op Fl aDdIiLlNnPpqSsTtwXxy are extensions to that specification. .Sh HISTORY A .Nm command appeared in .At v6 . diff --git a/usr.bin/diff/diff.c b/usr.bin/diff/diff.c index 03eb363fc323..d947c1e01705 100644 --- a/usr.bin/diff/diff.c +++ b/usr.bin/diff/diff.c @@ -1,657 +1,653 @@ /* $OpenBSD: diff.c,v 1.67 2019/06/28 13:35:00 deraadt Exp $ */ /* * Copyright (c) 2003 Todd C. Miller * * Permission to use, copy, modify, and distribute this software for any * purpose with or without fee is hereby granted, provided that the above * copyright notice and this permission notice appear in all copies. * * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR * ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE. * * Sponsored in part by the Defense Advanced Research Projects * Agency (DARPA) and Air Force Research Laboratory, Air Force * Materiel Command, USAF, under agreement number F39502-99-1-0512. */ #include #include #include #include #include #include #include #include #include #include #include #include "diff.h" #include "xmalloc.h" static const char diff_version[] = "FreeBSD diff 20220309"; bool lflag, Nflag, Pflag, rflag, sflag, Tflag, cflag; bool ignore_file_case, suppress_common, color, noderef; static bool help = false; int diff_format, diff_context, status; int tabsize = 8, width = 130; static int colorflag = COLORFLAG_NEVER; char *start, *ifdefname, *diffargs, *label[2]; char *ignore_pats, *most_recent_pat; char *group_format = NULL; const char *add_code, *del_code; struct stat stb1, stb2; struct excludes *excludes_list; regex_t ignore_re, most_recent_re; #define OPTIONS "0123456789aBbC:cdD:efF:HhI:iL:lnNPpqrS:sTtU:uwW:X:x:y" enum { OPT_TSIZE = CHAR_MAX + 1, OPT_STRIPCR, OPT_IGN_FN_CASE, OPT_NO_IGN_FN_CASE, OPT_NORMAL, OPT_HELP, OPT_HORIZON_LINES, OPT_CHANGED_GROUP_FORMAT, OPT_SUPPRESS_COMMON, OPT_COLOR, OPT_NO_DEREFERENCE, OPT_VERSION, }; static struct option longopts[] = { { "text", no_argument, 0, 'a' }, { "ignore-space-change", no_argument, 0, 'b' }, { "context", optional_argument, 0, 'C' }, { "ifdef", required_argument, 0, 'D' }, { "minimal", no_argument, 0, 'd' }, { "ed", no_argument, 0, 'e' }, { "forward-ed", no_argument, 0, 'f' }, { "show-function-line", required_argument, 0, 'F' }, { "speed-large-files", no_argument, NULL, 'H' }, { "ignore-blank-lines", no_argument, 0, 'B' }, { "ignore-matching-lines", required_argument, 0, 'I' }, { "ignore-case", no_argument, 0, 'i' }, { "paginate", no_argument, NULL, 'l' }, { "label", required_argument, 0, 'L' }, { "new-file", no_argument, 0, 'N' }, { "rcs", no_argument, 0, 'n' }, { "unidirectional-new-file", no_argument, 0, 'P' }, { "show-c-function", no_argument, 0, 'p' }, { "brief", no_argument, 0, 'q' }, { "recursive", no_argument, 0, 'r' }, { "report-identical-files", no_argument, 0, 's' }, { "starting-file", required_argument, 0, 'S' }, { "expand-tabs", no_argument, 0, 't' }, { "initial-tab", no_argument, 0, 'T' }, { "unified", optional_argument, 0, 'U' }, { "ignore-all-space", no_argument, 0, 'w' }, { "width", required_argument, 0, 'W' }, { "exclude", required_argument, 0, 'x' }, { "exclude-from", required_argument, 0, 'X' }, { "side-by-side", no_argument, NULL, 'y' }, { "ignore-file-name-case", no_argument, NULL, OPT_IGN_FN_CASE }, { "help", no_argument, NULL, OPT_HELP}, { "horizon-lines", required_argument, NULL, OPT_HORIZON_LINES }, { "no-dereference", no_argument, NULL, OPT_NO_DEREFERENCE}, { "no-ignore-file-name-case", no_argument, NULL, OPT_NO_IGN_FN_CASE }, { "normal", no_argument, NULL, OPT_NORMAL }, { "strip-trailing-cr", no_argument, NULL, OPT_STRIPCR }, { "tabsize", required_argument, NULL, OPT_TSIZE }, { "changed-group-format", required_argument, NULL, OPT_CHANGED_GROUP_FORMAT}, { "suppress-common-lines", no_argument, NULL, OPT_SUPPRESS_COMMON }, { "color", optional_argument, NULL, OPT_COLOR }, { "version", no_argument, NULL, OPT_VERSION}, { NULL, 0, 0, '\0'} }; static void checked_regcomp(char const *, regex_t *); static void usage(void) __dead2; static void conflicting_format(void) __dead2; static void push_excludes(char *); static void push_ignore_pats(char *); static void read_excludes_file(char *file); static void set_argstr(char **, char **); static char *splice(char *, char *); static bool do_color(void); int main(int argc, char **argv) { const char *errstr = NULL; char *ep, **oargv; long l; int ch, dflags, lastch, gotstdin, prevoptind, newarg; oargv = argv; gotstdin = 0; dflags = 0; lastch = '\0'; prevoptind = 1; newarg = 1; diff_context = 3; diff_format = D_UNSET; #define FORMAT_MISMATCHED(type) \ (diff_format != D_UNSET && diff_format != (type)) while ((ch = getopt_long(argc, argv, OPTIONS, longopts, NULL)) != -1) { switch (ch) { case '0': case '1': case '2': case '3': case '4': case '5': case '6': case '7': case '8': case '9': if (newarg) usage(); /* disallow -[0-9]+ */ else if (lastch == 'c' || lastch == 'u') diff_context = 0; else if (!isdigit(lastch) || diff_context > INT_MAX / 10) usage(); diff_context = (diff_context * 10) + (ch - '0'); break; case 'a': dflags |= D_FORCEASCII; break; case 'b': dflags |= D_FOLDBLANKS; break; case 'C': case 'c': if (FORMAT_MISMATCHED(D_CONTEXT)) conflicting_format(); cflag = true; diff_format = D_CONTEXT; if (optarg != NULL) { l = strtol(optarg, &ep, 10); if (*ep != '\0' || l < 0 || l >= INT_MAX) usage(); diff_context = (int)l; } break; case 'd': dflags |= D_MINIMAL; break; case 'D': if (FORMAT_MISMATCHED(D_IFDEF)) conflicting_format(); diff_format = D_IFDEF; ifdefname = optarg; break; case 'e': if (FORMAT_MISMATCHED(D_EDIT)) conflicting_format(); diff_format = D_EDIT; break; case 'f': if (FORMAT_MISMATCHED(D_REVERSE)) conflicting_format(); diff_format = D_REVERSE; break; case 'H': /* ignore but needed for compatibility with GNU diff */ break; case 'h': /* silently ignore for backwards compatibility */ break; case 'B': dflags |= D_SKIPBLANKLINES; break; case 'F': if (dflags & D_PROTOTYPE) conflicting_format(); dflags |= D_MATCHLAST; most_recent_pat = xstrdup(optarg); break; case 'I': push_ignore_pats(optarg); break; case 'i': dflags |= D_IGNORECASE; break; case 'L': if (label[0] == NULL) label[0] = optarg; else if (label[1] == NULL) label[1] = optarg; else usage(); break; case 'l': lflag = true; break; case 'N': Nflag = true; break; case 'n': if (FORMAT_MISMATCHED(D_NREVERSE)) conflicting_format(); diff_format = D_NREVERSE; break; case 'p': if (dflags & D_MATCHLAST) conflicting_format(); dflags |= D_PROTOTYPE; break; case 'P': Pflag = true; break; case 'r': rflag = true; break; case 'q': if (FORMAT_MISMATCHED(D_BRIEF)) conflicting_format(); diff_format = D_BRIEF; break; case 'S': start = optarg; break; case 's': sflag = true; break; case 'T': Tflag = true; break; case 't': dflags |= D_EXPANDTABS; break; case 'U': case 'u': if (FORMAT_MISMATCHED(D_UNIFIED)) conflicting_format(); diff_format = D_UNIFIED; if (optarg != NULL) { l = strtol(optarg, &ep, 10); if (*ep != '\0' || l < 0 || l >= INT_MAX) usage(); diff_context = (int)l; } break; case 'w': dflags |= D_IGNOREBLANKS; break; case 'W': width = (int) strtonum(optarg, 1, INT_MAX, &errstr); - if (errstr) { - warnx("Invalid argument for width"); - usage(); - } + if (errstr) + errx(1, "width is %s: %s", errstr, optarg); break; case 'X': read_excludes_file(optarg); break; case 'x': push_excludes(optarg); break; case 'y': if (FORMAT_MISMATCHED(D_SIDEBYSIDE)) conflicting_format(); diff_format = D_SIDEBYSIDE; break; case OPT_CHANGED_GROUP_FORMAT: if (FORMAT_MISMATCHED(D_GFORMAT)) conflicting_format(); diff_format = D_GFORMAT; group_format = optarg; break; case OPT_HELP: help = true; usage(); break; case OPT_HORIZON_LINES: break; /* XXX TODO for compatibility with GNU diff3 */ case OPT_IGN_FN_CASE: ignore_file_case = true; break; case OPT_NO_IGN_FN_CASE: ignore_file_case = false; break; case OPT_NORMAL: if (FORMAT_MISMATCHED(D_NORMAL)) conflicting_format(); diff_format = D_NORMAL; break; case OPT_TSIZE: tabsize = (int) strtonum(optarg, 1, INT_MAX, &errstr); - if (errstr) { - warnx("Invalid argument for tabsize"); - usage(); - } + if (errstr) + errx(1, "tabsize is %s: %s", errstr, optarg); break; case OPT_STRIPCR: dflags |= D_STRIPCR; break; case OPT_SUPPRESS_COMMON: suppress_common = 1; break; case OPT_COLOR: if (optarg == NULL || strncmp(optarg, "auto", 4) == 0) colorflag = COLORFLAG_AUTO; else if (strncmp(optarg, "always", 6) == 0) colorflag = COLORFLAG_ALWAYS; else if (strncmp(optarg, "never", 5) == 0) colorflag = COLORFLAG_NEVER; else errx(2, "unsupported --color value '%s' (must be always, auto, or never)", optarg); break; case OPT_NO_DEREFERENCE: rflag = true; noderef = true; break; case OPT_VERSION: printf("%s\n", diff_version); exit(0); default: usage(); break; } lastch = ch; newarg = optind != prevoptind; prevoptind = optind; } if (diff_format == D_UNSET && (dflags & D_PROTOTYPE) != 0) diff_format = D_CONTEXT; if (diff_format == D_UNSET) diff_format = D_NORMAL; argc -= optind; argv += optind; if (do_color()) { char *p; const char *env; color = true; add_code = "32"; del_code = "31"; env = getenv("DIFFCOLORS"); if (env != NULL && *env != '\0' && (p = strdup(env))) { add_code = p; strsep(&p, ":"); if (p != NULL) del_code = p; } } #ifdef __OpenBSD__ if (pledge("stdio rpath tmppath", NULL) == -1) err(2, "pledge"); #endif /* * Do sanity checks, fill in stb1 and stb2 and call the appropriate * driver routine. Both drivers use the contents of stb1 and stb2. */ if (argc != 2) usage(); checked_regcomp(ignore_pats, &ignore_re); checked_regcomp(most_recent_pat, &most_recent_re); if (strcmp(argv[0], "-") == 0) { fstat(STDIN_FILENO, &stb1); gotstdin = 1; } else if (stat(argv[0], &stb1) != 0) { if (!Nflag || errno != ENOENT) err(2, "%s", argv[0]); dflags |= D_EMPTY1; memset(&stb1, 0, sizeof(struct stat)); } if (strcmp(argv[1], "-") == 0) { fstat(STDIN_FILENO, &stb2); gotstdin = 1; } else if (stat(argv[1], &stb2) != 0) { if (!Nflag || errno != ENOENT) err(2, "%s", argv[1]); dflags |= D_EMPTY2; memset(&stb2, 0, sizeof(stb2)); stb2.st_mode = stb1.st_mode; } if (dflags & D_EMPTY1 && dflags & D_EMPTY2){ warn("%s", argv[0]); warn("%s", argv[1]); exit(2); } if (stb1.st_mode == 0) stb1.st_mode = stb2.st_mode; if (gotstdin && (S_ISDIR(stb1.st_mode) || S_ISDIR(stb2.st_mode))) errx(2, "can't compare - to a directory"); set_argstr(oargv, argv); if (S_ISDIR(stb1.st_mode) && S_ISDIR(stb2.st_mode)) { if (diff_format == D_IFDEF) errx(2, "-D option not supported with directories"); diffdir(argv[0], argv[1], dflags); } else { if (S_ISDIR(stb1.st_mode)) { argv[0] = splice(argv[0], argv[1]); if (stat(argv[0], &stb1) == -1) err(2, "%s", argv[0]); } if (S_ISDIR(stb2.st_mode)) { argv[1] = splice(argv[1], argv[0]); if (stat(argv[1], &stb2) == -1) err(2, "%s", argv[1]); } print_status(diffreg(argv[0], argv[1], dflags, 1), argv[0], argv[1], ""); } exit(status); } static void checked_regcomp(char const *pattern, regex_t *comp) { char buf[BUFSIZ]; int error; if (pattern == NULL) return; error = regcomp(comp, pattern, REG_NEWLINE | REG_EXTENDED); if (error != 0) { regerror(error, comp, buf, sizeof(buf)); if (*pattern != '\0') errx(2, "%s: %s", pattern, buf); else errx(2, "%s", buf); } } static void set_argstr(char **av, char **ave) { size_t argsize; char **ap; argsize = 4 + *ave - *av + 1; diffargs = xmalloc(argsize); strlcpy(diffargs, "diff", argsize); for (ap = av + 1; ap < ave; ap++) { if (strcmp(*ap, "--") != 0) { strlcat(diffargs, " ", argsize); strlcat(diffargs, *ap, argsize); } } } /* * Read in an excludes file and push each line. */ static void read_excludes_file(char *file) { FILE *fp; char *buf, *pattern; size_t len; if (strcmp(file, "-") == 0) fp = stdin; else if ((fp = fopen(file, "r")) == NULL) err(2, "%s", file); while ((buf = fgetln(fp, &len)) != NULL) { if (buf[len - 1] == '\n') len--; if ((pattern = strndup(buf, len)) == NULL) err(2, "xstrndup"); push_excludes(pattern); } if (strcmp(file, "-") != 0) fclose(fp); } /* * Push a pattern onto the excludes list. */ static void push_excludes(char *pattern) { struct excludes *entry; entry = xmalloc(sizeof(*entry)); entry->pattern = pattern; entry->next = excludes_list; excludes_list = entry; } static void push_ignore_pats(char *pattern) { size_t len; if (ignore_pats == NULL) ignore_pats = xstrdup(pattern); else { /* old + "|" + new + NUL */ len = strlen(ignore_pats) + strlen(pattern) + 2; ignore_pats = xreallocarray(ignore_pats, 1, len); strlcat(ignore_pats, "|", len); strlcat(ignore_pats, pattern, len); } } void print_status(int val, char *path1, char *path2, const char *entry) { if (label[0] != NULL) path1 = label[0]; if (label[1] != NULL) path2 = label[1]; switch (val) { case D_BINARY: printf("Binary files %s%s and %s%s differ\n", path1, entry, path2, entry); break; case D_DIFFER: if (diff_format == D_BRIEF) printf("Files %s%s and %s%s differ\n", path1, entry, path2, entry); break; case D_SAME: if (sflag) printf("Files %s%s and %s%s are identical\n", path1, entry, path2, entry); break; case D_MISMATCH1: printf("File %s%s is a directory while file %s%s is a regular file\n", path1, entry, path2, entry); break; case D_MISMATCH2: printf("File %s%s is a regular file while file %s%s is a directory\n", path1, entry, path2, entry); break; case D_SKIPPED1: printf("File %s%s is not a regular file or directory and was skipped\n", path1, entry); break; case D_SKIPPED2: printf("File %s%s is not a regular file or directory and was skipped\n", path2, entry); break; case D_ERROR: break; } } static void usage(void) { (void)fprintf(help ? stdout : stderr, "usage: diff [-aBbdilpTtw] [-c | -e | -f | -n | -q | -u] [--ignore-case]\n" " [--no-ignore-case] [--normal] [--strip-trailing-cr] [--tabsize]\n" " [-I pattern] [-F pattern] [-L label] file1 file2\n" " diff [-aBbdilpTtw] [-I pattern] [-L label] [--ignore-case]\n" " [--no-ignore-case] [--normal] [--strip-trailing-cr] [--tabsize]\n" " [-F pattern] -C number file1 file2\n" " diff [-aBbdiltw] [-I pattern] [--ignore-case] [--no-ignore-case]\n" " [--normal] [--strip-trailing-cr] [--tabsize] -D string file1 file2\n" " diff [-aBbdilpTtw] [-I pattern] [-L label] [--ignore-case]\n" " [--no-ignore-case] [--normal] [--tabsize] [--strip-trailing-cr]\n" " [-F pattern] -U number file1 file2\n" " diff [-aBbdilNPprsTtw] [-c | -e | -f | -n | -q | -u] [--ignore-case]\n" " [--no-ignore-case] [--normal] [--tabsize] [-I pattern] [-L label]\n" " [-F pattern] [-S name] [-X file] [-x pattern] dir1 dir2\n" " diff [-aBbditwW] [--expand-tabs] [--ignore-all-blanks]\n" " [--ignore-blank-lines] [--ignore-case] [--minimal]\n" " [--no-ignore-file-name-case] [--strip-trailing-cr]\n" " [--suppress-common-lines] [--tabsize] [--text] [--width]\n" " -y | --side-by-side file1 file2\n" " diff [--help] [--version]\n"); if (help) exit(0); else exit(2); } static void conflicting_format(void) { fprintf(stderr, "error: conflicting output format options.\n"); usage(); } static bool do_color(void) { const char *p, *p2; switch (colorflag) { case COLORFLAG_AUTO: p = getenv("CLICOLOR"); p2 = getenv("COLORTERM"); if ((p != NULL && *p != '\0') || (p2 != NULL && *p2 != '\0')) return isatty(STDOUT_FILENO); break; case COLORFLAG_ALWAYS: return (true); case COLORFLAG_NEVER: return (false); } return (false); } static char * splice(char *dir, char *path) { char *tail, *buf; size_t dirlen; dirlen = strlen(dir); while (dirlen != 0 && dir[dirlen - 1] == '/') dirlen--; if ((tail = strrchr(path, '/')) == NULL) tail = path; else tail++; xasprintf(&buf, "%.*s/%s", (int)dirlen, dir, tail); return (buf); } diff --git a/usr.bin/diff/diffreg.c b/usr.bin/diff/diffreg.c index 11aefb73cda6..8042003cd836 100644 --- a/usr.bin/diff/diffreg.c +++ b/usr.bin/diff/diffreg.c @@ -1,1685 +1,1702 @@ /* $OpenBSD: diffreg.c,v 1.93 2019/06/28 13:35:00 deraadt Exp $ */ /*- * SPDX-License-Identifier: BSD-4-Clause * * Copyright (C) Caldera International Inc. 2001-2002. * All rights reserved. * * Redistribution and use in source and binary forms, with or without * modification, are permitted provided that the following conditions * are met: * 1. Redistributions of source code and documentation must retain the above * copyright notice, this list of conditions and the following disclaimer. * 2. Redistributions in binary form must reproduce the above copyright * notice, this list of conditions and the following disclaimer in the * documentation and/or other materials provided with the distribution. * 3. All advertising materials mentioning features or use of this software * must display the following acknowledgement: * This product includes software developed or owned by Caldera * International, Inc. * 4. Neither the name of Caldera International, Inc. nor the names of other * contributors may be used to endorse or promote products derived from * this software without specific prior written permission. * * USE OF THE SOFTWARE PROVIDED FOR UNDER THIS LICENSE BY CALDERA * INTERNATIONAL, INC. AND CONTRIBUTORS ``AS IS'' AND ANY EXPRESS OR * IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES * OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. * IN NO EVENT SHALL CALDERA INTERNATIONAL, INC. BE LIABLE FOR ANY DIRECT, * INDIRECT INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES * (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR * SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, * STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING * IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE * POSSIBILITY OF SUCH DAMAGE. */ /*- * Copyright (c) 1991, 1993 * The Regents of the University of California. All rights reserved. * * Redistribution and use in source and binary forms, with or without * modification, are permitted provided that the following conditions * are met: * 1. Redistributions of source code must retain the above copyright * notice, this list of conditions and the following disclaimer. * 2. Redistributions in binary form must reproduce the above copyright * notice, this list of conditions and the following disclaimer in the * documentation and/or other materials provided with the distribution. * 3. Neither the name of the University nor the names of its contributors * may be used to endorse or promote products derived from this software * without specific prior written permission. * * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF * SUCH DAMAGE. * * @(#)diffreg.c 8.1 (Berkeley) 6/6/93 */ #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include "pr.h" #include "diff.h" #include "xmalloc.h" /* * diff - compare two files. */ /* * Uses an algorithm due to Harold Stone, which finds a pair of longest * identical subsequences in the two files. * * The major goal is to generate the match vector J. J[i] is the index of * the line in file1 corresponding to line i file0. J[i] = 0 if there is no * such line in file1. * * Lines are hashed so as to work in core. All potential matches are * located by sorting the lines of each file on the hash (called * ``value''). In particular, this collects the equivalence classes in * file1 together. Subroutine equiv replaces the value of each line in * file0 by the index of the first element of its matching equivalence in * (the reordered) file1. To save space equiv squeezes file1 into a single * array member in which the equivalence classes are simply concatenated, * except that their first members are flagged by changing sign. * * Next the indices that point into member are unsorted into array class * according to the original order of file0. * * The cleverness lies in routine stone. This marches through the lines of * file0, developing a vector klist of "k-candidates". At step i * a k-candidate is a matched pair of lines x,y (x in file0 y in file1) * such that there is a common subsequence of length k between the first * i lines of file0 and the first y lines of file1, but there is no such * subsequence for any smaller y. x is the earliest possible mate to y that * occurs in such a subsequence. * * Whenever any of the members of the equivalence class of lines in file1 * matable to a line in file0 has serial number less than the y of some * k-candidate, that k-candidate with the smallest such y is replaced. The * new k-candidate is chained (via pred) to the current k-1 candidate so * that the actual subsequence can be recovered. When a member has serial * number greater that the y of all k-candidates, the klist is extended. At * the end, the longest subsequence is pulled out and placed in the array J * by unravel. * * With J in hand, the matches there recorded are check'ed against reality * to assure that no spurious matches have crept in due to hashing. If they * have, they are broken, and "jackpot" is recorded -- a harmless matter * except that a true match for a spuriously mated line may now be * unnecessarily reported as a change. * * Much of the complexity of the program comes simply from trying to * minimize core utilization and maximize the range of doable problems by * dynamically allocating what is needed and reusing what is not. The core * requirements for problems larger than somewhat are (in words) * 2*length(file0) + length(file1) + 3*(number of k-candidates installed), * typically about 6n words for files of length n. */ struct cand { int x; int y; int pred; }; static struct line { int serial; int value; } *file[2]; /* * The following struct is used to record change information when * doing a "context" or "unified" diff. (see routine "change" to * understand the highly mnemonic field names) */ struct context_vec { int a; /* start line in old file */ int b; /* end line in old file */ int c; /* start line in new file */ int d; /* end line in new file */ }; enum readhash { RH_BINARY, RH_OK, RH_EOF }; -#define MIN_PAD 1 static FILE *opentemp(const char *); static void output(char *, FILE *, char *, FILE *, int); static void check(FILE *, FILE *, int); static void range(int, int, const char *); static void uni_range(int, int); static void dump_context_vec(FILE *, FILE *, int); static void dump_unified_vec(FILE *, FILE *, int); static bool prepare(int, FILE *, size_t, int); static void prune(void); static void equiv(struct line *, int, struct line *, int, int *); static void unravel(int); static void unsort(struct line *, int, int *); static void change(char *, FILE *, char *, FILE *, int, int, int, int, int *); static void sort(struct line *, int); static void print_header(const char *, const char *); static void print_space(int, int, int); static bool ignoreline_pattern(char *); static bool ignoreline(char *, bool); static int asciifile(FILE *); static int fetch(long *, int, int, FILE *, int, int, int); static int newcand(int, int, int); static int search(int *, int, int); static int skipline(FILE *); static int stone(int *, int, int *, int *, int); static enum readhash readhash(FILE *, int, unsigned *); static int files_differ(FILE *, FILE *, int); static char *match_function(const long *, int, FILE *); static char *preadline(int, size_t, off_t); static int *J; /* will be overlaid on class */ static int *class; /* will be overlaid on file[0] */ static int *klist; /* will be overlaid on file[0] after class */ static int *member; /* will be overlaid on file[1] */ static int clen; static int inifdef; /* whether or not we are in a #ifdef block */ static int len[2]; static int pref, suff; /* length of prefix and suffix */ static int slen[2]; static int anychange; -static int hw, padding; /* half width and padding */ +static int hw, lpad, rpad; /* half width and padding */ static int edoffset; static long *ixnew; /* will be overlaid on file[1] */ static long *ixold; /* will be overlaid on klist */ static struct cand *clist; /* merely a free storage pot for candidates */ static int clistlen; /* the length of clist */ static struct line *sfile[2]; /* shortened by pruning common prefix/suffix */ static int (*chrtran)(int); /* translation table for case-folding */ static struct context_vec *context_vec_start; static struct context_vec *context_vec_end; static struct context_vec *context_vec_ptr; #define FUNCTION_CONTEXT_SIZE 55 static char lastbuf[FUNCTION_CONTEXT_SIZE]; static int lastline; static int lastmatchline; static int clow2low(int c) { return (c); } static int cup2low(int c) { return (tolower(c)); } int diffreg(char *file1, char *file2, int flags, int capsicum) { FILE *f1, *f2; int i, rval; struct pr *pr = NULL; cap_rights_t rights_ro; f1 = f2 = NULL; rval = D_SAME; anychange = 0; lastline = 0; lastmatchline = 0; - /* - * hw excludes padding and make sure when -t is not used, - * the second column always starts from the closest tab stop - */ + /* + * In side-by-side mode, we need to print the left column, a + * change marker surrounded by padding, and the right column. + * + * If expanding tabs, we don't care about alignment, so we simply + * subtract 3 from the width and divide by two. + * + * If not expanding tabs, we need to ensure that the right column + * is aligned to a tab stop. We start with the same formula, then + * decrement until we reach a size that lets us tab-align the + * right column. We then adjust the width down if necessary for + * the padding calculation to work. + * + * Left padding is half the space left over, rounded down; right + * padding is whatever is needed to match the width. + */ if (diff_format == D_SIDEBYSIDE) { - hw = width >> 1; - padding = tabsize - (hw % tabsize); - if ((flags & D_EXPANDTABS) != 0 || (padding % tabsize == 0)) - padding = MIN_PAD; - - hw = (width >> 1) - - ((padding == MIN_PAD) ? (padding << 1) : padding) - 1; + if (flags & D_EXPANDTABS) { + if (width > 3) { + hw = (width - 3) / 2; + } else { + /* not enough space */ + hw = 0; + } + } else if (width <= 3 || width <= tabsize) { + /* not enough space */ + hw = 0; + } else { + hw = (width - 3) / 2; + while (hw > 0 && roundup(hw + 3, tabsize) + hw > width) + hw--; + if (width - (roundup(hw + 3, tabsize) + hw) < tabsize) + width = roundup(hw + 3, tabsize) + hw; + } + lpad = (width - hw * 2 - 1) / 2; + rpad = (width - hw * 2 - 1) - lpad; } - if (flags & D_IGNORECASE) chrtran = cup2low; else chrtran = clow2low; if (S_ISDIR(stb1.st_mode) != S_ISDIR(stb2.st_mode)) return (S_ISDIR(stb1.st_mode) ? D_MISMATCH1 : D_MISMATCH2); if (strcmp(file1, "-") == 0 && strcmp(file2, "-") == 0) goto closem; if (flags & D_EMPTY1) f1 = fopen(_PATH_DEVNULL, "r"); else { if (!S_ISREG(stb1.st_mode)) { if ((f1 = opentemp(file1)) == NULL || fstat(fileno(f1), &stb1) == -1) { warn("%s", file1); rval = D_ERROR; status |= 2; goto closem; } } else if (strcmp(file1, "-") == 0) f1 = stdin; else f1 = fopen(file1, "r"); } if (f1 == NULL) { warn("%s", file1); rval = D_ERROR; status |= 2; goto closem; } if (flags & D_EMPTY2) f2 = fopen(_PATH_DEVNULL, "r"); else { if (!S_ISREG(stb2.st_mode)) { if ((f2 = opentemp(file2)) == NULL || fstat(fileno(f2), &stb2) == -1) { warn("%s", file2); rval = D_ERROR; status |= 2; goto closem; } } else if (strcmp(file2, "-") == 0) f2 = stdin; else f2 = fopen(file2, "r"); } if (f2 == NULL) { warn("%s", file2); rval = D_ERROR; status |= 2; goto closem; } if (lflag) pr = start_pr(file1, file2); if (capsicum) { cap_rights_init(&rights_ro, CAP_READ, CAP_FSTAT, CAP_SEEK); if (caph_rights_limit(fileno(f1), &rights_ro) < 0) err(2, "unable to limit rights on: %s", file1); if (caph_rights_limit(fileno(f2), &rights_ro) < 0) err(2, "unable to limit rights on: %s", file2); if (fileno(f1) == STDIN_FILENO || fileno(f2) == STDIN_FILENO) { /* stdin has already been limited */ if (caph_limit_stderr() == -1) err(2, "unable to limit stderr"); if (caph_limit_stdout() == -1) err(2, "unable to limit stdout"); } else if (caph_limit_stdio() == -1) err(2, "unable to limit stdio"); caph_cache_catpages(); caph_cache_tzdata(); if (caph_enter() < 0) err(2, "unable to enter capability mode"); } switch (files_differ(f1, f2, flags)) { case 0: goto closem; case 1: break; default: /* error */ rval = D_ERROR; status |= 2; goto closem; } if (diff_format == D_BRIEF && ignore_pats == NULL && (flags & (D_FOLDBLANKS|D_IGNOREBLANKS|D_IGNORECASE|D_STRIPCR)) == 0) { rval = D_DIFFER; status |= 1; goto closem; } if ((flags & D_FORCEASCII) != 0) { (void)prepare(0, f1, stb1.st_size, flags); (void)prepare(1, f2, stb2.st_size, flags); } else if (!asciifile(f1) || !asciifile(f2) || !prepare(0, f1, stb1.st_size, flags) || !prepare(1, f2, stb2.st_size, flags)) { rval = D_BINARY; status |= 1; goto closem; } prune(); sort(sfile[0], slen[0]); sort(sfile[1], slen[1]); member = (int *)file[1]; equiv(sfile[0], slen[0], sfile[1], slen[1], member); member = xreallocarray(member, slen[1] + 2, sizeof(*member)); class = (int *)file[0]; unsort(sfile[0], slen[0], class); class = xreallocarray(class, slen[0] + 2, sizeof(*class)); klist = xcalloc(slen[0] + 2, sizeof(*klist)); clen = 0; clistlen = 100; clist = xcalloc(clistlen, sizeof(*clist)); i = stone(class, slen[0], member, klist, flags); free(member); free(class); J = xreallocarray(J, len[0] + 2, sizeof(*J)); unravel(klist[i]); free(clist); free(klist); ixold = xreallocarray(ixold, len[0] + 2, sizeof(*ixold)); ixnew = xreallocarray(ixnew, len[1] + 2, sizeof(*ixnew)); check(f1, f2, flags); output(file1, f1, file2, f2, flags); closem: if (pr != NULL) stop_pr(pr); if (anychange) { status |= 1; if (rval == D_SAME) rval = D_DIFFER; } if (f1 != NULL) fclose(f1); if (f2 != NULL) fclose(f2); return (rval); } /* * Check to see if the given files differ. * Returns 0 if they are the same, 1 if different, and -1 on error. * XXX - could use code from cmp(1) [faster] */ static int files_differ(FILE *f1, FILE *f2, int flags) { char buf1[BUFSIZ], buf2[BUFSIZ]; size_t i, j; if ((flags & (D_EMPTY1|D_EMPTY2)) || stb1.st_size != stb2.st_size || (stb1.st_mode & S_IFMT) != (stb2.st_mode & S_IFMT)) return (1); if (stb1.st_dev == stb2.st_dev && stb1.st_ino == stb2.st_ino) return (0); for (;;) { i = fread(buf1, 1, sizeof(buf1), f1); j = fread(buf2, 1, sizeof(buf2), f2); if ((!i && ferror(f1)) || (!j && ferror(f2))) return (-1); if (i != j) return (1); if (i == 0) return (0); if (memcmp(buf1, buf2, i) != 0) return (1); } } static FILE * opentemp(const char *f) { char buf[BUFSIZ], tempfile[PATH_MAX]; ssize_t nread; int ifd, ofd; if (strcmp(f, "-") == 0) ifd = STDIN_FILENO; else if ((ifd = open(f, O_RDONLY, 0644)) == -1) return (NULL); (void)strlcpy(tempfile, _PATH_TMP "/diff.XXXXXXXX", sizeof(tempfile)); if ((ofd = mkstemp(tempfile)) == -1) { close(ifd); return (NULL); } unlink(tempfile); while ((nread = read(ifd, buf, BUFSIZ)) > 0) { if (write(ofd, buf, nread) != nread) { close(ifd); close(ofd); return (NULL); } } close(ifd); lseek(ofd, (off_t)0, SEEK_SET); return (fdopen(ofd, "r")); } static bool prepare(int i, FILE *fd, size_t filesize, int flags) { struct line *p; unsigned h; size_t sz, j = 0; enum readhash r; rewind(fd); sz = MIN(filesize, SIZE_MAX) / 25; if (sz < 100) sz = 100; p = xcalloc(sz + 3, sizeof(*p)); while ((r = readhash(fd, flags, &h)) != RH_EOF) switch (r) { case RH_EOF: /* otherwise clang complains */ case RH_BINARY: return (false); case RH_OK: if (j == sz) { sz = sz * 3 / 2; p = xreallocarray(p, sz + 3, sizeof(*p)); } p[++j].value = h; } len[i] = j; file[i] = p; return (true); } static void prune(void) { int i, j; for (pref = 0; pref < len[0] && pref < len[1] && file[0][pref + 1].value == file[1][pref + 1].value; pref++) ; for (suff = 0; suff < len[0] - pref && suff < len[1] - pref && file[0][len[0] - suff].value == file[1][len[1] - suff].value; suff++) ; for (j = 0; j < 2; j++) { sfile[j] = file[j] + pref; slen[j] = len[j] - pref - suff; for (i = 0; i <= slen[j]; i++) sfile[j][i].serial = i; } } static void equiv(struct line *a, int n, struct line *b, int m, int *c) { int i, j; i = j = 1; while (i <= n && j <= m) { if (a[i].value < b[j].value) a[i++].value = 0; else if (a[i].value == b[j].value) a[i++].value = j; else j++; } while (i <= n) a[i++].value = 0; b[m + 1].value = 0; j = 0; while (++j <= m) { c[j] = -b[j].serial; while (b[j + 1].value == b[j].value) { j++; c[j] = b[j].serial; } } c[j] = -1; } static int stone(int *a, int n, int *b, int *c, int flags) { int i, k, y, j, l; int oldc, tc, oldl, sq; unsigned numtries, bound; if (flags & D_MINIMAL) bound = UINT_MAX; else { sq = sqrt(n); bound = MAX(256, sq); } k = 0; c[0] = newcand(0, 0, 0); for (i = 1; i <= n; i++) { j = a[i]; if (j == 0) continue; y = -b[j]; oldl = 0; oldc = c[0]; numtries = 0; do { if (y <= clist[oldc].y) continue; l = search(c, k, y); if (l != oldl + 1) oldc = c[l - 1]; if (l <= k) { if (clist[c[l]].y <= y) continue; tc = c[l]; c[l] = newcand(i, y, oldc); oldc = tc; oldl = l; numtries++; } else { c[l] = newcand(i, y, oldc); k++; break; } } while ((y = b[++j]) > 0 && numtries < bound); } return (k); } static int newcand(int x, int y, int pred) { struct cand *q; if (clen == clistlen) { clistlen = clistlen * 11 / 10; clist = xreallocarray(clist, clistlen, sizeof(*clist)); } q = clist + clen; q->x = x; q->y = y; q->pred = pred; return (clen++); } static int search(int *c, int k, int y) { int i, j, l, t; if (clist[c[k]].y < y) /* quick look for typical case */ return (k + 1); i = 0; j = k + 1; for (;;) { l = (i + j) / 2; if (l <= i) break; t = clist[c[l]].y; if (t > y) j = l; else if (t < y) i = l; else return (l); } return (l + 1); } static void unravel(int p) { struct cand *q; int i; for (i = 0; i <= len[0]; i++) J[i] = i <= pref ? i : i > len[0] - suff ? i + len[1] - len[0] : 0; for (q = clist + p; q->y != 0; q = clist + q->pred) J[q->x + pref] = q->y + pref; } /* * Check does double duty: * 1. ferret out any fortuitous correspondences due to confounding by * hashing (which result in "jackpot") * 2. collect random access indexes to the two files */ static void check(FILE *f1, FILE *f2, int flags) { int i, j, /* jackpot, */ c, d; long ctold, ctnew; rewind(f1); rewind(f2); j = 1; ixold[0] = ixnew[0] = 0; /* jackpot = 0; */ ctold = ctnew = 0; for (i = 1; i <= len[0]; i++) { if (J[i] == 0) { ixold[i] = ctold += skipline(f1); continue; } while (j < J[i]) { ixnew[j] = ctnew += skipline(f2); j++; } if (flags & (D_FOLDBLANKS | D_IGNOREBLANKS | D_IGNORECASE | D_STRIPCR)) { for (;;) { c = getc(f1); d = getc(f2); /* * GNU diff ignores a missing newline * in one file for -b or -w. */ if (flags & (D_FOLDBLANKS | D_IGNOREBLANKS)) { if (c == EOF && d == '\n') { ctnew++; break; } else if (c == '\n' && d == EOF) { ctold++; break; } } ctold++; ctnew++; if (flags & D_STRIPCR && (c == '\r' || d == '\r')) { if (c == '\r') { if ((c = getc(f1)) == '\n') { ctold++; } else { ungetc(c, f1); } } if (d == '\r') { if ((d = getc(f2)) == '\n') { ctnew++; } else { ungetc(d, f2); } } break; } if ((flags & D_FOLDBLANKS) && isspace(c) && isspace(d)) { do { if (c == '\n') break; ctold++; } while (isspace(c = getc(f1))); do { if (d == '\n') break; ctnew++; } while (isspace(d = getc(f2))); } else if (flags & D_IGNOREBLANKS) { while (isspace(c) && c != '\n') { c = getc(f1); ctold++; } while (isspace(d) && d != '\n') { d = getc(f2); ctnew++; } } if (chrtran(c) != chrtran(d)) { /* jackpot++; */ J[i] = 0; if (c != '\n' && c != EOF) ctold += skipline(f1); if (d != '\n' && c != EOF) ctnew += skipline(f2); break; } if (c == '\n' || c == EOF) break; } } else { for (;;) { ctold++; ctnew++; if ((c = getc(f1)) != (d = getc(f2))) { /* jackpot++; */ J[i] = 0; if (c != '\n' && c != EOF) ctold += skipline(f1); if (d != '\n' && c != EOF) ctnew += skipline(f2); break; } if (c == '\n' || c == EOF) break; } } ixold[i] = ctold; ixnew[j] = ctnew; j++; } for (; j <= len[1]; j++) { ixnew[j] = ctnew += skipline(f2); } /* * if (jackpot) * fprintf(stderr, "jackpot\n"); */ } /* shellsort CACM #201 */ static void sort(struct line *a, int n) { struct line *ai, *aim, w; int j, m = 0, k; if (n == 0) return; for (j = 1; j <= n; j *= 2) m = 2 * j - 1; for (m /= 2; m != 0; m /= 2) { k = n - m; for (j = 1; j <= k; j++) { for (ai = &a[j]; ai > a; ai -= m) { aim = &ai[m]; if (aim < ai) break; /* wraparound */ if (aim->value > ai[0].value || (aim->value == ai[0].value && aim->serial > ai[0].serial)) break; w.value = ai[0].value; ai[0].value = aim->value; aim->value = w.value; w.serial = ai[0].serial; ai[0].serial = aim->serial; aim->serial = w.serial; } } } } static void unsort(struct line *f, int l, int *b) { int *a, i; a = xcalloc(l + 1, sizeof(*a)); for (i = 1; i <= l; i++) a[f[i].serial] = f[i].value; for (i = 1; i <= l; i++) b[i] = a[i]; free(a); } static int skipline(FILE *f) { int i, c; for (i = 1; (c = getc(f)) != '\n' && c != EOF; i++) continue; return (i); } static void output(char *file1, FILE *f1, char *file2, FILE *f2, int flags) { int i, j, m, i0, i1, j0, j1, nc; rewind(f1); rewind(f2); m = len[0]; J[0] = 0; J[m + 1] = len[1] + 1; if (diff_format != D_EDIT) { for (i0 = 1; i0 <= m; i0 = i1 + 1) { while (i0 <= m && J[i0] == J[i0 - 1] + 1) { if (diff_format == D_SIDEBYSIDE && suppress_common != 1) { nc = fetch(ixold, i0, i0, f1, '\0', 1, flags); - print_space(nc, (hw - nc) + (padding << 1) + 1, flags); + print_space(nc, hw - nc + lpad + 1 + rpad, flags); fetch(ixnew, J[i0], J[i0], f2, '\0', 0, flags); printf("\n"); } i0++; } j0 = J[i0 - 1] + 1; i1 = i0 - 1; while (i1 < m && J[i1 + 1] == 0) i1++; j1 = J[i1 + 1] - 1; J[i1] = j1; /* * When using side-by-side, lines from both of the files are * printed. The algorithm used by diff(1) identifies the ranges * in which two files differ. * See the change() function below. * The for loop below consumes the shorter range, whereas one of * the while loops deals with the longer one. */ if (diff_format == D_SIDEBYSIDE) { for (i = i0, j = j0; i <= i1 && j <= j1; i++, j++) change(file1, f1, file2, f2, i, i, j, j, &flags); while (i <= i1) { change(file1, f1, file2, f2, i, i, j + 1, j, &flags); i++; } while (j <= j1) { change(file1, f1, file2, f2, i + 1, i, j, j, &flags); j++; } } else change(file1, f1, file2, f2, i0, i1, j0, j1, &flags); } } else { for (i0 = m; i0 >= 1; i0 = i1 - 1) { while (i0 >= 1 && J[i0] == J[i0 + 1] - 1 && J[i0] != 0) i0--; j0 = J[i0 + 1] - 1; i1 = i0 + 1; while (i1 > 1 && J[i1 - 1] == 0) i1--; j1 = J[i1 - 1] + 1; J[i1] = j1; change(file1, f1, file2, f2, i1, i0, j1, j0, &flags); } } if (m == 0) change(file1, f1, file2, f2, 1, 0, 1, len[1], &flags); if (diff_format == D_IFDEF || diff_format == D_GFORMAT) { for (;;) { #define c i0 if ((c = getc(f1)) == EOF) return; printf("%c", c); } #undef c } if (anychange != 0) { if (diff_format == D_CONTEXT) dump_context_vec(f1, f2, flags); else if (diff_format == D_UNIFIED) dump_unified_vec(f1, f2, flags); } } static void range(int a, int b, const char *separator) { printf("%d", a > b ? b : a); if (a < b) printf("%s%d", separator, b); } static void uni_range(int a, int b) { if (a < b) printf("%d,%d", a, b - a + 1); else if (a == b) printf("%d", b); else printf("%d,0", b); } static char * preadline(int fd, size_t rlen, off_t off) { char *line; ssize_t nr; line = xmalloc(rlen + 1); if ((nr = pread(fd, line, rlen, off)) == -1) err(2, "preadline"); if (nr > 0 && line[nr-1] == '\n') nr--; line[nr] = '\0'; return (line); } static bool ignoreline_pattern(char *line) { int ret; ret = regexec(&ignore_re, line, 0, NULL, 0); return (ret == 0); /* if it matched, it should be ignored. */ } static bool ignoreline(char *line, bool skip_blanks) { if (skip_blanks && *line == '\0') return (true); if (ignore_pats != NULL && ignoreline_pattern(line)) return (true); return (false); } /* * Indicate that there is a difference between lines a and b of the from file * to get to lines c to d of the to file. If a is greater then b then there * are no lines in the from file involved and this means that there were * lines appended (beginning at b). If c is greater than d then there are * lines missing from the to file. */ static void change(char *file1, FILE *f1, char *file2, FILE *f2, int a, int b, int c, int d, int *pflags) { static size_t max_context = 64; long curpos; int i, nc; const char *walk; bool skip_blanks, ignore; skip_blanks = (*pflags & D_SKIPBLANKLINES); restart: if ((diff_format != D_IFDEF || diff_format == D_GFORMAT) && a > b && c > d) return; if (ignore_pats != NULL || skip_blanks) { char *line; /* * All lines in the change, insert, or delete must match an ignore * pattern for the change to be ignored. */ if (a <= b) { /* Changes and deletes. */ for (i = a; i <= b; i++) { line = preadline(fileno(f1), ixold[i] - ixold[i - 1], ixold[i - 1]); ignore = ignoreline(line, skip_blanks); free(line); if (!ignore) goto proceed; } } if (a > b || c <= d) { /* Changes and inserts. */ for (i = c; i <= d; i++) { line = preadline(fileno(f2), ixnew[i] - ixnew[i - 1], ixnew[i - 1]); ignore = ignoreline(line, skip_blanks); free(line); if (!ignore) goto proceed; } } return; } proceed: if (*pflags & D_HEADER && diff_format != D_BRIEF) { printf("%s %s %s\n", diffargs, file1, file2); *pflags &= ~D_HEADER; } if (diff_format == D_CONTEXT || diff_format == D_UNIFIED) { /* * Allocate change records as needed. */ if (context_vec_start == NULL || context_vec_ptr == context_vec_end - 1) { ptrdiff_t offset = -1; if (context_vec_start != NULL) offset = context_vec_ptr - context_vec_start; max_context <<= 1; context_vec_start = xreallocarray(context_vec_start, max_context, sizeof(*context_vec_start)); context_vec_end = context_vec_start + max_context; context_vec_ptr = context_vec_start + offset; } if (anychange == 0) { /* * Print the context/unidiff header first time through. */ print_header(file1, file2); anychange = 1; } else if (a > context_vec_ptr->b + (2 * diff_context) + 1 && c > context_vec_ptr->d + (2 * diff_context) + 1) { /* * If this change is more than 'diff_context' lines from the * previous change, dump the record and reset it. */ if (diff_format == D_CONTEXT) dump_context_vec(f1, f2, *pflags); else dump_unified_vec(f1, f2, *pflags); } context_vec_ptr++; context_vec_ptr->a = a; context_vec_ptr->b = b; context_vec_ptr->c = c; context_vec_ptr->d = d; return; } if (anychange == 0) anychange = 1; switch (diff_format) { case D_BRIEF: return; case D_NORMAL: case D_EDIT: range(a, b, ","); printf("%c", a > b ? 'a' : c > d ? 'd' : 'c'); if (diff_format == D_NORMAL) range(c, d, ","); printf("\n"); break; case D_REVERSE: printf("%c", a > b ? 'a' : c > d ? 'd' : 'c'); range(a, b, " "); printf("\n"); break; case D_NREVERSE: if (a > b) printf("a%d %d\n", b, d - c + 1); else { printf("d%d %d\n", a, b - a + 1); if (!(c > d)) /* add changed lines */ printf("a%d %d\n", b, d - c + 1); } break; } if (diff_format == D_GFORMAT) { curpos = ftell(f1); /* print through if append (a>b), else to (nb: 0 vs 1 orig) */ nc = ixold[a > b ? b : a - 1] - curpos; for (i = 0; i < nc; i++) printf("%c", getc(f1)); for (walk = group_format; *walk != '\0'; walk++) { if (*walk == '%') { walk++; switch (*walk) { case '<': fetch(ixold, a, b, f1, '<', 1, *pflags); break; case '>': fetch(ixnew, c, d, f2, '>', 0, *pflags); break; default: printf("%%%c", *walk); break; } continue; } printf("%c", *walk); } } if (diff_format == D_SIDEBYSIDE) { if (color && a > b) printf("\033[%sm", add_code); else if (color && c > d) printf("\033[%sm", del_code); if (a > b) { - print_space(0, hw + padding , *pflags); + print_space(0, hw + lpad, *pflags); } else { nc = fetch(ixold, a, b, f1, '\0', 1, *pflags); - print_space(nc, hw - nc + padding, *pflags); + print_space(nc, hw - nc + lpad, *pflags); } if (color && a > b) printf("\033[%sm", add_code); else if (color && c > d) printf("\033[%sm", del_code); printf("%c", (a > b) ? '>' : ((c > d) ? '<' : '|')); if (color && c > d) printf("\033[m"); - print_space(hw + padding + 1 , padding, *pflags); + print_space(hw + lpad + 1, rpad, *pflags); fetch(ixnew, c, d, f2, '\0', 0, *pflags); printf("\n"); } if (diff_format == D_NORMAL || diff_format == D_IFDEF) { fetch(ixold, a, b, f1, '<', 1, *pflags); if (a <= b && c <= d && diff_format == D_NORMAL) printf("---\n"); } if (diff_format != D_GFORMAT && diff_format != D_SIDEBYSIDE) fetch(ixnew, c, d, f2, diff_format == D_NORMAL ? '>' : '\0', 0, *pflags); if (edoffset != 0 && diff_format == D_EDIT) { /* * A non-zero edoffset value for D_EDIT indicates that the last line * printed was a bare dot (".") that has been escaped as ".." to * prevent ed(1) from misinterpreting it. We have to add a * substitute command to change this back and restart where we left * off. */ printf(".\n"); printf("%ds/.//\n", a + edoffset - 1); b = a + edoffset - 1; a = b + 1; c += edoffset; goto restart; } if ((diff_format == D_EDIT || diff_format == D_REVERSE) && c <= d) printf(".\n"); if (inifdef) { printf("#endif /* %s */\n", ifdefname); inifdef = 0; } } static int fetch(long *f, int a, int b, FILE *lb, int ch, int oldfile, int flags) { int i, j, c, lastc, col, nc, newcol; edoffset = 0; nc = 0; /* * When doing #ifdef's, copy down to current line * if this is the first file, so that stuff makes it to output. */ if ((diff_format == D_IFDEF) && oldfile) { long curpos = ftell(lb); /* print through if append (a>b), else to (nb: 0 vs 1 orig) */ nc = f[a > b ? b : a - 1] - curpos; for (i = 0; i < nc; i++) printf("%c", getc(lb)); } if (a > b) return (0); if (diff_format == D_IFDEF) { if (inifdef) { printf("#else /* %s%s */\n", oldfile == 1 ? "!" : "", ifdefname); } else { if (oldfile) printf("#ifndef %s\n", ifdefname); else printf("#ifdef %s\n", ifdefname); } inifdef = 1 + oldfile; } for (i = a; i <= b; i++) { fseek(lb, f[i - 1], SEEK_SET); nc = f[i] - f[i - 1]; if (diff_format == D_SIDEBYSIDE && hw < nc) nc = hw; if (diff_format != D_IFDEF && diff_format != D_GFORMAT && ch != '\0') { if (color && (ch == '>' || ch == '+')) printf("\033[%sm", add_code); else if (color && (ch == '<' || ch == '-')) printf("\033[%sm", del_code); printf("%c", ch); if (Tflag && (diff_format == D_NORMAL || diff_format == D_CONTEXT || diff_format == D_UNIFIED)) printf("\t"); else if (diff_format != D_UNIFIED) printf(" "); } col = j = 0; lastc = '\0'; while (j < nc && (hw == 0 || col < hw)) { c = getc(lb); if (flags & D_STRIPCR && c == '\r') { if ((c = getc(lb)) == '\n') j++; else { ungetc(c, lb); c = '\r'; } } if (c == EOF) { if (diff_format == D_EDIT || diff_format == D_REVERSE || diff_format == D_NREVERSE) warnx("No newline at end of file"); else printf("\n\\ No newline at end of file\n"); return (col); } - /* - * when using --side-by-side, col needs to be increased - * in any case to keep the columns aligned - */ if (c == '\t') { - if (flags & D_EXPANDTABS) { - newcol = ((col / tabsize) + 1) * tabsize; - do { - printf(" "); - } while (++col < newcol && col < hw); + /* + * Calculate where the tab would bring us. + * If it would take us to the end of the + * column, either clip it (if expanding + * tabs) or return right away (if not). + */ + newcol = roundup(col + 1, tabsize); + if ((flags & D_EXPANDTABS) == 0) { + if (hw > 0 && newcol >= hw) + return (col); + printf("\t"); } else { - if (diff_format == D_SIDEBYSIDE) { - if ((col + tabsize) > hw) { - printf("%*s", hw - col, ""); - col = hw; - } else { - printf("\t"); - col += tabsize - 1; - } - } else { - printf("\t"); - col++; - } + if (hw > 0 && newcol > hw) + newcol = hw; + printf("%*s", newcol - col, ""); } + col = newcol; } else { if (diff_format == D_EDIT && j == 1 && c == '\n' && lastc == '.') { /* * Don't print a bare "." line since that will confuse * ed(1). Print ".." instead and set the, global variable * edoffset to an offset from which to restart. The * caller must check the value of edoffset */ printf(".\n"); edoffset = i - a + 1; return (edoffset); } /* when side-by-side, do not print a newline */ if (diff_format != D_SIDEBYSIDE || c != '\n') { if (color && c == '\n') printf("\033[m%c", c); else printf("%c", c); col++; } } j++; lastc = c; } } if (color && diff_format == D_SIDEBYSIDE) printf("\033[m"); return (col); } /* * Hash function taken from Robert Sedgewick, Algorithms in C, 3d ed., p 578. */ static enum readhash readhash(FILE *f, int flags, unsigned *hash) { int i, t, space; unsigned sum; sum = 1; space = 0; for (i = 0;;) { switch (t = getc(f)) { case '\0': if ((flags & D_FORCEASCII) == 0) return (RH_BINARY); goto hashchar; case '\r': if (flags & D_STRIPCR) { t = getc(f); if (t == '\n') break; ungetc(t, f); } /* FALLTHROUGH */ case '\t': case '\v': case '\f': case ' ': if ((flags & (D_FOLDBLANKS|D_IGNOREBLANKS)) != 0) { space++; continue; } /* FALLTHROUGH */ default: hashchar: if (space && (flags & D_IGNOREBLANKS) == 0) { i++; space = 0; } sum = sum * 127 + chrtran(t); i++; continue; case EOF: if (i == 0) return (RH_EOF); /* FALLTHROUGH */ case '\n': break; } break; } *hash = sum; return (RH_OK); } static int asciifile(FILE *f) { unsigned char buf[BUFSIZ]; size_t cnt; if (f == NULL) return (1); rewind(f); cnt = fread(buf, 1, sizeof(buf), f); return (memchr(buf, '\0', cnt) == NULL); } #define begins_with(s, pre) (strncmp(s, pre, sizeof(pre) - 1) == 0) static char * match_function(const long *f, int pos, FILE *fp) { unsigned char buf[FUNCTION_CONTEXT_SIZE]; size_t nc; int last = lastline; const char *state = NULL; lastline = pos; for (; pos > last; pos--) { fseek(fp, f[pos - 1], SEEK_SET); nc = f[pos] - f[pos - 1]; if (nc >= sizeof(buf)) nc = sizeof(buf) - 1; nc = fread(buf, 1, nc, fp); if (nc == 0) continue; buf[nc] = '\0'; buf[strcspn(buf, "\n")] = '\0'; if (most_recent_pat != NULL) { int ret = regexec(&most_recent_re, buf, 0, NULL, 0); if (ret != 0) continue; strlcpy(lastbuf, buf, sizeof(lastbuf)); lastmatchline = pos; return (lastbuf); } else if (isalpha(buf[0]) || buf[0] == '_' || buf[0] == '$' || buf[0] == '-' || buf[0] == '+') { if (begins_with(buf, "private:")) { if (!state) state = " (private)"; } else if (begins_with(buf, "protected:")) { if (!state) state = " (protected)"; } else if (begins_with(buf, "public:")) { if (!state) state = " (public)"; } else { strlcpy(lastbuf, buf, sizeof(lastbuf)); if (state) strlcat(lastbuf, state, sizeof(lastbuf)); lastmatchline = pos; return (lastbuf); } } } return (lastmatchline > 0 ? lastbuf : NULL); } /* dump accumulated "context" diff changes */ static void dump_context_vec(FILE *f1, FILE *f2, int flags) { struct context_vec *cvp = context_vec_start; int lowa, upb, lowc, upd, do_output; int a, b, c, d; char ch, *f; if (context_vec_start > context_vec_ptr) return; b = d = 0; /* gcc */ lowa = MAX(1, cvp->a - diff_context); upb = MIN(len[0], context_vec_ptr->b + diff_context); lowc = MAX(1, cvp->c - diff_context); upd = MIN(len[1], context_vec_ptr->d + diff_context); printf("***************"); if (flags & (D_PROTOTYPE | D_MATCHLAST)) { f = match_function(ixold, cvp->a - 1, f1); if (f != NULL) printf(" %s", f); } printf("\n*** "); range(lowa, upb, ","); printf(" ****\n"); /* * Output changes to the "old" file. The first loop suppresses * output if there were no changes to the "old" file (we'll see * the "old" lines as context in the "new" list). */ do_output = 0; for (; cvp <= context_vec_ptr; cvp++) if (cvp->a <= cvp->b) { cvp = context_vec_start; do_output++; break; } if (do_output) { while (cvp <= context_vec_ptr) { a = cvp->a; b = cvp->b; c = cvp->c; d = cvp->d; if (a <= b && c <= d) ch = 'c'; else ch = (a <= b) ? 'd' : 'a'; if (ch == 'a') fetch(ixold, lowa, b, f1, ' ', 0, flags); else { fetch(ixold, lowa, a - 1, f1, ' ', 0, flags); fetch(ixold, a, b, f1, ch == 'c' ? '!' : '-', 0, flags); } lowa = b + 1; cvp++; } fetch(ixold, b + 1, upb, f1, ' ', 0, flags); } /* output changes to the "new" file */ printf("--- "); range(lowc, upd, ","); printf(" ----\n"); do_output = 0; for (cvp = context_vec_start; cvp <= context_vec_ptr; cvp++) if (cvp->c <= cvp->d) { cvp = context_vec_start; do_output++; break; } if (do_output) { while (cvp <= context_vec_ptr) { a = cvp->a; b = cvp->b; c = cvp->c; d = cvp->d; if (a <= b && c <= d) ch = 'c'; else ch = (a <= b) ? 'd' : 'a'; if (ch == 'd') fetch(ixnew, lowc, d, f2, ' ', 0, flags); else { fetch(ixnew, lowc, c - 1, f2, ' ', 0, flags); fetch(ixnew, c, d, f2, ch == 'c' ? '!' : '+', 0, flags); } lowc = d + 1; cvp++; } fetch(ixnew, d + 1, upd, f2, ' ', 0, flags); } context_vec_ptr = context_vec_start - 1; } /* dump accumulated "unified" diff changes */ static void dump_unified_vec(FILE *f1, FILE *f2, int flags) { struct context_vec *cvp = context_vec_start; int lowa, upb, lowc, upd; int a, b, c, d; char ch, *f; if (context_vec_start > context_vec_ptr) return; b = d = 0; /* gcc */ lowa = MAX(1, cvp->a - diff_context); upb = MIN(len[0], context_vec_ptr->b + diff_context); lowc = MAX(1, cvp->c - diff_context); upd = MIN(len[1], context_vec_ptr->d + diff_context); printf("@@ -"); uni_range(lowa, upb); printf(" +"); uni_range(lowc, upd); printf(" @@"); if (flags & (D_PROTOTYPE | D_MATCHLAST)) { f = match_function(ixold, cvp->a - 1, f1); if (f != NULL) printf(" %s", f); } printf("\n"); /* * Output changes in "unified" diff format--the old and new lines * are printed together. */ for (; cvp <= context_vec_ptr; cvp++) { a = cvp->a; b = cvp->b; c = cvp->c; d = cvp->d; /* * c: both new and old changes * d: only changes in the old file * a: only changes in the new file */ if (a <= b && c <= d) ch = 'c'; else ch = (a <= b) ? 'd' : 'a'; switch (ch) { case 'c': fetch(ixold, lowa, a - 1, f1, ' ', 0, flags); fetch(ixold, a, b, f1, '-', 0, flags); fetch(ixnew, c, d, f2, '+', 0, flags); break; case 'd': fetch(ixold, lowa, a - 1, f1, ' ', 0, flags); fetch(ixold, a, b, f1, '-', 0, flags); break; case 'a': fetch(ixnew, lowc, c - 1, f2, ' ', 0, flags); fetch(ixnew, c, d, f2, '+', 0, flags); break; } lowa = b + 1; lowc = d + 1; } fetch(ixnew, d + 1, upd, f2, ' ', 0, flags); context_vec_ptr = context_vec_start - 1; } static void print_header(const char *file1, const char *file2) { const char *time_format; char buf[256]; struct tm tm1, tm2, *tm_ptr1, *tm_ptr2; int nsec1 = stb1.st_mtim.tv_nsec; int nsec2 = stb2.st_mtim.tv_nsec; time_format = "%Y-%m-%d %H:%M:%S"; if (cflag) time_format = "%c"; tm_ptr1 = localtime_r(&stb1.st_mtime, &tm1); tm_ptr2 = localtime_r(&stb2.st_mtime, &tm2); if (label[0] != NULL) printf("%s %s\n", diff_format == D_CONTEXT ? "***" : "---", label[0]); else { strftime(buf, sizeof(buf), time_format, tm_ptr1); printf("%s %s\t%s", diff_format == D_CONTEXT ? "***" : "---", file1, buf); if (!cflag) { strftime(buf, sizeof(buf), "%z", tm_ptr1); printf(".%.9d %s", nsec1, buf); } printf("\n"); } if (label[1] != NULL) printf("%s %s\n", diff_format == D_CONTEXT ? "---" : "+++", label[1]); else { strftime(buf, sizeof(buf), time_format, tm_ptr2); printf("%s %s\t%s", diff_format == D_CONTEXT ? "---" : "+++", file2, buf); if (!cflag) { strftime(buf, sizeof(buf), "%z", tm_ptr2); printf(".%.9d %s", nsec2, buf); } printf("\n"); } } /* * Prints n number of space characters either by using tab * or single space characters. * nc is the preceding number of characters */ static void -print_space(int nc, int n, int flags) { - int i, col; +print_space(int nc, int n, int flags) +{ + int col, newcol, tabstop; - col = n; + col = nc; + newcol = nc + n; + /* first, use tabs if allowed */ if ((flags & D_EXPANDTABS) == 0) { - /* first tabstop may be closer than tabsize */ - i = tabsize - (nc % tabsize); - while (col >= tabsize) { + while ((tabstop = roundup(col + 1, tabsize)) <= newcol) { printf("\t"); - col -= i; - i = tabsize; + col = tabstop; } } - printf("%*s", col, ""); + /* finish with spaces */ + printf("%*s", newcol - col, ""); }