diff --git a/usr.bin/diff/diff.c b/usr.bin/diff/diff.c index 03eb16211e86..a5966e74dbcc 100644 --- a/usr.bin/diff/diff.c +++ b/usr.bin/diff/diff.c @@ -1,553 +1,563 @@ /* $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 __FBSDID("$FreeBSD$"); #include #include #include #include #include #include #include #include #include #include #include "diff.h" #include "xmalloc.h" bool lflag, Nflag, Pflag, rflag, sflag, Tflag, cflag; bool ignore_file_case, suppress_common; int diff_format, diff_context, status; int tabsize = 8, width = 130; char *start, *ifdefname, *diffargs, *label[2], *ignore_pats; char *group_format = NULL; struct stat stb1, stb2; struct excludes *excludes_list; regex_t ignore_re; #define OPTIONS "0123456789aBbC:cdD:efHhI: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_HORIZON_LINES, OPT_CHANGED_GROUP_FORMAT, OPT_SUPPRESS_COMMON, }; 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' }, { "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 }, { "horizon-lines", required_argument, NULL, OPT_HORIZON_LINES }, { "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 }, { NULL, 0, 0, '\0'} }; -void usage(void) __dead2; -void conflicting_format(void) __dead2; -void push_excludes(char *); -void push_ignore_pats(char *); -void read_excludes_file(char *file); -void set_argstr(char **, char **); +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 *); 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(); 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 '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': 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(); } 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_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(); } break; case OPT_STRIPCR: dflags |= D_STRIPCR; break; case OPT_SUPPRESS_COMMON: suppress_common = 1; break; 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; #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(); if (ignore_pats != NULL) { char buf[BUFSIZ]; int error; if ((error = regcomp(&ignore_re, ignore_pats, REG_NEWLINE | REG_EXTENDED)) != 0) { regerror(error, &ignore_re, buf, sizeof(buf)); if (*ignore_pats != '\0') errx(2, "%s: %s", ignore_pats, buf); else errx(2, "%s", buf); } } 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); } -void +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. */ -void +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. */ -void +static void push_excludes(char *pattern) { struct excludes *entry; entry = xmalloc(sizeof(*entry)); entry->pattern = pattern; entry->next = excludes_list; excludes_list = entry; } -void +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_only(const char *path, size_t dirlen, const char *entry) -{ - if (dirlen > 1) - dirlen--; - printf("Only in %.*s: %s\n", (int)dirlen, path, entry); -} - 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; } } -void +static void usage(void) { (void)fprintf(stderr, "usage: diff [-aBbdilpTtw] [-c | -e | -f | -n | -q | -u] [--ignore-case]\n" " [--no-ignore-case] [--normal] [--strip-trailing-cr] [--tabsize]\n" " [-I pattern] [-L label] file1 file2\n" " diff [-aBbdilpTtw] [-I pattern] [-L label] [--ignore-case]\n" " [--no-ignore-case] [--normal] [--strip-trailing-cr] [--tabsize]\n" " -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" " -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" " [-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"); exit(2); } -void +static void conflicting_format(void) { fprintf(stderr, "error: conflicting output format options.\n"); usage(); } + +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/diff.h b/usr.bin/diff/diff.h index 04e75e631954..7ae700810fc6 100644 --- a/usr.bin/diff/diff.h +++ b/usr.bin/diff/diff.h @@ -1,109 +1,107 @@ /* $OpenBSD: diff.h,v 1.34 2020/11/01 18:16:08 jcs Exp $ */ /*- * 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. * * @(#)diff.h 8.1 (Berkeley) 6/6/93 * $FreeBSD$ */ #include #include #include /* * Output format options */ #define D_NORMAL 0 /* Normal output */ #define D_EDIT -1 /* Editor script out */ #define D_REVERSE 1 /* Reverse editor script */ #define D_CONTEXT 2 /* Diff with context */ #define D_UNIFIED 3 /* Unified context diff */ #define D_IFDEF 4 /* Diff with merged #ifdef's */ #define D_NREVERSE 5 /* Reverse ed script with numbered lines and no trailing . */ #define D_BRIEF 6 /* Say if the files differ */ #define D_GFORMAT 7 /* Diff with defined changed group format */ #define D_SIDEBYSIDE 8 /* Side by side */ #define D_UNSET -2 /* * Output flags */ #define D_HEADER 0x001 /* Print a header/footer between files */ #define D_EMPTY1 0x002 /* Treat first file as empty (/dev/null) */ #define D_EMPTY2 0x004 /* Treat second file as empty (/dev/null) */ /* * Command line flags */ #define D_FORCEASCII 0x008 /* Treat file as ascii regardless of content */ #define D_FOLDBLANKS 0x010 /* Treat all white space as equal */ #define D_MINIMAL 0x020 /* Make diff as small as possible */ #define D_IGNORECASE 0x040 /* Case-insensitive matching */ #define D_PROTOTYPE 0x080 /* Display C function prototype */ #define D_EXPANDTABS 0x100 /* Expand tabs to spaces */ #define D_IGNOREBLANKS 0x200 /* Ignore white space changes */ #define D_STRIPCR 0x400 /* Strip trailing cr */ #define D_SKIPBLANKLINES 0x800 /* Skip blank lines */ /* * Status values for print_status() and diffreg() return values */ #define D_SAME 0 /* Files are the same */ #define D_DIFFER 1 /* Files are different */ #define D_BINARY 2 /* Binary files are different */ #define D_MISMATCH1 3 /* path1 was a dir, path2 a file */ #define D_MISMATCH2 4 /* path1 was a file, path2 a dir */ #define D_SKIPPED1 5 /* path1 was a special file */ #define D_SKIPPED2 6 /* path2 was a special file */ #define D_ERROR 7 /* A file access error occurred */ struct excludes { char *pattern; struct excludes *next; }; extern bool lflag, Nflag, Pflag, rflag, sflag, Tflag, cflag; extern bool ignore_file_case, suppress_common; extern int diff_format, diff_context, status; extern int tabsize, width; extern char *start, *ifdefname, *diffargs, *label[2], *ignore_pats; extern char *group_format; extern struct stat stb1, stb2; extern struct excludes *excludes_list; extern regex_t ignore_re; -char *splice(char *, char *); int diffreg(char *, char *, int, int); void diffdir(char *, char *, int); -void print_only(const char *, size_t, const char *); void print_status(int, char *, char *, const char *); diff --git a/usr.bin/diff/diffdir.c b/usr.bin/diff/diffdir.c index 2b6e5f366454..ecb7c4a6c4ee 100644 --- a/usr.bin/diff/diffdir.c +++ b/usr.bin/diff/diffdir.c @@ -1,239 +1,248 @@ /* $OpenBSD: diffdir.c,v 1.45 2015/10/05 20:15:00 millert Exp $ */ /* * Copyright (c) 2003, 2010 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 __FBSDID("$FreeBSD$"); #include #include #include #include #include #include #include #include #include #include "diff.h" static int selectfile(const struct dirent *); static void diffit(struct dirent *, char *, size_t, char *, size_t, int); +static void print_only(const char *, size_t, const char *); #define d_status d_type /* we need to store status for -l */ /* * Diff directory traversal. Will be called recursively if -r was specified. */ void diffdir(char *p1, char *p2, int flags) { struct dirent *dent1, **dp1, **edp1, **dirp1 = NULL; struct dirent *dent2, **dp2, **edp2, **dirp2 = NULL; size_t dirlen1, dirlen2; char path1[PATH_MAX], path2[PATH_MAX]; int pos; edp1 = edp2 = NULL; dirlen1 = strlcpy(path1, *p1 ? p1 : ".", sizeof(path1)); if (dirlen1 >= sizeof(path1) - 1) { warnc(ENAMETOOLONG, "%s", p1); status |= 2; return; } if (path1[dirlen1 - 1] != '/') { path1[dirlen1++] = '/'; path1[dirlen1] = '\0'; } dirlen2 = strlcpy(path2, *p2 ? p2 : ".", sizeof(path2)); if (dirlen2 >= sizeof(path2) - 1) { warnc(ENAMETOOLONG, "%s", p2); status |= 2; return; } if (path2[dirlen2 - 1] != '/') { path2[dirlen2++] = '/'; path2[dirlen2] = '\0'; } /* * Get a list of entries in each directory, skipping "excluded" files * and sorting alphabetically. */ pos = scandir(path1, &dirp1, selectfile, alphasort); if (pos == -1) { if (errno == ENOENT && (Nflag || Pflag)) { pos = 0; } else { warn("%s", path1); goto closem; } } dp1 = dirp1; edp1 = dirp1 + pos; pos = scandir(path2, &dirp2, selectfile, alphasort); if (pos == -1) { if (errno == ENOENT && Nflag) { pos = 0; } else { warn("%s", path2); goto closem; } } dp2 = dirp2; edp2 = dirp2 + pos; /* * If we were given a starting point, find it. */ if (start != NULL) { while (dp1 != edp1 && strcmp((*dp1)->d_name, start) < 0) dp1++; while (dp2 != edp2 && strcmp((*dp2)->d_name, start) < 0) dp2++; } /* * Iterate through the two directory lists, diffing as we go. */ while (dp1 != edp1 || dp2 != edp2) { dent1 = dp1 != edp1 ? *dp1 : NULL; dent2 = dp2 != edp2 ? *dp2 : NULL; pos = dent1 == NULL ? 1 : dent2 == NULL ? -1 : ignore_file_case ? strcasecmp(dent1->d_name, dent2->d_name) : strcmp(dent1->d_name, dent2->d_name) ; if (pos == 0) { /* file exists in both dirs, diff it */ diffit(dent1, path1, dirlen1, path2, dirlen2, flags); dp1++; dp2++; } else if (pos < 0) { /* file only in first dir, only diff if -N */ if (Nflag) { diffit(dent1, path1, dirlen1, path2, dirlen2, flags); } else { print_only(path1, dirlen1, dent1->d_name); status |= 1; } dp1++; } else { /* file only in second dir, only diff if -N or -P */ if (Nflag || Pflag) diffit(dent2, path1, dirlen1, path2, dirlen2, flags); else { print_only(path2, dirlen2, dent2->d_name); status |= 1; } dp2++; } } closem: if (dirp1 != NULL) { for (dp1 = dirp1; dp1 < edp1; dp1++) free(*dp1); free(dirp1); } if (dirp2 != NULL) { for (dp2 = dirp2; dp2 < edp2; dp2++) free(*dp2); free(dirp2); } } /* * Do the actual diff by calling either diffreg() or diffdir(). */ static void diffit(struct dirent *dp, char *path1, size_t plen1, char *path2, size_t plen2, int flags) { flags |= D_HEADER; strlcpy(path1 + plen1, dp->d_name, PATH_MAX - plen1); if (stat(path1, &stb1) != 0) { if (!(Nflag || Pflag) || errno != ENOENT) { warn("%s", path1); return; } flags |= D_EMPTY1; memset(&stb1, 0, sizeof(stb1)); } strlcpy(path2 + plen2, dp->d_name, PATH_MAX - plen2); if (stat(path2, &stb2) != 0) { if (!Nflag || errno != ENOENT) { warn("%s", path2); return; } flags |= D_EMPTY2; memset(&stb2, 0, sizeof(stb2)); stb2.st_mode = stb1.st_mode; } if (stb1.st_mode == 0) stb1.st_mode = stb2.st_mode; if (S_ISDIR(stb1.st_mode) && S_ISDIR(stb2.st_mode)) { if (rflag) diffdir(path1, path2, flags); else printf("Common subdirectories: %s and %s\n", path1, path2); return; } if (!S_ISREG(stb1.st_mode) && !S_ISDIR(stb1.st_mode)) dp->d_status = D_SKIPPED1; else if (!S_ISREG(stb2.st_mode) && !S_ISDIR(stb2.st_mode)) dp->d_status = D_SKIPPED2; else dp->d_status = diffreg(path1, path2, flags, 0); print_status(dp->d_status, path1, path2, ""); } /* * Returns 1 if the directory entry should be included in the * diff, else 0. Checks the excludes list. */ static int selectfile(const struct dirent *dp) { struct excludes *excl; if (dp->d_fileno == 0) return (0); /* always skip "." and ".." */ if (dp->d_name[0] == '.' && (dp->d_name[1] == '\0' || (dp->d_name[1] == '.' && dp->d_name[2] == '\0'))) return (0); /* check excludes list */ for (excl = excludes_list; excl != NULL; excl = excl->next) if (fnmatch(excl->pattern, dp->d_name, FNM_PATHNAME) == 0) return (0); return (1); } + +void +print_only(const char *path, size_t dirlen, const char *entry) +{ + if (dirlen > 1) + dirlen--; + printf("Only in %.*s: %s\n", (int)dirlen, path, entry); +} diff --git a/usr.bin/diff/diffreg.c b/usr.bin/diff/diffreg.c index 94026007a1bf..45821ad96e8c 100644 --- a/usr.bin/diff/diffreg.c +++ b/usr.bin/diff/diffreg.c @@ -1,1683 +1,1666 @@ /* $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 __FBSDID("$FreeBSD$"); #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 isqrt(int); 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 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; context_vec_ptr = context_vec_start - 1; /* * hw excludes padding and make sure when -t is not used, * the second column always starts from the closest tab stop */ 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_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); 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")); } -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); -} - 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; } /* Code taken from ping.c */ static int isqrt(int n) { int y, x = 1; if (n == 0) return (0); do { /* newton was a stinker */ y = x; x = n / x; x += y; x /= 2; } while ((x - y) > 1 || (x - y) < -1); return (x); } 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 = isqrt(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); 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); free(line); return (ret == 0); /* if it matched, it should be ignored. */ } static bool ignoreline(char *line, bool skip_blanks) { if (ignore_pats != NULL && skip_blanks) return (ignoreline_pattern(line) || *line == '\0'); if (ignore_pats != NULL) return (ignoreline_pattern(line)); if (skip_blanks) return (*line == '\0'); /* No ignore criteria specified */ 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; 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]); if (!ignoreline(line, skip_blanks)) 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]); if (!ignoreline(line, skip_blanks)) 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_ptr == context_vec_end - 1) { ptrdiff_t 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 (a > b) { print_space(0, hw + padding , *pflags); } else { nc = fetch(ixold, a, b, f1, '\0', 1, *pflags); print_space(nc, hw - nc + padding, *pflags); } printf("%c", (a > b) ? '>' : ((c > d) ? '<' : '|')); print_space(hw + padding + 1 , padding, *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') { 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 = 0; for (j = 0, lastc = '\0'; j < nc; j++, lastc = c) { 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 { if (diff_format == D_SIDEBYSIDE) j++; printf(" "); } while (++col < newcol && j < nc); } else { if (diff_format == D_SIDEBYSIDE) { if ((j + tabsize) > nc) { printf("%*s", nc - j, ""); j = col = nc; } else { printf("\t"); col += tabsize - 1; j += tabsize - 1; } } else { printf("\t"); col++; } } } 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') { printf("%c", c); col++; } } } } 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); 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: 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; while (pos > last) { 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) { buf[nc] = '\0'; buf[strcspn(buf, "\n")] = '\0'; if (isalpha(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); } } } pos--; } 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)) { f = match_function(ixold, lowa - 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)) { f = match_function(ixold, lowa - 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 buf1[256]; char buf2[256]; char end1[10]; char end2[10]; 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); strftime(buf1, 256, time_format, tm_ptr1); strftime(buf2, 256, time_format, tm_ptr2); if (!cflag) { strftime(end1, 10, "%z", tm_ptr1); strftime(end2, 10, "%z", tm_ptr2); sprintf(buf1, "%s.%.9d %s", buf1, nsec1, end1); sprintf(buf2, "%s.%.9d %s", buf2, nsec2, end2); } if (label[0] != NULL) printf("%s %s\n", diff_format == D_CONTEXT ? "***" : "---", label[0]); else printf("%s %s\t%s\n", diff_format == D_CONTEXT ? "***" : "---", file1, buf1); if (label[1] != NULL) printf("%s %s\n", diff_format == D_CONTEXT ? "---" : "+++", label[1]); else printf("%s %s\t%s\n", diff_format == D_CONTEXT ? "---" : "+++", file2, buf2); } /* * 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; col = n; if ((flags & D_EXPANDTABS) == 0) { /* first tabstop may be closer than tabsize */ i = tabsize - (nc % tabsize); while (col >= tabsize) { printf("\t"); col -= i; i = tabsize; } } printf("%*s", col, ""); }